HSEHealth & Safety

Executive

Development of a methodology to designand evaluate effective risk messages

Electroplating case study

Prepared by the University of Birmingham

for the Health and Safety Executive

CONTRACT RESEARCH REPORT

400/2002

HSEHealth & Safety

Executive

Development of a methodology to designand evaluate effective risk messages

Electroplating Case Study

Professor Judith Petts, Dr Sheila McAlpine &Dr Jaqueline Homan

Centre for Environmental Research and Training

Dr Steven SadhraInstitute of Occupational Health

Dr Helen PattisonThe Medical School

Dr Sandy MacRaeSchool of Psychology

The University of BirminghamEdgbaston

Birmingham B15 2TT

This report presents the findings of an application of the mental-models approach to the design andtesting of risk information. This approach focuses on identifying the beliefs and knowledge of workerscompared to experts. The aim is to identify any deficiencies in knowledge that may affect behaviourand to design more effective risk information. The electroplating sector was the focus of the study andspecifically small firms within the sector.

The study found a relatively good level of understanding of the chemical risks associated withelectroplating amongst platers. However, there was less understanding of the potential long-term risksand confusion over some of the technical terms. Platers’ mental models are strongly affected byexperience and by the oral tradition in the workplace. New risk messages were designed to presentinformation as practical advice expressed in ‘spoken’ language. They proved effective in raisingawareness of the long-term risks and supporting experience.

The report concludes that to rely solely on the mental-models approach for the design of new riskinformation would be inappropriate. It over-emphasises expert knowledge, and under-emphasisesbehaviour and the influence of organisational culture. However, the importance of the user-centredapproach to the design of risk information is stressed.

This report and the work it describes were funded by the Health and Safety Executive (HSE). Itscontents, including any opinions and/or conclusions expressed, are those of the authors alone and donot necessarily reflect HSE policy.

HSE BOOKS

ii

© Crown copyright 2002Applications for reproduction should be made in writing to:Copyright Unit, Her Majesty’s Stationery Office,St Clements House, 2-16 Colegate, Norwich NR3 1BQ

First published 2002

ISBN 0 7176 2245 2

All rights reserved. No part of this publication may bereproduced, stored in a retrieval system, or transmittedin any form or by any means (electronic, mechanical,photocopying, recording or otherwise) without the priorwritten permission of the copyright owner.

iii

CONTENTS

EXECUTIVE SUMMARY V

CHAPTER 1: BACKGROUND 1

1.1 OCCUPATIONAL RISK PERCEPTION AND COMMUNICATION 1

1.2 MENTAL MODELS 2

1.3 INFORMATION SOURCES 3

1.4 SMALL FIRMS & SAFETY 4

CHAPTER 2: RESEARCH OBJECTIVES & METHODS 5

2.1 OBJECTIVES 5

2.2 ELECTROPLATING SECTOR FOCUS 5

2.3 ETHNICITY 7

2.4 RESEARCH COMPONENTS 7

2.5 ACCESSING SMALL COMPANIES 9

CHAPTER 3: MENTAL MODELS - RESEARCH FINDINGS 11

3.1 INTRODUCTION 11

3.2 CHARACTERISTICS OF ELECTROPLATERS 11

3.3 KNOWLEDGE OF THE CHEMICAL HAZARDS 12

3.4 PERCEPTIONS OF THE RISKS 15

3.5 KNOWLEDGE OF TECHNICAL TERMS 18

3.6 SOURCES OF KNOWLEDGE 20

3.7 RELATIVE IMPORTANCE OF DIFFERENT INFORMATION SOURCES 23

3.8 PROTECTION AND CONTROLS 24

CHAPTER 4: TESTING THE NEW MESSAGES 26

4.1 DESIGN OF THE NEW MESSAGES 26

4.2 PILOT TEST 27

CHAPTER 5: EFFECTIVENESS OF THE NEW MESSAGES 31

5.1 OBJECTIVES & METHOD 31

5.2 CHARACTERISTICS OF ELECTROPLATERS 32

5.3 ACCESSIBILITY OF INFORMATION 32

5.4 KNOWLEDGE OF IMMEDIATE & DELAYED HEALTH EFFECTS 33

5.5 EXPOSURE & HEALTH RISKS 34

5.6 PRECAUTIONS AND PROTECTION 37

5.7 PUNJABI INFORMATION 38

5.8 WORKING IN PLATING 39

5.9 PERCEPTIONS OF NEW INFORMATION 42

5.10 USEFULNESS OF NEW INFORMATION FOR TRAINING 43

5.11 USE OF THE NEW INFORMATION 44

5.12 PREFERRED MEANS OF RECEIVING INFORMATION 45

iv

CHAPTER 6: DISCUSSION 47

6.1 INTRODUCTION 47

6.2 ELECTROPLATERS’ KNOWLEDGE AND BELIEFS 47

6.3 SOURCES OF INFORMATION 50

6.4 PROTECTION AND CONTROLS 52

6.5 ETHNICITY 52

6.6 DESIGNING THE NEW MESSAGES 53

6.7 IMPACT OF THE NEW INFORMATION 56

6.8 VALUE OF THE MENTAL MODELS APPROACH 60

CHAPTER 7: CONCLUSIONS 63

CHAPTER 8: RECOMMENDATIONS 65

8.1 GENERIC PRINCIPLES FOR THE DESIGN OF NEW RISK MESSAGES 65

8.2 RESEARCH NEEDS 66

ACKNOWLEDGEMENTS 67

REFERENCES 68

APPENDICES 70

APPENDIX 1 – PLATERS’ QUESTIONNAIRE 71

APPENDIX 2

A – 1ST ROUND EXPERT QUESTIONNAIRE 78

B – 2ND ROUND EXPERT QUESTIONNAIRE 87

APPENDIX 3 – PHASE 2 PLATERS’ QUESTIONNAIRE 96

APPENDIX 4 – NEW RISK MESSAGES 103

v

EXECUTIVE SUMMARY

This study has tested the potential application of the mental-models approach to the design of new information about occupational chemical risks. It is one of two studies funded by the Health and Safety Executive (HSE) for this purpose, the second undertaken at the University of East Anglia. The mental-models approach (Chapter 1.3) aims to understand individuals’ knowledge, beliefs and attitudes and to compare these with those of experts. Where deficiencies of knowledge and beliefs are identified new information is developed to enable people to make informed decisions about risks to which they are exposed. The electroplating sector was chosen as the focus of the study, and specifically small firms in the sector. Electroplating has well-known, accepted and significant occupational risks, defined controls and a range of formal information and guidance available in the workplace. Chromic acid was selected as the chemical focus, being a Category 1 carcinogen and used in the majority of plating firms. Electroplating was considered to provide a good ‘indicator’ sector for testing of the mental-models approach. The method (Chapter 2) largely followed previous applications of the mental-models approach. Initial, face-to-face, interviews (21) were conducted with managers and workers in small electroplating companies in the West Midlands. A structured questionnaire was distributed to test the prevalence of the beliefs identified from the initial interviews amongst a larger sample of electroplaters (84 responses). The study specifically included platers from the Asian communities. Culture and language are thought to have significant impact on the understanding and interpretation of risk messages. An expert questionnaire was designed to collect ‘expert’ beliefs. The ‘experts’ included factory inspectors, occupational health specialists within HSE and independent occupational safety and health consultants (13 participants). The expert questionnaire was designed as a two-step process, the aim being to identify consensus views and reasons for disagreement. This questionnaire replaced the interview process used to construct expert influence diagrams in other studies. Finally, new risk messages were designed to respond to the apparent deficiencies in beliefs amongst platers and tested in a small number of firms (Chapter 4). The study shows (Chapter 3) that electroplating is characterised by a workforce which is predominantly male and very stable, the majority of people having worked in plating for more than 10 years. This length of experience and the evident nature of the risks which platers face day-in and day-out impacts on their beliefs and knowledge. Platers are aware of the hazardous nature of the chemicals which they work with, but do not draw distinctions between hazards and risks and talk less readily about the effects of exposure and how this might be caused. There is a degree of acceptance of the risks and faith in the precautions. The questionnaire confirmed that platers have a poorer understanding of the potential long-term effects than the acute/short-term effects. They have a very good knowledge of the latter based on experience. They are aware of the importance of personal protective equipment to deal with the acute risks of burns, etc. Experts place greater emphasis on the long-term as opposed to short-term risks in terms of seriousness. A significant minority of platers displayed deficiencies of understanding of key technical terms – such as Category 1 carcinogen, oxidising and sensitisation by skin contact. The latter phrase also revealed different expert interpretations, which served to emphasise the fact that a single expert mental model does not exist.

vi

Platers stress that written information on the risks is important to them and is used, to a greater extent than the experts perceived. However, colleagues are a more important source of information than written sources, emphasising the importance of the oral tradition (Chapter 5). There is overall agreement that information in lay language is needed. New risk messages were designed (Chapter 4 and Appendix 4) to complement Safety Data Sheets, but to provide practical guidance on safe working with substances in the situations which occur in day-to-day working. The new material recast Safety Data Sheet information as practical advice but expressed in written language as spoken by platers. The new messages were subject to a test based on a pilot oral protocol (Chapter 4). Following this initial test further formats were produced to complement the A4 sheets (Appendix 4), including an A2 poster. In addition the information was translated into Punjabi including an audio-tape to provide for a majority of Asian platers working in plating who do not have English as their first language. The new information was distributed to all companies who had participated in the phase one survey. After a period of 5 months to allow for day-to-day use, a more extensive survey to gauge the effectiveness of the new information was undertaken, including a questionnaire and open-ended nterviews. The quantitative results from the initial oral protocol, in terms of correct statements made before and after reading the new messages compared with Safety Data Sheet information, did not suggest a significant impact of the new messages. However, the qualitative results showed a significant preference for the latter. The phase two study (Chapter 5) did identify a better understanding of some of the long-term health effects, particularly lung cancer, amongst those who had read the new information compared to those who had not. The new messages also had an impact in terms of raising awareness of practical precautions to take in the plating shop that were not based in direct experience. Again the qualitative data from the interviews identified significant preference for the new information both in style and format. The new information was regarded as particularly useful for training The mental-models approach was known to present a number of problems (Chapter 6.8), not least in undervaluing lay knowledge and beliefs, in assuming that expert beliefs are rational and in not addressing behavioural and organisational culture issues. The measurement of significant effects of new risk messages was known to be difficult. Despite the positive response to the new messages the study has confirmed the difficulties of measuring the effects of information. While the application of the approach to the electroplating sector has proved possible, the report describes (Chapter 8) 16 generic principles for the design of risk messages to overcome the deficiencies of the mental-models approach. In summary these emphasise: • The user-centred approach to the design of risk messages. • The value of face-to-face discussion to understand people’s knowledge and beliefs– i.e. the

‘conversation with a purpose’. • The need to focus not only on what people believe, but why, and how they reduce risks. • The need to recognise the potential for, and to elicit, divergence of views due to age, length

of experience, culture (ethnicity), and possibly disability. • The need to understand different expert mental models and not to assume that an expert’s

model is necessarily better than that of a worker. • The need for risk information to answer key questions – i.e. what can it do to me? how

might this happen? how can I protect myself? • The need to base risk information on what people already know and believe. • The need to present risk information in lay language and to respond to the needs of those

whose first language is not English. • The value of specialist design input.

vii

Finally the project identifies a number of outstanding research requirements (Chapter 8): (i) Investigation of the impact of knowledge and beliefs on behaviour (ii) Identification of the use made of printed information sources within companies. (iii) Analysis of the impact of the incorporation of information into training and of training

material on practical experience and accidents. (iv) Analysis of the extent to which mistaken beliefs about the properties of chemicals make

a real or potential difference to the way that employees behave. (v) Identification of methods for evaluating the effectiveness of information and training. (vi) Comparison of the chemical risk perceptions of different experts (vii) Analysis of differences of beliefs, knowledge and behaviour amongst ethnic minorities. (viii) Analysis of differences of beliefs, knowledge and behaviour amongst the old and the

young, the able and the disabled. (ix) Analysis of different mental models about specific chemicals between industrial sector.

Printed and published by the Health and Safety ExecutiveC30 1/98

viii

1

1.BACKGROUND

1.1 OCCUPATIONAL RISK PERCEPTION AND COMMUNICATION Communication, as a vehicle for attitude change, is known to be a key organisational process - organisations with good safety records generally have effective communication systems. However, communication in this context covers a range of activities - formal and informal, defined activities and communication which is an inherent part of the management structure and culture. Written communication is one element of these activities. Indeed, definitions of risk communication now stress a two-way process of listening and engaging people rather than simply a one-way process of information provision from "expert" to any other party (NRC, 1989; ILGRA, 1998). The goal of risk communication is more than just imparting scientific facts. It is also about ensuring that people fully understand risks and that they are able to make informed decisions under conditions of uncertainty (Nicholson, 2000. The literature, and often experts, have tried to draw distinctions between risk and hazard communication. Risk communication normally describes the direct transmission of information concerning the probability that an adverse event will occur (NRC, 1989; ILGRA, 1998). Hazard communication is the dissemination of information that a chemical substance, physical or biological agent or physical condition can lead to harm if exposure occurs, without indicating any probability of this harm occurring. However, much of the occupational risk literature focuses on safety communication related to what people should actually do to minimise the potential for harm to occur. It is argued that safety communication has no value unless it actually informs or changes behaviour so as to ensure safety, therefore, it is about "winning hearts and minds" (Cox & Tait, 1998). The communication process is founded on the original source-transmitter-receiver model of communication (Lasswell, 1948) but is extended to address the phenomenon by which information processes, institutional structures, social group behaviour and individual responses shape the experience of risk and the resulting risk consequences (Renn, 1991; Cox and Tait, 1991; Cox and Flin, 1998). Important are, (i) the credibility, skill and knowledge of the communicator (Covello, 1992), (ii) the relevance of the message to the needs and circumstances of the receiver and the perceived importance of the message, (iii) the form in which the message is delivered (the power of face-to-face communication in the workplace being unchallenged - Lee, 1995), (iv) the characteristics of the receiver (including age, intelligence, personality), (v) the receiver’s values, needs and interests, (vi) the extent and impact of safety subcultures (varying by age, seniority, occupation, prior accident involvement (Mearns et al, 1998)), and (vii) safety attitudes, including familiarity with hazardous activities/tasks (e.g. Rundmo, 1992)) and organisational factors (physical working conditions, job stress and commitment and involvement in safety (e.g., Cox et al., 1997; Rundmo et al., 1998)). Even when given the same information on workplace hazards people's responses are known to vary. Different behavioural effects may result from differences in perception by different groups of employees, such as managerial and non-managerial groups or groups that differ in experience, attitude, personality and skill, context and hazard knowledge (ACSNI, 1993; Ferguson et al, 1994; Glendon & McKenna, 1995; Holmes et al, 1997; Cox & Tait, 1998). However, as noted in the Workshop which preceded the letting of the contract for this research project by the Health and

2

Safety Executive (HSE) (Anon, 1999), there has been relatively little work investigating the variables that underlie these broad factors. Much public risk communication research now seeks to understand and analyse risk-related decisions and behaviour in society so as to mediate at the public-expert interface (e.g Fischhoff, 1995; Leiss, 1996; Petts, 1997; Löfstedt and Renn, 1997). Similarly, in the occupational environment it is argued that campaigns to change safety attitudes must be developed in a manner that addresses underlying problems. They, thus, need to be based upon, (i) an analysis of existing attitudes, their architecture and "drivers", (ii) appropriately designed and focused messages, (iii) preliminary testing to see if they have a reasonable chance of succeeding (Lee, 1987), and (iv) evidence of overall success, i.e., impact on safety performance. Other challenges for communicators include how information ‘overload’ can be managed and how to impart more than just scientific facts (Eiser, 1998).

1.2 MENTAL MODELS For this study the HSE favoured the use of mental models. This was seen as allowing for a user-centred approach to the formulation of risk communication media particularly at the micro (company/organisation/industrial sector) level. The psychometric paradigm has provided a useful framework for conceptualising the contextual elements that influence public or lay risk perceptions. However, it gives a less complete picture of how people reason about specific risks and how new information about a risk is integrated with what they already know, perceive or believe. The notion of a "mental model" (Craik, 1943) seeks to explain how the manipulation of internal representations of reality are used in the course of thinking about the world. Mental models are cognitive tools that allow people to reason and put into order what would otherwise be an incomprehensible and disorderly world. They are a primary mechanism by which individuals construct explanations of events and expertise (MacGregor & Fleming, 1996). As such, people’s mental models are comprised of a system of knowledge, attitudes, beliefs, impressions and images (Johnson-Laird, 1983). They reflect the fact that in everyday life people apprehend reality in two fundamentally different ways, one described variously as intuitive, experiential, non-verbal, narrative, automatic, and the other analytical, deliberative, verbal and rational (Epstein, 1994). A mental-models approach to research aims to understand individuals’ cognitive tools. It starts from the premise that people process new information within the context of their existing beliefs. If they know nothing about a topic then a new message will be incomprehensible. If they hold erroneous beliefs ("bugs" - (Fischhoff et al, 1997)) then a new message may be misconstrued. Therefore, knowledge of the extent and nature of a receiver's knowledge and beliefs is essential to the design of effective messages. The latter can be defined as a message that will not be ignored, dismissed or misinterpreted, and one which supports risk-reducing behaviour. However, mental models do not need to be accurate to be useful. It is argued that people can still derive plausible explanations for experiences from a model that bears little resemblance to that of another (say expert’s) model (MacGregor & Fleming, 1996). Using the mental-models research approach should assist in the identification of the means of presenting information to people to contribute to (but not provide the sole mechanism for) making informed judgements. The approach, as used, for example, by Jungermann et al (1988) in relation to drugs; MacGregor (1989) in evaluating product warning labels, Morgan et al., (1992) in a study of radon perceptions, and MacGregor et al (1994) looking at electro-magnetic fields (EMF), is based upon: (i) elicitation of people's beliefs about a hazard allowing expression of both correct and incorrect concepts, (ii) structured questionnaires designed to determine prevalence of these beliefs, (iii) comparison of these beliefs with expert understanding of the risks, (iv) development of communications based upon what people need to know to make informed decisions and a

3

psychological assessment of their current beliefs and (v) iterative testing of successive versions of the communications using open-ended, closed-form and problem-solving instruments administered before during and after receipt of the messages. Comparison of "lay" perceptions with expert beliefs is important, since the objective is to present information that bridges the gap between the "at risk" population's beliefs and expert understandings of risk. However, given the view that accuracy of the mental model does not necessarily impact on utility it would be wrong to assume that any mismatch between lay and expert beliefs is automatically detrimental to appropriate lay response and action. Furthermore, it must be noted that information is always communicated in a social/cultural context that may predispose the information to be misconstrued, misunderstood or ignored even if its purpose is to bridge this gap. In the workplace the social/cultural context is not only that of the company but is influenced also by the "domestic" context in which people live, including the influence of other information sources such as family and friends and the media (Nicholson, 2000). Effective risk communication must cope with lack of knowledge of science, scientific complexity, uncertainties, mistrust of technology and the evident difference of opinion between experts. The relationship between grounded and mediated beliefs, values and perceptions and those which are affected by the workplace culture and climate must be a key question in understanding risk responses, but it is a relationship relatively poorly studied. Work analysing the effectiveness of risk communication in relation to radon risks (Morgan et al., 1992) has highlighted the importance of information that emphasises action and facilitates inferences. This raises fundamental questions about the definition of "effective" risk and risk control messages. For example, is effectiveness a matter of outcomes - such as changing or supporting risk control behaviours? Or should it be defined in terms of the arguably more limited, but fundamental, requirement of promoting personal and group knowledge and understanding of risks? Evaluation of the impact of risk communication is a key methodological issue in occupational risk communication (Schulte et al, 1993), as it is in public risk communication. While this research did not extend to the measurement of behaviour and safety performance, the latter is important not only in the context of HSE's responsibilities but also against the background of the Healthy Workplace Initiative of the Department of Health, which stresses the importance of access to occupational health advice and support.

1.3 INFORMATION SOURCES There are numerous vehicles for formally communicating risk information in the occupational environment - notices, posters, in-house bulletins, information sheets, circulars, safety committee minutes, incident and near-miss reports, meetings and team briefings. This research has not attempted to evaluate the potential impact of information from multiple sources such as these. In the context of chemical risks the HSE wished to focus this work on suppliers’ labels and Safety Data Sheets which provide an important statutory source of risk information, being required by the Chemicals (Hazard Information and Packaging for Supply) Regulations 1997. The regulations require chemical suppliers to identify the hazards of chemicals, provide hazard information and to package them safely. The Control of Substances Hazardous to Health Regulations (1999) require employers to assess, and to inform all employees about risks to their health to which they may be exposed and how they could be minimised. Potentially important information sources, albeit primarily focused on key chemicals, are the HSE's own chemical risk leaflets (e.g. Working with Chromium Are You at Risk?), Guidance Notes (e.g. Chromium and Its Inorganic Compounds: Health Hazards and Precautionary Measures), and Information Sheets (e.g. Health Surveillance Requirements in the Electroplating Industry), and information provided directly by Trade Associations (e.g. Chromium Code of Approved Practice published by the Metal Finishing Association) or in trade magazines.

4

The important point about information sources in the occupational environment is that they have rarely been evaluated in terms of their impact on a specific audience’s understanding: hazard warning labels being a notable exception (Wogalter & Laughery, 1996; Wogalter et l, 1999). There is a tendency to focus on distribution and awareness success as opposed to impact.

1.4 SMALL FIRMS AND SAFETY The conclusions of the 1999 Workshop (Anon, 1999) were that risk messages may need to be tailored to the meet the needs and understanding of the target audience. Small firms were considered a potential target in this respect, being perceived to be (i) isolated from advice and guidance, and (ii) reluctant to ask questions of regulators for fear of enforcement action relating to incorrect behaviour. Research relating to small and medium-sized enterprises (SMEs) and the environment, which also partly addressed health and safety compliance attitudes (Petts et al, 1998), suggests that small companies are unlikely to have specialist staff with dedicated health and safety functions, specialist chemical risk knowledge and access to information. A possible exception is where management have come from larger companies - a trend noted in some sectors. Whilst the culture of health and safety compliance is likely to be positive, a negative safety climate may exist in terms of ability to perform. The majority of research has focused on larger companies. However, given that companies with less than 250 employees represent 99% of total businesses - 38% in the size range 10-249 employees, with micro firms (those with up to 9 employees) representing about 60% of all businesses and employing 56% of all employees - it is right that focus should be upon their information needs to ensure health and safety. There is limited information concerning how micro firms, in particular, respond to health and safety messages or even how they understand the role of the HSE. There are, however, known problems in accessing small companies in research, not necessarily because of a lack of interest in the topic, but a lack of time to take part, and because they are often under significant market and business pressures.

5

2. RESEARCH OBJECTIVES AND METHOD

2.1 OBJECTIVES

The HSE's objectives provided the framework for the research, i.e.: to develop a methodology to

(a) evaluate the effectiveness of written messages on risk and risk control, and (b) design written messages on risk and risk control

The focus was the methodology for designing and evaluating risk information that could be replicated for different industry sectors and risk scenarios. It was considered that it might be possible to provide a methodological ‘blueprint’ for HSE and industry for how they should go about generating and formulating health and safety warnings and hazards guidance information. The purpose was not to provide a full understanding of the factors affecting workplace perceptions of risks and behavioural response. This was important in terms of focusing the research in a relatively short time frame (9 months for phase 1 and 4 for phase 2). It also pointed to the need for supporting objectives, i.e.: (i) To identify an effective method of understanding how management and non-management

in small companies perceive chemical risks to which they are exposed, including their knowledge, attitudes, and understanding of information supplied to them.

(ii) To evaluate an objective test of management and non-management understanding of chemical-related risk messages and to identify the key factors affecting this understanding.

(iii) To design new, sample, risk messages that could be incorporated into existing information sources such as Safety Data Sheets based upon the findings of (i) and (ii) above and upon identified discrepancies between employee and expert understanding of objective risks.

(iv) To test a method for evaluating understanding of the new risk messages amongst the same population.

(v) To present conclusions and recommendations to the HSE on the method as implemented for (i)-(iv) and the results achieved, in particular addressing the potential extension of the methodology to other risks, socio-cultural contexts, audiences and media.

2.2 ELECTROPLATING SECTOR FOCUS Electroplating is an important sector which has many small companies. The majority of firms in Britain employ less than 20 people. The Surface Engineering Association's (SEA) 1999 statistics indicated that some 30% of the UK's surface engineering activity is based in the Midlands, the Region selected as the focus for this research. Electroplating is a chemical or electrochemical process for applying a metallic layer to the product – for example, nickel to protect against corrosion, hard chromium to improve surface properties and gold for aesthetic purposes. The coatings are produced as a result of the passage of a low voltage current through a specially formulated solution. The components to be plated act as a cathode in an electrolytic cell: the anode normally the metal that is to be deposited completes the circuit. The study focused on decorative and hard chromium plating.

6

Decorative chromium is used as a top coat for bright nickel, and the dual plating provides the bright, shiny, mirror-like finish as well as corrosion resistance on everything from cars to bicycles, toys, furniture appliances and plumbing fixtures. Hard chromium is applied from slightly different plating solutions and is thicker, designed normally for engineering purposes. Hard chromium plating provides wear resistance on parts such as pistons, cylinders, cutting tools, etc. Bright nickel contributes to corrosion resistance, brightness and cost effectiveness as the middle layer within the decorative copper/nickel/chromium system as well as the undercoat for brass, silver and gold fancy finishes. Importantly, electroplating involves exposure to chemicals with well-established associations with occupational ill-health, and is the subject of official guidance and codes of practice. The most common electroplating processes are chromium (hexavalent) (referred to as chromic acid throughout this report) and nickel. Chromic acid is a Category 1 respiratory carcinogen (as classified by the International Agency for Research on Cancer (IARC)) – i.e. it has the potential to cause cancer in humans by inhalation. A recently published study (Sorahan & Harrington, 2000) suggests that platers working with chromium may be at increased risk from death due to lung cancer. Chromic acid also has the potential to cause respiratory sensitisation by inhalation, which can lead to occupational asthma; severe burns and chrome ulcers by skin contact; serious eye damage from splashes; damage to the nose, including ulcers and holes in the septum (the flap of tissue separating the nostrils); irritation to the lungs, and kidney damage, and occupational contact dermatitis and skin sensitisation. Chromic acid has the potential to create a fire hazard as it is a strong oxidising agent. Because of its high toxicity, chromic acid has been assigned a Maximum Exposure Limit (MEL) in the workplace based on an 8-hour reference period, of 0.05 mg/m3 . This means that the exposure should be reduced as far below the limit as is reasonably practicable and should not exceed the limit when averaged over the 8 hours. HSE’s Guidance Note EH2 on Chromium points out that electroplating is an activity that requires special attention to the assessment of exposure to chromic acid. Work carried out in confined spaces or locations with poor ventilation is likely to result in higher exposures. Provision of effective ventilation control or treating the plating solution with a mist suppressant is advised to achieve adequate control. Personal Protective Equipment (PPE) which includes protective clothing, footwear, and eye protection should be provided and worn. There is guidance on work planning and housekeeping (e.g. providing facilities to separate contaminated from clean and work from street clothes; personal hygiene practices; washing of floors and work surfaces) in electroplating operations and requirements for the monitoring airborne exposure levels and health surveillance. There is the potential for substitution of chromic acid by compounds of a lower toxicity such as trivalent chromium compounds. In decorative chrome-plating trivalent compounds have been found to be an effective replacement. However, resistance to substitution is often based on concerns about product quality, customer resistance and additional costs. Nickel compounds are also classified by IARC as a Category 1 carcinogen. Compounds such as nickel sulphate and nickel chloride used in plating are harmful if swallowed and may cause an acute reaction such as nausea. Skin contact may cause irritation dermatitis or allergic dermatitis among individuals already sensitised to nickel. Excessive exposure to nickel as mists may irritate the eyes and the respiratory tract and may cause an allergic reaction if inhaled. In the light of these well-known, accepted and significant occupational risks, the defined controls, and the range of formal information and guidance available in the workplace it was considered that electroplating would form a good ‘indicator’ sector for application of the mental-models research approach. Other work for HSE (by WS Atkins, on The Characteristics of Individuals who use Chemicals in Small Enterprises), has suggested that electroplaters are consistently better at understanding chemical symbols and hazard phrases compared to those employed in sectors such as

7

hairdressing, drycleaning, woodyards and garages, and that electroplaters can correctly identify the short-term health effects.

2.3 ETHNICITY The research has attempted to deal with ethnicity in the light of the known significant number of companies in the Midlands employing people from the ethnic minorities and, particularly in plating, from the Indian subcontinent. Culture and language are thought to have a significant impact on the understanding and interpretation of risk messages, although insufficient work has been undertaken in the UK in this regard. This research has specifically identified Asian owned/managed companies where the majority of the workforce is Asian, and also companies where only a small percentage of the workforce is Asian. The new risk messages tested were translated into Punjabi in an attempt to respond to those who do not have English as their first language. This is the first time that such information has been made available in another language and also formats (in this case including an audio-tape).

2.4 RESEARCH COMPONENTS The mental-models approach is based upon a five-step process: (i) elicitation of people's beliefs about a hazard allowing expression of both correct and

incorrect concepts; (ii) structured questionnaires designed to determine how prevalent these beliefs are amongst a

study group; (iii) comparison of these beliefs with expert understanding of the risks; (iv) development of new messages and communications based upon what people need to know

to make informed decisions, particularly drawing upon differences between their beliefs (ii) and expert understanding (iii), and

(v) iterative testing of successive versions of the new communications, comparing old and new knowledge and beliefs.

Therefore, the research project comprised five main elements. (i) initial face-to-face interviews in a small sample of electroplating companies; (ii) a structured questionnaire; (iii) an expert questionnaire, and (iv) design and pilot testing of new messages (v) evaluation of the effectiveness of the new messages/information after a period of use by

companies. The project was completed in two phases: phase 1 comprising elements (i)-(iv) over the period October 1999-July 2000, and phase 2 comprising element (v) over the period March-July 2001. The initial interviews used a semi-structured approach with a set of prompt questions to get participants to (i) think about the nature of their job and their work environment, (ii) talk about their view of themselves in relation to their work risks, (iii) describe the hazards that they thought they were exposed to, and (iv) consider whether the risks identified are significant and ways in which they may be controlled and minimised. The prompt questions included length of experience in plating; perceptions of the main safety issues (not just chemicals); perceptions of the hazards presented by the chemicals and the harm that could result; where information on the chemical risks is obtained; the role of written information (particularly Safety Data Sheets) compared to personal

8

experience in influencing people's understanding and perception of the hazards and risks, and the use of personal protective equipment. The interviews were tape recorded and transcribed so that individuals’ own words and statements could be used to inform the development of the second element. In the first phase 21 interviews were conducted with platers and managers in 8 firms (4 with less than 20 employees, 2 with between 20 and 49 employees, and 2 with over 50 employees). The interviews were stopped at 21 when no significantly new comments or views were being heard. One of the interviews was with an Asian manager and three were with Asian platers. Of these four, two of the platers were bilingual native-speakers, having been in England since birth or early childhood, while the manager and the third plater were coping with English as a foreign language. The purpose of the structured questionnaire (see Appendix 1) was to test the prevalence of the beliefs identified from the initial interviews amongst a larger sample of electroplaters. The questionnaire included a section designed to establish facts relating to demographic and other background issues (age, education, number of years in current employment, occupational history, etc.). Attitude statements used Likert-type scales relating to degree of agreement with the statement and also tests of whether statements were true or false. The statements included those elicited in the initial interviews, using similar words as far as possible. They also included "expert" statements about risks as presented in various information sources such as Safety Data Sheets, trade association leaflets and HSE chemical risks information. Other statements derived from the initial interviews related to the sources and consequences of chemical exposure, perceived seriousness of harm, control measures (including use of Personal Protective Equipment (PPE)), management systems and safety culture, and credibility of information sources. The questionnaire was piloted with some people interviewed in the initial phase plus health and safety advisers within the Surface Engineering Association (SEA) and their consultants. 19 firms (out of 54 contacted) agreed to participate in this phase and in total 170 questionnaires were delivered to the firms. This was known to be an overestimate of the likely response rate. Hand delivery and collection of questionnaires was used (rather than a postal approach), to enable the purpose of the survey to be clear to the manager and to optimise the number of participants. In one Asian-owned company the questionnaire was used and completed in an English class which had been organised by the firm. 84 questionnaire returns were received from a total of 15 companies (2 with less than 20 employees, 10 with between 20 and 50 employees, and 3 with over 50 employees). The expert questionnaire (see Appendix 2a & b) was designed to collect expert understanding of the risks of electroplating. 'Expert' was defined to include health expertise (i.e. occupational medicine and hygiene experts in HSE and also independent consultants) and also occupational safety and operational expertise (i.e. factory inspectors). The open-ended questionnaire was designed as a two–step process, the aim being to identify consensus views and also reasons for any disagreements, relating to the perceived high risks; the factors influencing their occurrence in small electroplating firms; safe working practices; the use and effectiveness of control measures, and use and relevance of different safety information. Some of the results of the first questionnaire were included in the second so that individuals could understand the views of other participants allowing for points of agreement and disagreement to emerge. 15 experts were contacted and 13 took part in the 2-phase process. This questionnaire-based method was chosen instead of the interview-based interative approach which leads to the production of expert influence diagrams relating to the source-transport/pathway-target framework which underpins the risk assessment process. These influence diagrams have formed the basis of previous mental models studies (e.g. Morgan et al, 1992), the aim in part being to help the researchers themselves to understand the hazard sources and risk effects but also to produce diagrams which can be compared with lay beliefs. Some members of the

9

research team had expert understanding and experience of the hazards and risks of electroplating, which meant that the initial interviews with platers could be commenced from a point of existing knowledge. Certainly, to apply the mental-models approach requires an expert knowledge starting point. The process of acquiring this knowledge could not be replicated in any single project. However, a model of this knowledge can be derived. The influence diagram helps linkages between hazards and risks to be graphically illustrated, but it is time-consuming to produce with any degree of detail. The choice of a questionnaire which engaged divergent ‘expertise’ enabled expert information to be gathered relatively rapidly and efficiently (compared with interviews), ensuring structured responses but also iteration and comparison amongst the group. Importantly it engaged experts in a process that could be relatively easily replicated. A Workshop was held at the University of Birmingham in June 2000 to which academics, representatives of participating companies, Trade Association representatives and HSE policy and factory inspectorate staff were invited. The preliminary results of the project were discussed as well as the usefulness and replicability of the mental models approach. The new risk messages (see Appendix 4) were designed to respond to the apparent deficiencies of understanding of chemical risks amongst electroplaters. The rationale for the design is discussed in Chapter 4. There was not time within the project to design and produce a complete new information format – such as a new information leaflet. Such an undertaking also raises issues of access to design consultants. The project, in testing the mental-models approach aimed to test appropriate forms of text and possible formats. The evaluation of the effectiveness of risk messages presents significant methodological issues, as has been recognised in previous mental models work (e.g. Bostrom et al, 1994). Reader-based evaluation methods vary from completely open-ended, in which respondents formulate their own responses, to completely closed in which respondents select from researcher-generated responses. The former have advantages in that they reduce the risk of either underestimating or overestimating people’s understanding, they can measure what ‘sticks’ in people’s minds as well as what information people are able to use to respond to risks. However, they are open to subjective scoring by the researcher. Closed-ended methods are relatively cheap to administer by comparison, but they potentially change people’s beliefs through the cues offered and can restrict the expression of lay beliefs which may be useful to the person’s understanding of appropriate risk reduction responses although they may differ from how an expert would express the risk. The research team chose the open-ended approach (discussed in Chapter 4) to initially test the messages during phase 1. However, the major problem with the evaluation process is that it cannot measure the effect of the new risk messages in terms of changed behaviour. Phase 2 of the project provided an opportunity to combine open-ended and closed-ended methods (Chapter 5) to evaluate the effectiveness of the new messages after a period of use in the workplace on a day-to-day basis. It also provided an opportunity to extend the formats available. Phase 2 was designed to identify whether understanding, particularly of the long-term health effects had been increased and any impacts on working practice.

2.5 ACCESSING SMALL COMPANIES Identifying and accessing small firms who undertake a specific activity is notoriously difficult. The first problem lies in identifying a firm. No single official data source exists which lists companies by industrial activity, size and includes address details. In this case the research team were pleased to received the assistance of the SEA who were able to provide a list of their members and also of non-members, which included a general listing of types of surface engineering activities. However, phone calls to companies were required to confirm whether or not electroplating activities were being undertaken and also the number of people engaged in the

10

activity. In the light of the project’s decision to concentrate on chromium plating so as to provide focus to the discussion of chemical risks given the range of potentially hazardous chemicals used in plating, it was also necessary to contact firms to check what type of plating they undertook on the site. Some companies listed as doing chromium plating did not actually do this themselves but could arrange for it to be done. The second problem lies in the fact that although a company may employ 20 people, it may only have 3-4 people directly involved in the plating activity. Therefore, the number of platers which could be accessed per company was relatively small and placed reliance on being able to access a large number of companies if the sample size, particularly for the questionnaire, was to be meaningful. The third problem linked to the latter point, and common to all such research, was being able to gain the support of companies to undertake interviews and to complete questionnaires. Time limitations and business pressures were the main reasons why companies could not participate. Lack of interest was not an issue. The research team chose to identify the project as being linked to the HSE. This created no real problems, only a few saying that they would not take part in any work related to HSE. It had benefits in that companies and the trade organisation were able to understand that the research would have direct potential use by, and impact on, the regulator. Overall a good level of participation was achieved. However, the time resources required by the research team to this effect were significant.

11

3. MENTAL MODELS: RESEARCH FINDINGS

3.1 INTRODUCTION The following discussion presents the findings from elements (i)-(iii) of the research: i.e. the initial interviews, the electroplaters’ questionnaire and the expert questionnaire. The purpose is to identify people's beliefs about a hazard and the prevalence of these beliefs amongst the study group; and to compare these beliefs with expert understanding and perceptions. The discussion is organised in terms of the characteristics of the respondents; their knowledge of the chemical risks and their perceptions of these risks; their knowledge of the technical terms that form the basis of the information particularly on Safety Data Sheets; how they acquire this knowledge both formally and informally, and finally knowledge and perceptions of how the risks can be reduced.

3.2 CHARACTERISTICS OF ELECTROPLATERS From both the interviews and the questionnaire it emerges that the workforce in this sector is predominantly male and very stable. Apart from one female managing director all those interviewed were men. Around half of those interviewed had worked in plating for 10 years or more. In one small family business bought by the managing director’s father in 1978 the plater had been with the firm for over 40 years; the youngest worker was 21, and had been with the firm since leaving school. Of the 35 employees in an Asian-owned firm, several Punjabi-speakers had been with the firm for the 12 years of its existence. Of the valid questionnaire responses: • 90% were male (N=76) • 49% had worked in their present firm for more than 10 years (N=83) • 63% had worked in electroplating for more than 10 years (N=79) • 39% were aged between 35 and 49 years; 27% were aged 50 and over (N=79) In addition, the questionnaire showed that educational qualifications are low: 49% had no educational qualification (N=84). The questionnaire offered four job classifications: Manager, Supervisor, Plater and Jigger. A further category of “Laboratory worker” was added to accommodate two respondents who did not fit any of these descriptions. Of the 84 respondents 16 were managers (19%). There were 16 supervisors (19%), 35 platers (42%), 15 jiggers (18%) and 2 laboratory workers (2%). In the larger companies these categories tend to be discrete. Management is not in direct contact with the plating shop and may consist of people who have no plating expertise. A supervisor may have responsibility for quality control or health and safety, and spend more time in the office or the laboratory than in the factory. Those who do the plating do not do the preparatory and finishing work. This is the responsibility of jiggers. They attach the pieces to be plated to the jigs which will be immersed in the plating tanks, and remove them after they have been plated and neutralised. In small companies (particularly <10 employees), by contrast, these categories tend to be fluid. A respondent may describe himself as a “hands-on manager”. A supervisor may also work

12

mainly as a plater; a plater may do the whole process including jigging and dejigging. Thus, any way of grouping these categories in order to bring out differences between platers and managers will be imperfect. Also, the small numbers do not warrant analysis of the returns by job category. For these reasons the main results of the questionnaire are reported for all respondents – simply referred to as ‘platers’. However, in the few cases where there is a significant difference between the responses, results are also reported by job category. Where the terms “Manager” and “Plater” are used in a comparison, the definition of “Manager” includes Supervisors, giving a total of 32 potential respondents and the definition of “Plater” includes platers and jiggers, giving a total of 50 potential respondents. The results need to be interpreted with caution as the number of respondents, particularly managers, is small and the identification of different beliefs between managers and platers was not the main purpose of the research. Thirteen of the respondents (16% N=82) were not native speakers of English. Of the other first languages (Punjabi, Urdu, Hindi, Gujarati) Punjabi was mentioned most frequently. Non-native speakers were younger than the native speakers of English and had been less time in their present job and in electroplating. They were more highly qualified, some having studied in India. Of the six people reporting a degree qualification, three were non-native speakers. A fourth had a Certificate in Computer Software.

3.3 KNOWLEDGE OF THE CHEMICAL HAZARDS Eight of the nine companies which gave interviews did some form of chrome plating. One, in the Jewellery Quarter of Birmingham, did nickel, gold, silver and copper. Chromic acid was used in 93% (N=82) of the companies participating in the questionnaire, and by 77% (N=79) of the individual respondents. Nickel was used in 80% (N=79) of the companies, and by 57% (N=75) of the individual respondents. The interviews commenced with a prompt about the general risks (i.e. not just chemicals) in plating. There was some mention of the physical hazards such as moving fork-lift trucks, and one person who thought that working in construction was more hazardous, however, the chemical hazards were perceived as the most important and most evident. Hazards and risks were not differentiated. Most people were aware of the two terms but could not explain the difference. Nor did they regard the distinction as useful. As one managing director put it: Don’t give me two words, give me one. Some experts expressed doubts that the concept of risk was understood by platers – e.g. few platers understand the concept of risk when related to their work

they may understand the hazards associated with the chemcials they use but are unlikely to know the risks they present

However, there was not a consensus amongst the expert group on this issue with some noting that awareness of the hazards and of the precautions to take is in itself important. Platers tended to talk automatically about the fact that the chemicals are “dangerous”, “hazardous”, “serious”, although they usually referred to them as chemicals and acids rather than naming specific chemicals. They talked less readily about what the effects are, or how they are actually caused and often had to be prompted. There was a difference between managers/supervisors and platers in the interviews, with the former more likely to explain in some detail a process being used and to refer to more specific health effects.

13

The acute effects, particularly burns to skin and eyes, splashes and skin contact were often the first things to be mentioned and the most frequently discussed: You can burn yourself If they get chrome on them it is nasty Obviously it can burn your skin Long-term effects were relatively rarely mentioned without prompting and even then people were often unsure about what the effects of long-term exposure might be: Trouble with lungs I suppose I don’t know if it affects your lungs

It can affect your nasal passages but I’m not aware that it can cause any other problems

However, holes through the nose, and ulceration to the nose and septum were mentioned. Protective clothing was sometimes the only control measure to be mentioned. One supervisor summed up his safety instructions: If they don’t touch anything they can’t harm themselves. Overall, the questionnaires suggest a better level of knowledge than the interviews. Table 3.1 shows the responses to a number of statements in the questionnaires testing knowledge of chemical properties. As in all of the following Tables, numbers preceding the statements refer to the question numbers in the main questionnaire (See Appendix 1). N gives the total number who answered the question. Variations in N are due to missing responses to some questions. Of the 10 statements included to test knowledge of the hazards, all except three – numbers 25, 26 and 42 – could be considered “correct”. It should be noted that all were statements generated by platers in the interviews as being correct. Table 3.1 suggests that the majority have correct knowledge of the chemical risks they are exposed to. Statement 25 relating to the risks of solvents and 42 relating to chrome stains on the skin suggest a more significant minority are not aware of the risks compared with the other statements. There is evident confusion in relation to the cancer-causing properties of nickel (statement 43), and whether there is a known occupational exposure limit for chromic acid mist (statement 45) with both statements only generating a slight majority correct response. Of the 9 people incorrectly disagreeing with the statement that ‘Inhaling chrome mists gives you cancer’ (Statement 30), 3 were managers (20% of managers) contrasting with only 6 out of 46 platers/jiggers (13%) who disagreed with the statement. The experts were asked to estimate the significance of the risk of a list of health outcomes from exposure to chromic acid, bearing in mind the seriousness of the outcome and the likelihood of its occurrence. They agreed that nasal septum perforation is the most significant health outcome, with allergic contact dermatitis, and occupational asthma placed second and third. Blindness caused by burns to the eyes was ranked 4th, lung cancer 5th and severe burns to the skin 6th. Burns to the skin were the most frequently recognised risks by platers, being directly experienced. The platers’ questionnaire did not include statements on the first three health hazards, since the interviews suggested relatively low awareness of these specific effects. No

14

one mentioned the potential for asthma in the interviews. One interviewee considered dermatitis to be just as likely to be caused by wearing protective gloves.

Table 3.1 Knowledge of chemical hazards (%)

Statement Agree Disagree N 25. Solvents may make your skin tingle but cause no harm

21

79 82

26. Chromic acid used in plating is not harmful because it is in dilute form

4

96

80

28. Chromic acid is more dangerous than nickel salts

86

14 81

30. Inhaling chrome mists can give you lung cancer

89

11

79

32. Chromic acid burns the skin

93

7

82

34. Nickel produces skin rashes

92

8

80

37. Chromic acid mist causes ulcers in the nose

96

4

80

42. Chrome stains on the skin are not harmful

17

83

78

43. Nickel causes cancer

57

43

68

45. There is no known safe limit for chromic acid mist in the atmosphere

56 44

77

In questions 22 to 28 of the expert questionnaire respondents were asked to rate the degree of risk presented by various substances used in plating relative to the risk presented by chromic acid. The majority of experts rated hydrochloric acid, nickel salts, sodium hydroxide, sulphuric acid and zinc salts as presenting an equal or lower risk. The only point of comparison with the platers’ questionnaire is Question 28, where there is agreement with the expert view (as reported in Table 3.1 above, 86% of platers also regarded chromic acid as more dangerous than nickel salts). Five out of 13 experts rated trichloroethylene as a greater risk. Only cyanide salts were rated more risky than chromic acid by a majority of experts (9 out of 13 – 69%). Again the interviewees often talked in generic terms about the chemicals being used – e.g. as ‘acids’ or ‘solvents’ and chromic acid tended to be a consistent focus. In the light of this and (as identified in Chapter 2.2) the expert concerns about chromic acid as a carcinogen and the large amount of guidance on the chemical it was decided to focus on this in the questionnaire.

15

3.4 PERCEPTIONS OF THE RISKS There was a degree of acceptance of the risks and faith in the precautions. The following are examples of statements made in the interviews:

When you use acids it’s going to cause you some problems It’s not a very nice game to be in. But if you understand the risks, they don't bother you Providing you are not sort of sticking your head in the stuff, I can't really see any harm in it You’ve all the protective gear and so on, there’s not a lot of risks to it really Everything is safe as long as it is handled properly Chemical hazards are serious but they’re well understood When you are working you know what to do So long as you just swill it out immediately its okay

The experts tended to a view that platers are aware of the risks but treat them as an occupational hazard and part of the job (comment in the second round of the expert questionnaire). Table 3.2 compares the responses of experts and platers to a number of statements relating to risk perceptions. The statements were derived directly from comments made during the interviews. Numbers preceding the statements refer to the question numbers in the main questionnaire; numbers in brackets and prefaced by E give the question number in the expert questionnaire. In some cases the formulation was slightly different to allow for the fact that platers were being asked for their own views while experts were being asked for their perceptions of platers’ views. The version given in Table 3.2 is as it appeared in the questionnaire distributed to platers. For the version which appeared in the expert questionnaire, see Appendix 2. N gives the total number who answered the question in the main questionnaire. Question 40 on the platers and expert questionnaires were not identical and are reported separately in the Table. There is most divergence in agreement between platers and the experts where experts have been asked about their perceptions of platers – i.e. whether platers understand the risks of the work that they do (statement 29) and whether plating companies take the risks of working with chemicals seriously (statement 44). The experts have less confidence in the understanding of platers in both cases. For example, in relation to statement 44 (most companies take the risks of working with chemicals very seriously) the follow expert comments were made: They say they do but have little idea Why should they, they think that they are immune to accidents at work Familiarity breeds contempt To most companies working with chemicals is just part of being in the industry

16

There was agreement between experts and platers that there may be special safety problems for workers who do not understand English and that it is easy for platers to become complacent about the risks of the work that they do. The last two statements (46 and 49) were included as direct quotes from interviewees in the initial interviews. While the majority of platers agreed with the experts, it is noticeable that a significant minority (27%) believed that skin rashes are more likely to be caused by wearing gloves than by nickel, although Table 3.1 shows that 92% of platers understand that nickel can produce skin rashes.

Table 3.2 Responses of experts and platers to statements relating to risk perceptions (%)

Agree Disagree N Statement

Experts Platers Experts Platers Platers (Exp)

27. (E32) There’s no special safety problem for workers who don’t understand English

0

14

100

86

78 (13)

29. (E33) I have a good understanding of the risks from the work I do

23

98

62

2

81 (11)

31. (E34) It is easy to become complacent about the risks in my work

100

85

0

14

82 (13)

35. (E36) People don’t take safety seriously until they have an accident

85

85

15

15

82 (13)

38. (E38) Working in plating causes no long-term health problems

15

31

85

69

78 (13)

40. All chemicals are safe as long as they are handled correctly

N/A

84

N/A

16

80

(E40). All chemicals used in plating are dangerous

54

N/A

46

N/A

(13)

44. (E43) My company takes the risks of working with chemicals very seriously

23

74

69

26

77 (12)

46. (E44) Skin rashes are more likely to be caused by wearing protective gloves than by nickel

0

27

100

73

74 (13)

49. (E47) As long as you swill your eyes with water, a chrome splash will not harm you

0

13

100

87

79 (13)

17

The 31% of respondents who did not believe that working in plating causes long-term health problems were made up of 7 out of 14 managers (50%) and 6 out of 16 supervisors (38%), compared to only 11 out of 46 platers/jiggers (24%). This can be compared with statement 30 in Table 3.1 where 89% of platers know that inhaling chrome mist causes lung cancer. This provides one indication of a potential difficulty of perceiving concepts of long-term risks, although the hazard is recognised. In the interviews there was little evidence of understanding of long-term problems.

One question (40) in both the plater and experts questionnaire was asked slightly differently. While 84% of platers agreed that “All chemicals are safe as long as they are handled correctly”, experts were divided as to whether “All chemicals used in plating are dangerous”. Table 3.3 shows platers’ and experts’ estimation of the risks of other jobs and activities compared to plating. This question was included to identify primarily whether platers perceive themselves as facing greater occupational risks than those working in other industrial sectors and also to explore whether people have any understanding of the comparative risks which they face in their occupational compared to their home environments. In the interviews a few comments had been made about the relative risks from other activities: e.g. You’ve probably got more of a hazard off fags. Its health and safety and we are all puffing away with fags. It is important to note that the word ‘risk’ was not defined in the question and therefore how people interpreted this – e.g. risk of death, or risk of injury, or risk of a ill-health effect, etc – cannot be judged. The results are presented for platers (P) and experts (E) in terms of whether there were perceived to be more (greater), similar or less risks from plating compared with the identified activities.

Table 3.3 Risks of plating compared to the risks of other activities

% More % Similar % Less Comparison P E P E P E

Construction site 31 77 46 23 23 -

Steel foundry 34 77 54 23 12 -

Long distance HGV 14 15 23 23 63 62

Clothing manufacture 2 - 11 8 87 92

General engineering 3 - 53 54 44 46

Driving a car 14 15 24 31 62 54

Smoking 41 38 41 54 18 8

P = platers; E=experts Table 3.3 suggests a significant difference between expert and platers’ views of the comparative risks of the two main industrial sectors - construction and steel -, with 77% of experts perceiving these as more risky activities than plating compared with only a third of platers. More detailed analysis of the figures suggests that managers were more likely to perceive these industries as having greater risks (40%) than platers (26%) but significantly less than the experts. It could be expected that the experts should have a better understanding of the relative risks of the different industrial sectors, not least factory inspectors responsible for regulation and investigation of accidents.

18

There was much greater agreement between experts and platers in relation to the other risks, with clothing manufacturing viewed by the majority as less risky. There was also commonality of views in relation to driving a car and smoking, although both actually produce more deaths per year than working in electroplating. Managers displayed a slightly better understanding of the relative risks of smoking than platers or experts (46% saying than smoking was more risky compared to 33% platers and 38% of experts).

3.5 KNOWLEDGE OF TECHNICAL TERMS The following hazard phrases common to Safety Data Sheets were printed on three separate cards and presented to interviewees in the first phase interviews: 1. Chromic acid has been classified as a Category 1 carcinogen i.e. a substance known to be

carcinogenic to man 2. Chromic acid is a respiratory irritant 3. May cause sensitisation by skin contact There was a wide range of responses even in this small sample. In a small family firm both the managing director and the very experienced plater showed full understanding of the statements and the health implications, although chromic acid was not used by the company. By contrast, two Punjabi-speakers, one a manager, the other a plater in a different company, showed no understanding of the technical terms. In a large firm where the manager was conversant with all the technical and medical terms, two platers – one experienced, one inexperienced – understood the second statement, but not the first or the third. In another firm, an experienced plater understood all the statements, but was slightly hesitant about the health implications of sensitisation. One plater thought that the third statement meant the same as the second one. The questionnaire also tested knowledge of hazard phrases in two sections. In the first section Questions 52 to 60 listed phrases taken from Safety Data Sheets and asked respondents to say whether they were true or false in relation to chromic acid. In fact all the statements did apply to chromic acid and so the correct answer was True in every case. Table 3.4 shows the percentage of managers and platers who wrongly answered ‘false’ or who entered a response of ‘don’t know’. The most significant finding from Table 3.4 is the number of people who did not know that chromic acid is a Category 1 carcinogen (39% of managers, 35% of platers). While Table 3.1 indicates that platers are aware that lung cancer can result from inhaling chrome mists, the words ‘category 1 carcinogen' are not necessarily understood. Other significant findings are that 43% of platers did not know that chromic acid is an oxidising agent and 31% that it is assigned an MEL, managers performing better on both of these terms. While both managers and platers were more likely to think that they understood the phrase `sensitisation by skin contact', 10% of platers gave the wrong answer in relation to chromic acid. In the next section of the questionnaire five hazard phrases were given, this time with four possible definitions for each. Respondents were asked to tick one definition which they considered closest in meaning to the hazard phrase. They performed better on this multiple-choice form of the questions compared with the previous true/false questions, although an element of ‘guessing’ can not be discounted. Table 3.5 shows the percentages of correct responses, wrong responses and no response.

19

Table 3.4 % of wrong responses and don’t know to Hazard Phrases

% False % Don’t know

Chromic acid M P T M P T

52. Is toxic if swallowed - - - 3 2 2

53. Can cause severe burns 6 8 7 3 4 4

54. Is a respiratory carcinogen 7 - 2 19 27 24

55. Is a category 1 carcinogen - 4 3 39 35 35

56. May cause sensitisation by skin contact

3 10 8 14 6 9

57. Is assigned a Maximum Exposure Limit (MEL)

3 8 6 19 31 26

58. Is a respiratory irritant 3 - 1 13 13 13

59. Is an oxidising agent - - - 10 43 29

60. Prolonged exposure can cause serious damage to health

-

- - 6 10 8

M= Managers %, P =Platers %, T= Total %

Table 3.5 Responses to definitions of hazard phrases (N=84)

Hazard phrase % Correct % Wrong % No response Respiratory irritant 77 20 2 Category 1 carcinogen 56 25 19 Oxidising agent 51 27 21 Toxic 18 72 11 Sensitisation by skin contact 42 44 14

The expert interpretation of toxic (Can cause serious health effects) was taken as correct. However, 63% of respondents ticked ‘Poisonous’. This lay interpretation is understandable and the multiple choice responses in this case may not have helped as ‘poisonous’ is arguably not completely wrong. When offered 4 possible explanations of the term ‘Category 1 carcinogen’ (likely to cause cancer; highly corrosive; poisonous; contains tar), only 56% chose the right answer, 25% getting the answer wrong (14% ‘poisonous’, 11% highly corrosive’) and 19% not able to answer. It appeared to be the word carcinogen itself that was not known as was also the case with the word ‘oxidising’ (only 51% correct, 21% not able to select from 4 options and 27% wrong). The phrase sensitisation by skin contact caused significant problems – only 42% able to chose the correct answer when offered 4 options (causes blistering of the skin; causes an allergic reaction; causes skin burns; causes the skin to peel), 44% choosing the wrong answer (29% choosing ‘causes skin burns’) and 14% unable to make a choice.

20

Interestingly, amongst the experts a number of different explanations for the term ‘sensitisation by skin contact’ were offered. Only one of the 13 definitions proposed by the experts was not considered to be wrong by other experts in the second round of the expert questionnaire. The following are examples of definitions proposed in the first expert questionnaire and marked wrong or alternative definitions were proposed by about half of the experts in the second round:

If your skin becomes irritated by X then this may become prolonged or affect other parts of your body e.g. lungs

Repeated exposure will overcome your body’s tolerance, so that in future you will suffer a catastrophic reaction to the slightest exposure.

Where there was overlap between these two sections of the main questionnaire some respondents correctly answered that the phrases applied to chromic acid, but in the next section could not define them (e.g 13 respondents in the case of Respiratory irritant; 15 in the case of Category 1 carcinogen). Other respondents knew their meanings, but were not sure if they were true of chromic acid (e.g 8 respondents in the case of Respiratory irritant; 14 in the case of Category 1 carcinogen).

3.6 SOURCES OF KNOWLEDGE Experience and learning from others are the primary means by which platers in particular find out about the risks. Typical responses to the question in the interviews about where information comes from were: I think it’s just what you get over the years

It’s just…learning. Watching and learning. See I’ve been here 9 years this time. You just pick things up

That’s how I learned, from other people

That’s the way I was taught: That’s what that does. Be careful of that. Be a bit more careful of that than that

He always splashed himself. And in a way he learnt the hard way

The experts also commented on the importance of experience: They [platers] are not stupid. They can see and smell the solutions

Older platers who have seen the health problems have [an understanding of the health risks] younger ones do not

They have learnt from experience of others and of their own There was a range of attitudes to written sources of information and comments on their usefulness. Most regarded information sources as valuable, however, not always used. The following statements were made by managers/supervisors in the interviews:

The three platers I’ve worked with have learnt me different things over the years that I couldn’t’ve read up on

21

We have got an acid information folder. I haven’t seen it for a bit

They all know about this book, but they never need to consult it. You know, it’s experience. They’re very experienced people

People’s experience counts for a lot but as far as what I’ve got in documentation form it’s what the supplier sends me

I read the data sheets if it’s something I don’t know about. There is a Canning’s Handbook as well which everybody’s used in the plating industry since the year dot, but I don’t think it’s been updated since about 1968. The risks are still pretty much the same.

The technical manager used to tell me everything: This chemical can do this to you: this is the way to handle it. And I have read a couple of books on chemicals.

Word-of-mouth, I don’t think at times is quite good enough. I think it ought to be backed up by written…documented information

Table 3.6 shows the responses of platers and experts to statements in the questionnaires relating to sources of information. In the case of all expert responses N=13. There was significant disagreement between experts and platers in relation to two of the five statements in Table 3.6. As in relation to perceptions of risks (Table 3.2) experts appeared to be sceptical of company practices and knowledge - only 23% believing that companies make sure that people read safety information, compared to 77% of platers saying that their companies do this. The qualitative elements of the expert questionnaire stressed some of the sceptical views

Table 3.6 Responses of experts and platers to statements relating to sources of information (%)

Agree Disagree N Statement

Experts Platers Experts Platers Platers

33. (E35) Written information about the chemicals is essential to my job

92

94

8

6

81

36. (E37) My company makes sure everyone reads written safety information

23

77

77

23

81

47. (E45) I learned most of what I know about the risks of plating from the people I have worked with

92

74

8

26

81

48. (E46) I don’t need to know what a chemical could do to me; I just need to know how to handle it safely

0

12

100

88

81

50. (E48) A safety data sheet on a substance I use is not going to tell me anything I don’t know already

8

22

92

78

82

A significant minority (22%) of platers believed that Safety Data Sheets cannot add to their own knowledge, which given the length of time people spend in electroplating is not surprising. By contrast, experts remained convinced of the importance of formal information. However, experts were also convinced that platers learn from those they work with, whereas a smaller majority of

22

platers (72%) believed this. The only real point of agreement between platers and experts was that some form of written information is essential (92% of experts and 94% of platers agreeing). Question 41 in the main questionnaire was the statement “Safety data sheets are too complex for me to understand”. The corresponding question to experts was split into two, E41: “Safety data sheets are too complex for most platers to understand”, and E42: “Safety data sheets are too complex for most plating managers to understand”. Twelve of the 13 experts agreed that Safety Data Sheets are too complex for platers, with 8 of the 13 agreeing they are too complex for managers. By contrast platers and managers were more positive, 68% of platers and 88% of managers believing that they are not too complex. This confidence needs to be tempered by the results in Tables 3.4 and 3.5 which indicated problems in understanding some key hazard phrases. Respondents were asked to indicate how frequently they consulted key safety data information. Table 3.7 shows managers’ responses and experts’ estimations of managers’ use of the information. Table 3.8 shows platers’ responses and experts’ estimations of platers’ use of the information.

Table 3.7 Managers’ use of written safety information (%)

Often Now and then Never Managers Experts Managers Experts Managers Experts Labels 84 0 13 77 3 23 Safety Data Sheets 52 15 48 77 0 8 HSE leaflets 41 8 47 69 13 23 HSE inspection reports

23 0 52 85 26 15

Reference books 44 0 41 62 16 38 COSHH assessment 45 8 42 69 13 23 Trade Association literature

31 23 47 62 22 15

Trade magazines 26 23 39 69 35 8 Company H+S literature

43 0 33 0 23 0

Table 3.8

Platers’ use of written safety information (%) Often Now and then Never Platers Experts Platers Experts Platers Experts Labels 76 15 18 85 6 0 Safety Data Sheets 40 0 40 46 19 54 HSE leaflets 31 0 38 54 31 46 HSE inspection reports

19 8 36 31 45 62

Reference books 13 0 23 15 65 85 COSHH assessment 18 0 38 54 44 46 Trade Association literature

6 0 19 31 75 69

Trade magazines 0 0 17 38 83 62 Company literature 27 0 55 0 18 0

23

There is further evidence of the scepticism of experts about the use of information in both Tables, particularly in relation to labels, where only 15% of the experts considered that they were used often by platers and none considered that they were used by managers. However, 84% of managers and 76% of platers said that they used labels often. Indeed, they are identified as the most valuable source of information. HSE leaflets and inspectors’ reports, COSHH assessments and Safety Data Sheets were all considered by experts to be used by managers to a lesser extent than managers claimed. Other noticeable figures are the fact that 26% of managers and 45% of platers never read HSE inspectors’ reports (although this question may have been answered in terms of whether an inspection report had ever been received by a company - it being likely that some had never had an inspection), and the fact that 19% of platers had never used a Safety Data Sheet, although 40% say they use them often. Trade magazines are least important sources of health and safety information to managers and of no importance to platers (83% never using them).

3.7 RELATIVE IMPORTANCE OF DIFFERENT INFORMATION SOURCES In a final section on written sources of information both questionnaires presented 15 of the headings used in Safety Data Sheets (Sections 2-15, with Section 15 split into two, Risk Phrases and Safety Phrases). These questions were considered important in understanding mental models given the importance of the Safety Data Sheets to electroplating and the evident potential problems in their use where key hazard phrases are not understood. Platers were asked to indicate how important (Very important, Important, Not important and Don’t know) each kind of information was to them. Experts were asked to indicate how important they thought each type of information was first for managers and secondly for plating shop workers. The Don’t know box was omitted from the Expert Questionnaire. Most of the headings were marked as important by platers and managers. 12 of the sections were regarded as important by 80% or more of respondents, and of these 7 were valued by 90% or more. The three which attracted less than 80% of support were Ecological information (79%), Transport (74%) and Toxicological information (72%). Similarly, these were the three which fewer experts identified as important to managers, with 77% giving this response in the case of Ecological information and Transport, and only 46% responding that Toxicological information is important for managers. The same three sections were regarded as important to plating shop workers by even fewer of the experts, this time in a slightly different order: Transport (39%), Ecological information (31%), and Toxicological information (15%). In addition to these sections, four others were thought important to plating workers by less than 80% of the experts. These were Stability and Reactivity (77%), Disposal (77%), Physical properties (62%) and chemical composition (54%). However, when asked whether any sections might be removed (if legally possible) from Safety Data Sheets only one of the 13 experts considered this could be done without detracting from the information people need to know, although there was agreement amongst experts that all of the technical detail was not necessary and some was just jargon – e.g LD50 - which may be relevant to specialists only. In several cases over 10% responded ‘Don’t know’ to the question of important information. The Don’t know box was ticked by 20% of platers in the case of Toxicological information; by 15% in the case of Ecological information; by 13% for Physical properties; also by 13% for Stability and reactivity; and by 12% in the case of Safety phrases. There was overall agreement amongst managers, platers and experts that information in lay language is needed. The value of pictograms was also noted. However, written safety information was felt to be sufficient in itself only by the experts. Platers referred to the value of videos and to verbal briefings. Experts did not favour supply companies providing verbal information nor did

24

they value presenting leaflets in other languages, although they agreed with the platers that there may be special safety problems for people whose first language is not English. One accident was mentioned in an interview where a new employee’s poor command of English had been a contributory factor. Procedures had been altered following the accident.

3.8 PROTECTION AND CONTROLS Wearing the protective clothing was seen by both managers and platers as the most important piece of advice for any new plater. In the interviews people said their advice would be, for example, Just wear the protective gear and Protect your skin. Protect the eyes. Some of those interviewed, however, said that their main advice would be to ask an experienced person if unsure of anything. Also, when asked in interview about the relative importance of controls, some people identified ventilation as most important. Also in the questionnaire responses the main advice from platers was to wear protective equipment, with a strikingly contrasting response from the experts. In both questionnaires respondents were asked to tick up to two of the pieces of advice listed in Table 3.9 below as the most important to give to someone new to plating. Table 3.9 shows the percentage of platers and experts who ticked each item. Experts, perhaps reflecting their regulatory responsibilities, supported reading of Safety Data Sheets to a greater extent than platers. Platers do not read Safety Data Sheets to find information on the hazards and risks, their priority is practical measures with a focus on prevention before reading about the risks.

Table 3.9 Most important pieces of advice to a new plater

% Platers % Experts Always wear the protective clothing 68 23 If you’re not sure, ask the person in charge 42 39 Wash off chemical splashes immediately 40 31 Listen to the experienced people 34 31 Read the safety Data Sheets 30 46

As in the interviews there was confidence in barrier methods of control. 32% (N=81) believed that “As long as you wear protective clothing you cannot be harmed”, whereas none of the experts agreed with this statement. However, only 5% (N=82) thought there was no danger if the extractors were not working (Question 51). Only one of the 13 experts agreed that safety equipment in small plating firms is monitored regularly (Question E49). There was expert consensus that poor working practices, poor working conditions and ineffective health and safety training are the most important factors influencing risks in small electroplating firms. Manual addition of chemicals to plating baths, failure to maintain extraction systems and cleaning of plating tanks were considered to be the most frequent causes of accidents or ill health. Control failures were not explored in detail with the platers through the questionnaire, although in the interviews examples were given of accidents, some fatal. As noted above, general and local exhaust and ventilation systems were identified as the important controls, with personal protection for skin and eyes. The following are examples of statements made in the interviews.

25

Without extraction we wouldn’t be able to work there cos you’d have to hold your breath to put the work in the chrome The extraction is important. If it was turned off you can taste it, it gets to your eyes. (asked which control was most important) the extraction fan I’ve had acid in my face. It left a scar for about six months. That’s why we have to wear goggles. You can have what you want (PPE). If you will wear it they will get it for you We have got signs saying safety spectacles shall be worn whilst processing. They are available, but we can’t force them to wear them (manager) Advice: Mainly protect your skin and eyes, be careful. Just listen to the experienced workers

26

4. TESTING THE NEW MESSAGES

4.1 DESIGN OF THE NEW MESSAGES What emerged from the phase 1 interviews and questionnaires was that platers want practical guidance on safe working with substances in the situations that occur in day-to-day working. While this information exists in Safety Data Sheets, it is distributed over sections that are theoretically structured rather than being presented in the order in which people need to know the information. The new materials aimed to recast data sheet information as practical advice, expressed not in written language but in ‘spoken’ language - i.e. written in the way that people speak. It sought to deal with the evident difficulty that platers have in understanding technical phrases such as ‘Category 1 carcinogen’ and ‘skin sensitisation’. Chapter 6 discusses further the logic applied to the new messages. The information was structured in terms of (i) substance form and characteristics, (ii) how exposure occurs, (iii) what the ill-health effects are, and (iv) the precautions to take. The messages aimed to ensure that the long-term risks were explained and also that the logic in the selection of controls was demonstrated. The new messages focused on safe working with chromic acid as this is used by the majority of platers. The same format and approach could be used readily for other chemicals. The new messages consisted of three A4 sheets printed on both sides in colour and laminated so that they were easy to hold and did not become damaged by handling. The content and lay-out are reproduced in Appendix 4. Further formats were added for the phase 2 evaluation (see Chapter 5). The new messages covered broadly the same information as would be included in four specific sections of a Safety Data Sheet, namely Section 3: Hazard identification, Section 8: Exposure controls and personal protection, Section 11: Toxicological information, and Section 15: Regulatory information. These were the sections over which there was full agreement of importance from the questionnaires and also the key sections for conveying chemical risks and risk reduction measures. The style in which this information was conveyed was, however, very different. A question-and-answer structure was used throughout the first two pages. Answers were given in the context of detailed job situations, with an invitation to readers to think of other examples of their own. Technical and medical terms were either replaced by everyday words or explained in everyday words. The aim was to go through the information in the order that people would be likely to need it as they worked, rather than presenting it distributed over a number of formal headings. The first section discussed the physical properties of chromic acid in practical terms (e.g. What does it look like?) rather than in terms of chemical formulae. Then the acute/chronic distinction was formally discussed, using the more accessible terms of ‘immediate’ and ‘delayed’ effects. This distinction was repeated in every answer after that. The material then followed the same pattern in each section of asking three basic questions that platers need to know about – What harm can the substance do? How is this likely to come about? How can it be prevented? A final section listed the appropriate hazard and risk messages with an explanation of their meanings in simple words and a translation of latinate and technical terms. Icons were included where appropriate without further comment.

27

One advantage of the question-and-answer form is that the material could be used as a form of quiz, which makes it potentially engaging, and could be used to check if training had been effective. The quiz form also seemed appropriate, given the popularity of pub quizzes and the relatively well-informed, and experienced target group. The formulation of the material in terms of behaviour in the plating shop was designed to respond to the apparent difficulty people have in making the link between the potential effects of the chemicals and the realisation of these effects due to poor working practice or inappropriate controls. The new material also aimed to respond to our finding that people were less aware of long-term than short-term health effects. Finally, the new material aimed to deal with the evident difficulty people had with technical and medical terms. While it was clear from the interviews and questionnaires that there were comprehension problems, nothing had been done at that stage to explain the things that people did not understand. This was made good by the inclusion of the final page listing and explaining the formal messages which apply to chromic acid. The quiz is designed to complement the existing Safety Data Sheets and to provide a useful training resource. It aims to clarify what has not been understood in a data sheet or to act as a primer which will make data sheets less formidable. It has been formulated for use by both managers and platers (though because of the small numbers involved it was tested only with platers). It is likely to be most useful to workers who do not have access to Safety Data Sheets, for the training of new employees, and for the evaluation of health and safety knowledge and awareness.

4.2 PILOT TEST

4.2.1 Design of Test Procedure As discussed in Chapter 2.4, the design of effective evaluation methods in relation to new risk messages is difficult. Key questions have to relate to what is being evaluated – e.g. is it understanding of the information being provided?, or whether understanding is better given the new information compared to the old?, or is it ease of use of the information?, or is it the effectiveness of the new information in changing knowledge or dispelling misconceptions?, or is it the ability to take appropriate action as a result of the information being provided? The project could not provide for any evaluation of change of behaviour as ‘pre’ behaviour was not measured. Nor could it provide in the first phase for any test of whether the new risk messages were actually used in the workplace as there was not time to leave the material in the companies. Therefore, the focus in the pilot test was whether the new information produced better understanding than the old (i.e. Safety Data Sheets) – i.e. the test was against material which was already available. The open-ended, interview-based approach was selected primarily because it is known to reduce the risk of underestimating or overestimating people’s understanding in cases where their frame of reference differs from that of the expert. A small number of people (target 30) were to be tested before and after reading safety information. There were two groups. 15 were to be tested with the new materials and 15 with excerpts from a Safety Data Sheet for chromic acid. Assignment to groups was random, subject to maintaining a balance of jobs in the two groups. Although it may have been better to include selection criteria, such as experience based on length of time in plating, this was not possible given the number of platers who were available.

28

An oral protocol was developed. The experimental group were asked to read through the new material on the laminated A4 sheets. The control group were presented with a complete Safety Data Sheet, which had been laser printed to enhance quality, and asked to turn to Sections 3, 8, 11 and 15, and read them in turn. These were the sections deemed to include the same information as the new material. The interviews comprised three parts. People were first asked to name the parts of the body which could be harmed by chromic acid and to explain what damage the substance could do and how this could be prevented. Points made were ticked off on a list as they were mentioned. Secondly, participants were given the material to read, either the new messages or the Safety Data Sheet. Thirdly, they were asked to add to their first statements anything which they had learned or been reminded of by the material they had just read. The original statements were scored by the number of correct points made. The points made after reading were scored according to whether the reading material had simply jogged the memory or informed people of things they did not already know. Finally, people were asked a number of questions, which were aimed at getting a general sense of their response to the material – its layout, ease of reading, potential value, etc. Similar questions were also put to managers in each company.

4.2.2 Participating companies The original intention was to return to companies who had already taken part in the questionnaires and interviews. However, it was decided that a second approach could overburden firms who had already given generously of their time. Therefore, new companies had to be identified. The aim was to visit 10 companies and test a total of 30 platers. Since there were few companies left in the West Midlands which had not either already participated or refused to participate, it was necessary to go as far as Cheltenham and Gloucester. Because of this it was decided not to complete the final 3 tests when 27 tests in 9 companies showed no significantly different new responses emerging. Of the nine companies taking part, 4 had less than 20 employees, 2 had between 20 and 50 employees, and 3 had over 50 employees. In two cases the only person available for testing was a plater/managing director. Apart from this, the range of people tested in each firm was 2 – 6.

4.2.3 Results Table 4.1 shows the number of correct points made before and after reading the safety information. The number of points made before reading either set of safety material was low, considering that in the order of 70 aspects of safe working with chromic acid could have been mentioned. The two highest scores (33 and 21) were outliers with the majority range between 7-17. The average for the two groups was comparable. Table 4.1 also shows the additional points made after reading the information, either points made which evidently arose from memory jogging or new points which previously had not been known. There was no difference in the total number of points made after reading excerpts from the Safety Data Sheet and after reading the new material. However, there was a slightly larger number of new points made following reading of the new information compared to the Safety Data Sheets although the numbers are not significant. The reasons for the absence of any quantitative effect are discussed in chapter 6.

29

Table 4.1 Number of correct points made before and after reading safety information (N=13)

Correct points made before

reading Additional points made after reading

SDS Quiz SDS Quiz Jm dk Total Jm dk Total Total 210 201 36 10 46 33 14 47 Range 7-33 7-21 0-8 0-4 0-8 0-6 0-4 0-7 Average 16 15.5

Jm = jogged memory of points actually known dk = identified new points which previously did not know The attempt to gather qualitative information produced more useful data. Spontaneous comments on the quiz included:

Sheet’s very interesting

Quite comprehensive. Covers in detail all the things we’re aware of

I think that leaflet’s pretty good. It tells you a lot. I think we should have that pinned in our section

One manager pointed out the training potential of concentrating on chromic acid. If workers could internalise the principles involved in safe handling of hazardous substances, they would be able to apply them to less hazardous substances. Platers were equally enthusiastic about the usefulness of both sets of material for training someone new to plating. Asked whether a new employee would be likely to find it easy to read and understand, however, there were consistent differences in the replies. In the case of the quiz these included:

Wouldn’t be a big difficulty – no, not at all Yes, fairly straightforward Nothing daunting, pretty well defined. Can’t see any problems

Responses to the Safety Data Sheet included the following:

Probably a new starter might have some problems but if you go through it with them Depends – out of school they would have no idea A novice would need it in plain English, in layman’s terms I’ve been doing this for 8 years and still – what? (indicating non-comprehension) The terminology is confusing – it’s good if they ask but some don’t ask – fresh from school

30

The question How would you change it to make it more useful to people working with chromic acid? also brought out some differences in response to the two sets of material. Responses on the quiz included:

No because that’s spot-on, that is. That’s good stuff, that is. Not too long: best to cover it from all angles. Once a year you could sit somebody down and have a little mock test, an updater I wouldn’t change it – pretty spot-on. Very good that. Excellent No, I don’t think so, I think it’s more or less straightforward Seems all right to me, covered everything Have an even more detailed check-list of precautions for each job Not too long, no, I think it’s very good. They should have time to sit and read, if they’re going to do a job let them read first. We could do with a copy downstairs Put risk in context e.g. smoking, household goods. A young plater would be scared by this material though a smoker; could become a traffic warden instead and breathe in exhaust fumes. Risk assessment and a balanced view are needed It takes a while to read but it’s not worth cutting bits out just to make it quicker to read. You’ve got to know it all Make the writing bigger – blokes out there would find it hard to read. It’s long, but there is a lot of information that needs to be said – fine Probably a little bit simpler – a shorter quick guide. Give the long version first and then the quick guide as a reminder Maybe some pictures of things. I saw a brochure on dermatitis when I was an apprentice – it made me think about making sure not to get it on me when saw what it did to you

Responses on the Safety Data Sheet included:

Making it easier to understand, just spell out dangers and what it does – it has to be technical but explain more simply in terms that everyone can understand Break down into layman’s terms – what it can do to you without all this terminology – it would be easier to understand. It doesn’t tell you how to treat (chrome) before putting it down the effluent Plainer language Highlight signs, home in on skin – we use sharp knives; warn against cuts (leading to ulcers) Sift some of the … how shall I say, not exactly garbage, but long-winded. Break down, make simple, plain English. Just make it more understandable

31

5.EFFECTIVENESS OF THE NEW MESSAGES

5.1 OBJECTIVES & METHOD The objectives of the second phase of the project were to determine whether: • The new information is being used and is considered useful • Platers feel confident about their risk understanding • Platers have improved their risk understanding as a result of having the new information

available In order to explore more fully the effectiveness of the new information, it was decided companies should use it on a day-to-day basis before any survey was undertaken. Also the decision was taken to expand the information formats. Specifically (i) production of an A2 poster containing just the hazard phrases and associated symbols/signs designed to go on the wall of a plating shop, (ii) translation of the A4 sheets into Punjabi, and (iii) production of an audio tape in Punjabi of the information on the A4 sheets. Punjabi had been identified as the most common other language used by platers (Section 3.2). Production of the tape followed advice from a local community translation organisation and sought to respond to the problem of low reading skills amongst those who do not have English as a first language. The tape comprised a single voice reading out the A4 sheets. The font size on the A4 sheets was enlarged in response to comments during the pilot test. The A4 sheets and A2 poster were sent, during the autumn 2000, to all companies who had taken part in Phase 1. The Punjabi A4 sheets were sent in addition to those companies with Asian platers. The audio-tape was sent to one company identified as having a majority of Asian platers. A letter accompanying the information informed companies that they would be contacted in 4-5 months time and revisited to either conduct interviews or to leave questionnaires to establish how successful and useful the new information had been. All of the original companies involved were sent a reminder letter that they would be contacted in February 2001. They were then all telephoned to ascertain their willingness to be involved in the second phase. Of the 30 companies originally involved, 20 agreed to take part. Two had gone out of business in the interim period, one was in the process of moving premises. A problem became apparent in that some information that had been sent out had either not reached the companies concerned or had been mislaid or filed away and forgotten. This inevitably delayed the second phase as new copies of the information had to be supplied. Three methods were used to determine the effectiveness of the new information:

(i) Telephone interviews When the companies were re-contacted a brief telephone interview was conducted, usually with the manager or the health and safety officer. This was to ascertain where the information was being made available and how it was being used. (ii) Questionnaire A closed-ended, questionnaire (Appendix 3) was designed to explore use of the new information, and whether it had had any effect on knowledge and confidence in terms of working with chromic acid. Some questions were duplicated from the phase 1 questionnaire. Copies of the questionnaire were delivered to companies, and managers were asked to

32

distribute them. Completed questionnaires were either collected or returned by free-post. Unfortunately, questionnaires sent by two companies were lost by the latter method (iii) Face-to-face interviews Face-to-face, open-ended interviews were conducted at eight of the companies (18 people). The interview followed a similar format to that in the pilot test of the new messages, although in this case platers were not given the new messages to read during the interview nor Safety Data Sheets used. They were asked questions on the basis that they had read the information previously. The interview explored the immediate and long-term effects of chromic acid on the individual; the precautions that can be taken to prevent or reduce these effects, and views about the new information, including its potential use. The results cannot be compared directly with the pilot test open-ended protocol because of these changes, because people had access to the material over a much longer time period, and because a different interviewer was used. Nevertheless, the primary format of the interviews maintains the open-ended oral protocol that was considered important.

5.2 CHARACTERISTICS OF ELECTROPLATERS As with the first phase, there was a predominantly male bias in those interviewed and completing the questionnaires, reflecting the composition of the workforce. All nineteen of the interviewees were men. Of the valid questionnaire responses (N=43):

• 91% were male (same as in the first phase); • 42% had worked in electroplating for more than 10 years (compared with 63% in phase

1), with 30% having worked in plating between 1 and 5 years; • 35% were aged between 35 and 49 years; 28% between 25 and 34; and 16% were 50 or

over (similar to the first phase). Of those that completed a questionnaire, 90% were directly involved in the chrome plating process, with only 5% (N=2) being indirectly involved and 5% not providing any information in relation to this question. All were platers. No managers took part in this second phase. Twelve of the respondents (28%,) were not native speakers of English, a much larger proportion than the first phase due largely to the targeting of a single company having a majority of Asian platers (27) in its workforce of 36. Most people (60%) completing a questionnaire for this stage had not completed a questionnaire in the first phase. Only 28% had previously completed a questionnaire, the remainder being uncertain or not answering.

5.3 ACCESSIBILITY OF INFORMATION Initial telephone interviews revealed that most of the companies took the information seriously and had considered ways to optimise its use. In a number of companies the small sheets as well as the poster had been displayed on the wall. The location of the information varied from where people clocked-in, to the canteen wall, to actually placing them in the plating shop per se. In some circumstances, however, the small sheets were kept in the office with a view to using them for training or for lending out on request. By contrast in three of the companies each plater had been given their own copy of the information.

33

When the information first arrived it had been distributed in different ways. Where the chrome plating department was only small, the information was discussed between the platers. In larger plating shops a more formal meeting was held in order to introduce people to the material and inform them where it would be kept. Whilst many of the managers thought the information was good, it was also noted that in plating: Practical experience counts for more – effort goes into getting people trained but how much people remember is not certain. Some managers noted that the information was being used as part of a training package or that there were plans to incorporate it more formally into training materials:

We have used it as part of a training package – it is a good format with the questions and answers.

[The information has been used to train] a number of people in the factory to recognize the signs.

In terms of whether the information had made any difference to practice, or whether people had asked any more questions, there was a variety of response:

Most people have worked with it for a long time so know how to respect it [It has resulted in a definite difference in approach because] we didn’t know it was carcinogenic before. Operators have shown an increased interest in how various chemicals can affect their health It has increased operators’ awareness of everyday hazards in the plating shop. The sheets have started people discussing some of the information more now. It [the new information] is in simpler language than COSHH and Safety Data Sheets so people understand some of the more complex terms, but it has not made any practical differences.

Seven of the 20 companies took a different view, however, and some had either lost the information or thrown it away upon receipt. Others had made the information accessible only when they knew a researcher would be visiting. One manager mentioned that both he and the platers had thought that the information was humorous and, as a consequence, little new knowledge was acquired by the platers. In two cases, the health and safety contacts said that they had felt the information to be too frightening to be distributed for general use on the shop floor and therefore it had not been used in practice

5.4 KNOWLEDGE OF THE IMMEDIATE AND DELAYED HEALTH EFFECTS The questionnaire and the interviews sought to establish whether the information had raised awareness of the short- and long-term health effects of chromic acid. In the interview context, all except one of the platers had read the information; the exception was a new plater who had not had his attention drawn to the information sheets. The questionnaires demonstrated a higher proportion of people who had not seen and/ or read the information and this seemed to be dependent on the attitude of their company to its value and the need to make it accessible.

34

Table 5.1

Have you seen the new information sheets on chromic acid?

Poster

Total

Yes No Not sure No data Yes 8 6 2 9 25

Small sheets No 4 1 5 Not sure 1 2 3 No data 6 2 2 10

Total 15 12 4 12 43 Table 5.1 shows that only 8 of the 43 participants had seen both the small sheets and the poster, most people had seen one or other of the formats with the highest proportion having seen the small sheets. Only 4 indicated that they had seen neither, although it is likely that most if not all of the ‘no-data’ also had not seen the information.

Table 5.2 Have you read the new information sheets on chromic acid?

Poster Total

Yes No Not sure No data Yes 8 8 10 26 Small sheets No 5 1 6 Not sure 1 1 No data 6 2 2 10 Total 14 15 1 13 43 Table 5.2 shows that 16 out of 30 respondents (53%) had read the small sheets, but only 8 of these had also read the poster. 5 definitely had read neither the small sheets nor the poster. The non-respondents probably also falling into this category. The results have been analysed according to whether or not the new information (sheets and/or poster) had been read. Although someone who had read the poster only would not have been exposed to the information on precautions and working practice contained in the A4 sheets, this only accounted for 4 of the people.

5.5 EXPOSURE & HEALTH RISKS There was a clear difference in awareness of the effects of chromic acid amongst those who completed a questionnaire and those who were interviewed. In this phase, unlike the first, the greater awareness was amongst those who were interviewed (all bar one of whom had read the information). For example, during interviews, only 2 of the 18 platers did not mention that chromic acid was carcinogenic and that, more specifically, it can cause lung cancer. A direct comparison with the statistics in Table 4.1 is not relevant as on this occasion no comparison with Safety Data Sheets and the ‘before and after’ knowledge situation is possible. The average number of correct points made was 14.5, seemingly slightly lower than in the pilot test prior to reading of the new messages. The total number of correct points made was 261 with a range from 8-32 (similar to the pilot test). A further 55 points were made once memories

35

were jogged. In this phase people seemed to concentrate on the new information that they had read and particularly made a point of emphasising the new understanding that they had gained (i.e. about long-term effects). This concentration probably accounts for the apparent lack of increased understanding suggested by a straight statistical comparison with Table 4.1. The questionnaire comprised a multiple-choice section relating to the hazard phrases that required selection of one definition from four. This caused a certain amount of confusion, a number of people ticking more than one option for each question, others left all four blank and wrote ‘don’t know’ by the side of the question. These cases are represented by ‘no data’ in the following tables. Platers were required to answer questions about practical as well as theoretical issues covering both long- and short-term health topics as well as more general information about health and safety practice. For example, the first question required the chemical name for chromic acid: 53% of those who had read the sheets and 64% who had not read the sheets knew that this was hexavalent chromium with 9% in each category being unsure and the remainder giving one of the other three answers (trivalent chromium, chrome III or chromium compound).

5.5.1 Health Risks Given the displayed paucity of understanding of delayed effects in phase 1, chronic health risks were a particular focus of attention in this second phase. As with the first phase, the interviews indicated considerable awareness of many of the immediate effects of chromic acid in terms of skin/ eye contact:

Eyes…if you get it in your eyes it can affect your eyes – it’s really painful ….if you’ve got any cuts and you get it in a cut it can come up like a volcano Your nose can bleed…soreness, blockages…with chromic acid it does burn when you get it in your eyes ….when I’m here my nose ’ll run and it’ll bleed – you expect it to bleed about once a month You’ll only get burns on your skin if you have a cut – it’ll come up in an ulcer and go septic….sometimes if you get chrome on your hands and you don’t touch it it’ll stain…..

However, unlike the first phase, delayed effects, particularly lung cancer, were more frequently mentioned in conjunction with the shorter-term effects:

Eyes, lungs, the nasal cavity….dermatitis, chrome warts, chrome ulcers, cancer….it’s a carcinogen, isn’t it, so you can get cancer from it as well…… It can give you cancer….lung cancer….if you get it on your skin if you’ve got any cuts it can cause chrome ulcers…. Can get cancer – it can give you cancer of the lungs, it can affect your breathing, it can blind you, it can give you desensitisation of the skin

These immediate responses indicate an increased awareness possibly as a direct result of reading and using the new information. This was stated by one plater:

….like I was saying just now you can get cancer from it by listening to what is on the new data sheets…..which I must admit we didn’t really realize before these sheets came out.

36

Despite the increased recognition of lung cancer, other long-term health effects were not so well-known. In the interviews the development of ulcers and the potential damage that might be caused through the percutaneous nature of chromic acid were the other main long-term effects mentioned. Few of the interviewees mentioned dermatitis, asthma, or skin sensitisation. Table 5.3 illustrates an increased recognition by those who had read the information of some of the hazard phrases associated with chromic acid. For example, 75% understood what a category 1 carcinogen is, as opposed to 36% amongst those who had not read the information (56% in phase 1). However, people struggled with the term ‘respiratory carcinogen’ with the respective percentages of those understanding this concept being 50% and 27%. Only 40% made the association that if a carcinogen is something that causes cancer then a respiratory carcinogen must be something causing lung cancer. Indeed, 22% of those that had read the sheets, and 46% of those who had not, identified a respiratory carcinogen as being something that caused difficulty in breathing indicating that the addition of the word ‘respiratory’ resulted in association to a different physical condition. Other issues were less successfully communicated with less than half the questionnaire respondents being able to correctly identify a delayed and an immediate health effect from a choice of four, although the percentages amongst those that had read the sheets were higher in both cases (44% for each category). However, Question 20 (on the meaning of skin sensitization) indicated a high level of knowledge in both groups, with 63% of those that had, and 82% of those that had not read, the information sheets knowing the correct answer.

Table 5.3 Health Risks of chromic acid (%)

Read

sheets Not read

sheets

False True No data

False True No data

8. An example of a delayed ill-health effect is dermatitis

34 44 22 46 36 18

9. An example of an immediate health effect is eye damage from splashes

22 44 34 64 27 9

10. Delayed health effects may occur when platers breathe in chromic acid mists over a number of years

9 75 16 9 82 9

16. Swallowing chromic acid may cause burns inside the body

49 38 13 27 55 18

18. A category 1 carcinogen means a chemical is likely to cause cancer

12 75 13 46 36 18

19. A respiratory carcinogen may cause lung cancer

31 50 19 73 27 -

20. Sensitization by skin contact means a chemical may cause an allergic reaction

28 63 9 9 82 9

The questionnaire asked people to name short- and long-term effects from breathing in chromic acid and ways that these might be reduced. Of the 32 people that had read the information, 28 answered these 3 questions noting immediate health effects such as nose bleeds; tightness in the chest; wheezing; burns to the nose and throat; shortness of breath; itchy or runny nose, coughing. There was some evident confusion with people mentioning kidney failure and lung cancer as immediate effects.

37

Of the long-term effects from breathing in chromic acid, 18 of the 32 respondents that had read the information recognized that chromic acid caused (lung) cancer. Other answers included asthma; holes to the nose; allergic reaction; mouth and nose ulcers; kidney cancer; residual damage to lungs. There is a question about what information is being understood. There is evidence that once a concept has been introduced – such as long-term effects of cancer - it tends to be become pervasive. Thus, 19% of those that had read the information thought that swallowing chromic acid could cause cancer. Even though words have been explained they are still unfamiliar in their own right:

I think that there’s a couple of words there that I might have had difficulty with myself especially, and I know other people would have done, if they hadn’t been explained cos, to be honest with you, although we’re in the game it’s words we don’t use

I’ve never been sat down and explained the long-term effects…..you don’t hear about the long-term effects so much really cos by the time that person might have left that factory…..you don’t get to hear how their health’s gone

5.6 PRECAUTIONS AND PROTECTION Questions were asked about the practical aspects of working with chromic acid related to protective clothing and action in the case of short-term elevated exposure. Table 5.4 summarises the responses. Almost all respondents knew what immediate actions to take if there was an accident involving chromic acid. Many of those who had not read the sheets returned a 100% ‘correct’ response. Responses indicated that platers are still more aware of how to reduce the risks than they are about the risks per se. The interviews also explored ways in which exposure to chromic acid could be reduced with regard to nose; throat, chest and lungs; skin; eyes; and mouth, throat, stomach, kidneys, liver and blood. Most of the platers were aware of the precautions that should be taken both in terms of personal protective equipment (PPE) and ventilation:

…if we have to get into the vat to do maintenance or take the solution out we have overalls on, green armlets, strong gloves, wellies, a face mask with a filter on it and I also put a bag over my head as well. We wear full face masks, respirators, face shields, gauntlets, wellies, tip it out in a safe place and keep the distance you’re carrying it to as short as possible to reduce the hazard of tripping, aprons, gloves, glasses… We’ve got extraction….we’ve got a good extraction system really…and they teach us how to lift…they’ve got lifting training so we don’t, like, splash it in…. ….if the extraction goes down we do….we either shut the vat down or we have been known to put a suppressant in rather than shut the vat down but we rely on a good extraction

People also mentioned the importance of monitoring the atmosphere within the plating shop. Although interviewees also knew much of the basic and practical health and safety information, as with the 9% who would neutralize chromic acid with an alkali (incorporated in the ‘false’ responses to Q 13 on Table 5.4), one of the interviewees said:

38

If I get it [chromic acid] on my hands I usually wash it off with [hydrochloric] acid as soon as possible.

The new information has had an impact in terms of the effects that are not directly/ immediately experienced or where solutions are not necessarily based on common sense. For example, in terms of the usage of gloves, there was a significant difference between those who have and who have not read the information. Whilst 44% of those who had read the information recognized that gloves are prone to contamination, for those who have not read the information, the most frequent response was that gloves cause skin rashes, perhaps inferred from experience. Similarly, 28% of those who had read the information recognized the label shown as indicating that skin contact should be avoided, whilst all those who had not read the information thought that the sign meant that gloves should be worn – a clear association of the symbol with a practical action. One interviewee raised the possible confusion relating to the symbol denoting safety glasses:

Some of the signs I think you could perhaps change to mean more….the one about eye irritation……to say that it’s an eye irritant……we’ve got them all around the plant to say it’s a safety glasses area and it looks similar.

Table 5.4 Precautions for working with chromic acid (%)

Read

sheets Not

read sheets

False True No data

False True No data

12. The problem with gloves is that they can become contaminated

31 44 25 55 18 27

13. If you get chromic acid on your skin the first thing you should do is wash it with clean water

13 84 3 - 100 -

14. The first thing you should do if a colleague has difficulty breathing while working with chromic acid is move them into the fresh air and inform the supervisor

- 94 6 - 100 -

15. The first thing you should do if chromic acid gets in the eyes of yourself or a colleague is to swill it out with plenty of clean water

3 94 3 9 91 -

17. This symbol tells you that skin contact should be avoided

53 28 19 100 - -

5.7 PUNJABI INFORMATION Effective communication of chromic acid risks to the platers whose first language is not English was an important aspect of this project (28% of respondents in this phase). However, there were some problems in collecting data as there was a need to rely on other platers to take time from

39

their shifts in order to translate questionnaires. In the first phase, questionnaire completion had been assisted significantly by its use in an English class being run in one company. Unfortunately, of those who did not have English as a first language only 7% (N=3) had read the information. This might be because (i) the information was not produced in a language accessible to these platers; (ii) the Punjabi information had not been made available to them; or (iii) they could speak but not read Punjabi. With such a small number of respondents, the statistical significance of the results is low. However, in summary 2 of the 3 thought that the new information was better than any that they had seen before. The most interesting information came from the company that had a very high proportion of Punjabi speakers that had been given the audio tape. One of the platers was interviewed on behalf of those that had used the tape:

I found the tape more helpful than [the new information sheets]….you think about it see….you think about it more…I listened to it at home…it sinks in more.

A suggestion for improving the tape was to have the questions and answers read by two voices rather than one (considered by the research team but more expensive). The platers considered the tape to support the written information not replace it. The package of information (i.e. tape, small sheets and poster) had made a difference to the platers in this company primarily because they had not been able to read or access health and safety information before:

They’re more health conscious now….they make sure that they have their respirators on when they go to the top of the tank….It’s had a big impact on practice.

I find it really helpful….there’s a couple of things in there that I didn't really realize like skin ulcers and actual effects over a number of years [I was just aware] of the short-term effects

5.8 WORKING IN PLATING The third section of the questionnaire explored perceptions of working in plating. Table 5.5 summarizes the results (note N/D = No Data in the tables from this point). Only two questions show any significant differences between those who had and had not read the new information. In relation to Question 32, ‘Labels on containers are more useful than safety data sheets’, 91% of those who had not read the information considered labels more useful whereas those who had read the new information were evenly split in their responses. In relation to Question 35 “Work hazards are part and parcel of the job” those who had read the information were more likely to disagree with this compared with those who had not. In relation to Question 40 “ No one has explained to me how chromic acid could affect my health, there seemed to be greater clarity that this was not true amongst those who had read the new information. 5.8.1 Chemical safety As in the first phase a high proportion of people believed that whilst wearing protective clothing does not necessarily guarantee safety, as long as chemicals are handled correctly then the chances of being harmed are reduced (75% of those who had read the sheets and 73% of those who had not). This was supported by some of the interviewees:

40

I think it just comes back to commonsense all the time – as long as you wear you protective clothing…..you use your loaf and you know…..it’s something you work with and not play with….I think chrome’s safe. It’s the same as a lot of other different chemicals – you’ve got to treat them with respect – if you don’t, they’ll damage you.

5.8.2 Oral Tradition The oral tradition of learning emerged as a strong element (Question27) endorsing the findings from the first phase. This was confirmed in the interviews:

…if you go through it [the new information] with your job as well…..take them through the stages and that….break it down with the job they’ll take it in a lot more….when I was an apprentice you didn’t always read it but if you were doing it whilst you were working and you were told about it then you took it in a lot more To be totally honest with you the best way to get information across is to have every person .. explained to about what it can do because a lot of people will look at that [the new information sheets] and take no notice.

5.8.3 Health and Safety Information Most respondents indicated that health and safety information was important to them (Question 29), and that it influenced the way that they work (Question 34). However, over half of the respondents in both of the groups felt that written information was of little use by itself. Respondents also expressed a willingness to read health and safety information. Most agreed that they made time to read it (85% of those that read the sheets and 73% of those that had not). However, 63% of those who had not read the new information sheets said that they always read new health and safety information! 5.8.4 Trust and Information Provision A significant number of respondents indicated that they would like someone from the Health and Safety Executive to come to the workplace to explain health and safety information (84% of those who had read the information and 92% of those that had not). Whether this is a case of wanting someone perceived as ‘independent’ to explain the information or just the fact that the risks are being communicated in an interactive way is not apparent. One person commented on their questionnaire: Actual talks should take place at work on a 6 months basis. Most of the respondents expressed trust in the information provided by a chemical supplier on health and safety (62% of those who read the information and 82% of those that had not). There was only one interviewee who expressed reservations about information from chemical suppliers:

I think a lot of the problem is the suppliers of the chemicals – they obviously don’t want to scare people away from using them and they’ve already put it in small print on their health and safety data sheet, it’s not necessarily put on the containers.

41

Table 5.5 Working in plating (%)

Read Information Not Read Information Agree Disagree

N/D Agree Disagree N/D

24. Working in plating causes no long-term health effects

25 69 9 82 9

25 As long as you wear protective clothing you cannot be harmed

44 56 2 55 36 9

26 All chemicals are safe as long as they are handled correctly

75 22 3 73 18 9

27 I learned most of what I know about the risks from plating from people I have worked with

79 19 3 64 18 18

28 There is no danger if the extraction on the plating tank doesn’t work

9 91 - 18 73 9

29. Information on health and safety is important to help me protect my health

97 - 3 91 - 9

30. I am concerned about the effects of chromic acid on my health in the long term

78 16 6 73 9 18

31 Written health & safety information is of little use by itself

56 38 6 64 18 18

32 Labels on containers are more useful than safety data sheets

47 44 9 81 - 9

33. Safety data sheets don’t provide the information I need

34 59 6 46 36 18

34 Health & safety information influences the way I work

84 13 3 91 - 9

35 Work hazards are part and parcel of the job

59 31 9 82 9 9

36 Someone from the HSE should come to the workplace to explain health & safety information

84 16 - 92 - 8

37 Health & safety information is not important unless you have work-related health problems

12 85 3 - 91 9

38. I would believe information a chemical supplier provided on health & safety

62 25 13 82 9 9

39. I don’t have time to read health & safety information

12 85 3 18 73 9

40. No one has explained to me how chromic acid could affect my health

38 62 - 45 45 9

41 I always read new health & safety posters

75 13 3 63 27 9

42

5.8.5 Changes in the Plating Industry Interviews with more experienced platers revealed wide-ranging changes in the plating industry including in the use of information. For example:

The plating industry for years was regarded as a ‘railway arches’ industry – no one wanted to know anything about it really but everyone wanted the product. But now it’s getting far more upfront – there’s a lot more regulations to it which aren’t always a good thing but in most cases they are.

The new information provided was viewed as part of this wider cultural shift. One plater commented:

I will say these days people do tend to ask more questions than they used to….we’re a very open company here – we try not to keep things a secret – but I’m not saying it was always that way. And I don’t think it’s because of any sinister reasons – I think people just….a lack of information being given to people….[but] it’s [the new information sheets] very good information coming out – it’s a pity it’s taken so long for someone to grasp the nettle, really, isn’t it?

5.9 PERCEPTIONS OF THE NEW INFORMATION Both the questionnaire and the interviews provided opportunity for people to comment on the impact and usefulness of the new information. Table 5.6 indicates the extent to which people who had read the information felt it had affected their knowledge of working with chromic acid.

Table 5.6 Impact of information sheets on awareness and working practice (%)

Considerably A

little Not at all

N/D

Has the new information:

44. Made you more aware of how chromic acid could harm you?

50 44 3 3

45. Affected the way you handle and work with chromic acid?

41 25 31 3

46. Made you more aware of how contact with chromic acid could be reduced?

50 41 9 -

47. Helped you understand other information sheets or labels?

38 47 16 -

Respondents found the information most useful in terms of acquiring knowledge about the more ‘theoretical’ or ‘hidden’ aspects of chromic acid compared with the practical issues of working with it. However, the interviews indicated that the information sheets had made a difference to the way that people were working with chromic acid:

Practical experience…..yes we know it…..the problems are people tend to become complacent and I think it’s a good thing to prompt even the oldest hand

In one company an interviewee was talking about two of the platers who had working in plating for 37 years:

43

It’s a culture thing…they’ve been brought up through the ‘chromic way’ and didn’t realize the implications of it. So now you’ve put something in front of them saying ‘Look….this really is not good….you really shouldn’t be doing it and you really need to change for your own health and safety issues’…..so a few of the old ones are now completely kitted out after 37 years and I thought ‘Wow! That’s a culture change!’

In total, 94% of people who had used the information sheets felt that they were now more aware of how chromic acid could harm them. 91% were now more aware of how contact could be reduced. This is supported by interviewees:

I was more or less aware of the eyes and the hands but since I’ve read, you know, masks, over-vat ventilation and all that I think I’ve become more aware in the job that I do and I make sure that I’m doing it right instead of jumping the gun. I’m more aware of the mist and the fumes now

…things on the sheet – most of them we already knew about. The only things we didn’t know about was mask over the work which we have asked for but they [the managers] say it’s safe enough – I’ll tell you in a few years time.

Others mentioned that they were already taking the safety precautions indicated on the sheets:

I’d say it would make me more aware of what I was dealing with but it didn’t really change me much anyway cos I was…..I knew chrome was bad for you in every way so it didn’t really change me much….. No….not really…..the cancer part did worry me though….I said to x, ‘I didn’t know about that!’

5.10 USEFULNESS OF THE NEW INFORMATION FOR TRAINING Both the questionnaire and the interviews explored the usefulness of the new information for training, differentiating between training of existing and new platers. As Table 5.7 shows the majority considered the information to be helpful for training but most useful for new platers. However, 44% of respondents believed that the information would be very helpful in training experienced platers as it served as a refresher. As one interviewee suggested: [It] would be good to have it as compulsory to read over every 12 months. The information was considered to have particular training value for new platers:

I think something like that would be good for induction, test them and drill it into them right from the word go. …it makes them totally aware of what they’re using Very much so….very much so…..we’ve already started putting it in the place of some of our own health and safety details because people said how good it was Yeah….this actually tells you some of the problems – what it can cause and that – so they can [be made aware] of what nasty stuff it is and what it can do to you….it’s all down there

However, some concern was expressed about the material being too frightening:

44

It’s fairly straightforward so….you probably wouldn’t want to work with chrome once you’d read it……

Table 5.7 Usefulness of new information for training (%)

Very

helpful Partly helpful

Not at all helpful

N/D

Would the new information sheets be helpful in training:

48. People who are already platers? 44 47 6 3

49. New platers? 91 6 3 -

The interviews suggested that the material would be best used in an interactive way for training new platers, either by going through it with a supervisor or to supplement what they were told as they learnt ‘on the job’. Use of the quiz format on the sheets was also considered potentially beneficial.

5.11 USE OF THE NEW INFORMATION People have used the information to identify both specific and general information on chromic acid risks. Table 5.8 illustrates the reasons why people had used the new information.

Table 5.8 Use of the information sheets (%)

Yes No N/DHave you used the new information sheets or poster to obtain information on

50a. some particular aspect of chromic acid? 63 28 9

50b. chromic acid generally? 63 22 16

Figure 5.1 shows the specific aspects of interest. The graph illustrates the spread of the responses rather than specific combinations (respondents were asked to tick any that they felt were relevant). The most interesting elements appear to be the information about chromic acid itself rather than the practical aspects that are learnt ‘on-the-job’. This rather simplistic overview belies the combinations of reasons for using the information. For example, 28% said that they were interested in all the aspects and 16% that they were interested in all the aspects apart from ‘Ways that you can get chromic acid on you’. Platers also referred to the information being distributed outside their companies. One plater mentioned that they had passed the information on to their chemical supplier. A second who had suffered from a respiratory problem for a number of years took it to his doctor, who took a photocopy of the information and sent him to hospital for a chest x-ray. He was diagnosed as having asthma.

45

Figure 5.1 Specific aspects of the information platers found interesting

5.12 PREFERRED MEANS OF RECEIVING INFORMATION Table 5.9 shows how platers would prefer to receive risk information. Audio tape was the least favoured, although written information alone was also unpopular. Visual formats were the most popular, with pictures and video scoring highly, corresponding with the results in the first phase (section 3.7). It is important to note that platers often had no experience of alternative formats and the pros and cons of the options were not discussed. However, amongst those who had read the information the poster scored well, with 34% noting this as the best means of information provision. Indeed, all of the15 respondents who had read the poster considered it useful. The reasons for this included because it was a quick reference and served as a visual reminder, particularly when placed in areas that are used frequently: e.g. by the chrome tank, by the clocking-in point or in the canteen. Interviewees also commented on the effectiveness of the poster as a means of displaying information:

Yeah….cos they’re like the actual symbols that is used…it’s not like some drawing that people do….that’s like the proper symbols that we see around our place…so yeah, it does help…. ….if it’s in a poster form it encourages people to put it on the wall…usually…hopefully where the operations are taking place I learnt this quite a few years ago, that a picture on a notice of any sort brings the fact home a lot better than half a dozen lines. If you’ve got a picture there then I think that 9 out of 10 people take more notice…..

0%

10%

20%

30%

40%

50%

60%

70%

80%

The health effects ofchromic acid

Ways that the healtheffects can be reduced

Ways that you can getchromic acid on you

Ways that you canreduce the chance of

getting chromic acid onyou

Explanation of termsand symbols used in

official health warnings

No data

Perc

enta

ge o

f res

pond

ents

that

read

shee

ts

46

However, one person mentioned that the poster might frighten some people:

I think some people, with all those hazardous chemical symbols, would think, ‘Hang on, I’m not touching that!’ so I think some of the pictures may get people to over react to it….

Table 5.9

Preferred ways of receiving information (%)

Read sheets Not read sheets Video 19 9 Audio - - Written 3 9 Written with pictures 34 18 Video and written with pictures 13 18 Written and written with pictures 19 - All except audio 13 27 No data - 18 Interviewees suggested the form a video might take, including the use of ‘shock tactics’ to reinforce some of the effect; as an introduction to the actual chemistry of chromic acid and the plating process, or as a general health and safety introduction. People also noted that it requires very little effort to engage in such a form of communication. However, it cannot replace written information as once it has been seen it is easily forgotten. Many people expressed their satisfaction with the new information:

They’re brilliant [the new information sheets] – the more things like that the better It’s nice and simple – it’s what we need It’s good someone’s done something I think it’s fine – I had no problems with it at all I found that it explained things in, like I said, layman’s terms….much more simpler and more likely for the person on the shop floor to understand….. You’ve put everything in plain English – it’s better than what we did have before It’s not giving people too much to worry or think about all at once – it’s nicely compact and it’s general, good information

47

6. DISCUSSION

6.1 INTRODUCTION The following sections discuss the results presented in Chapter 3 and the design and testing of the new messages presented in Chapters 4 and 5. The sections offer some explanation for platers’ mental models, particularly in terms of the importance of experience and the strength of the oral tradition in the workplace as a means of conveying risk messages. The differences between expert and platers’ mental models are discussed with questioning of whether these differences are significant. The impact of the new messages in terms of the increased understanding of the long-term effects of chromic acid are discussed as well as the difficulties of evaluating impact on behaviour. Lessons learnt in the design and formatting of new messages/information are considered. Finally, some generic principles relating to the application of the mental models approach and specifically the design of risk messages that ‘tell people what they need to know’ are presented.

6.2 ELECTROPLATERS’ KNOWLEDGE AND BELIEFS

6.2.1 Importance of Experience Knowledge does not necessarily translate into perceptions. Other occupational safety culture studies suggest that individuals frequently possess quite accurate perceptions of the physical risks in the workplace (e.g. Rundmo, 1992; Rundmo & Sjoberg, 1997), but have weaker perceptions in relation to hazardous chemicals where risks may be less immediate, observable or are not experienced (Weyman & Scobie, 1998). However, this does not appear to be the case in relation to electroplating. Mental models are constructed partly on the basis of experience. They are a way of organising and understanding experience. Plating is an activity dependent on, indeed composed solely of, the direct use and application of hazardous chemicals with a large amount of manual handling. This sets it apart from some of the other sectors which have been looked at in the second study for the HSE on mental models at the University of East Anglia, particularly soldering where infrequency of exposure to solder fumes impacts on perceived risks. It could be expected that platers’ experience of the risks of chromic acid in everyday working would fundamentally affect their knowledge and perceptions. Platers experience personally, and observe amongst their colleagues, splashes, acid burns, skin rashes, ulcers. They experience and observe frequency and seriousness of effect. Accidents were used by platers in the interviews as a means of describing these effects, whereas they were less able to articulate the effects in abstract terms – i.e. give a ‘textbook’ explanation. The fact that 86% of platers thought that nickel salts were less risky than chromic acid possibly partly reflects the fact that immediate burns and damage to eyes are not an issue with nickel. Wearing the PPE and learning to work with it, even when conditions (such as heat) lead to its removal (e.g. goggles), has a significant impact on beliefs about the short-term or immediate effects. In plating, the potential for harm is part of everyday working. The need to wear the PPE and the daily routines required to put it on, work in it, and then to change back into ‘street’ clothes, provides a constant reminder. Hence the strong understanding amongst all platers that

48

plating is an inherently risky activity, although not one necessarily perceived as more risky than other occupations such as working on a construction site on in the steel industry (Table 3.3). The risks are something that have to be acknowledged but not dwelt upon otherwise it is doubtful whether platers would stay in the job. There was strong evidence that platers had rationalised the potential for harm and had faith in their own experience, common sense and adoption of correct handling procedures. The experts on the other hand did not see this as rationalisation but as ignorance or even deliberate avoidance of the issue. It was not the aim of the project to determine in detail why experts thought this, but certainly the qualitative replies in the expert questionnaire seemed to imply that a plater’s inability to articulate a risk in terms which an inspector might recognise seemed to be interpreted by some experts as meaning that the plater does not understand. The survey results would not support such a view. People do not ‘experience’ lung cancer in the same way as the acute effects. Even if a colleague or friend was known to have lung cancer this secondary experience would not have the same impact, not least as it is likely that the cancer will have been caused by other activities or sources which are non-occupational. 41% of platers thought that smoking was a more risky activity than plating, the one activity that a significant minority of platers considered a greater risk (Table 3.3) Despite the higher potential for lung cancer in plating than many other occupations, the chances of an individual getting cancer is still low with few platers suffering the disease (none in the interviews had experience of a colleague getting cancer). Lung cancer from inhaling chromic acid is a theoretical risk to most platers. It is a risk that is well understood by experts as a potential outcome with the classification of Category 1 carcinogen not subject to disagreement. The potential for cancer was known by over 89% of platers when prompted in the questionnaire, although only 57% knew that nickel also causes cancer (Table 3.1). However, cancer was not an experience that ensured that they articulated this risk in the interviews. Despite appearing to know that chrome mists can cause cancer, a third of platers did not believe that working in plating causes long-term health effects (Table 3.2). If over 50% of them have worked in plating for more than 10 years without evidence of long-term effects then the seriousness of the risks remains theoretical. It is important to recognise that the concept of a ‘long-term’ risk lacks clarity. A lay interpretation of the phrase might include the potential for harm where the effects might last for a long time; the potential for exposure to a chemical but no evidence of the effects until some point in the future; the potential for effects which are irreversible; or the potential for effects immediately which might recur in the future. For experts, long-term versus short-term, or acute versus chronic is a key distinction, albeit one which has shifting definitions depending on context. While platers are aware that they can have an accident with immediate effects or that they may develop a skin problem or nasal septum perforation after some time in the job, they are less likely to use a two-part classification to describe these risks the focus being outcome or harm as opposed to when this might occur. They do not draw distinctions between brief exposure to a high concentration of a substance, and exposure over time to a low concentration. This is not surprising when it is the effects or harm which are known to be of most importance to people. It is inevitable that health professionals and factory inspectors will have a different experiential model of safe working in electroplating from the people who do the job day-in and day-out. The framework for inspectors and occupational health experts is largely theoretical, despite many inspectors having an industrial background. Inspecting equipment, operations and management systems is not the same as actually having to work with that equipment, experiencing problems, working to deadlines, learning the effective short cuts, etc. Some occupational health experts may never have seen a plating shop, but are asked for advice on prevention or management of ill-health effects.

49

Platers’ experience derives from a practical context, from experience of normal, abnormal and accidental operations. This practical experience contributes to the evidence of confidence, perhaps overconfidence, of platers in their knowledge of the chemical risks, although evidence of overconfidence in questionnaire responses in general is not uncommon. Given the stability of experience over many years, with processes and products in many firms having changed little and many platers doing the same job for more than 10 years, the impact of personal experience on mental models in electroplating is powerful. Platers are experiential experts about the risks of electroplating. As stressed in the mental models approach, if people know nothing about a topic then a message about that topic will not be comprehensible, but this will not stop people from grafting the information onto their existing beliefs. The replacement of those beliefs by ‘correct’ understanding will only happen when the information provided matches with experience and is seen as having personal salience. Cancer in general is known to heighten concerns, it is a feared health outcome. In the semantic images of risk in public perception it is a ‘slow killer’ (Renn, 1998) where knowledge is reliant on information from others rather than on personal experience. In that one of the primary deficiencies of platers’ mental models seems to be the potential for lung cancer the production of a message which can be adopted as correct when personal experience is missing requires close attention to language that is immediately identifiable.

6.2.2 Knowledge of Expert Terminology The risk literature has referred to a lay tendency to focus on outcomes not probabilities – hence the concern that people worry about high hazard consequences with low probabilities. The mental models of platers do not suggest such a concern. However, the distinction between hazard and risk (like that of acute and chronic risks) does differentiate the mental models of experts and platers. While the distinction is fundamental to the experts’ understanding of risk management it is an unnecessary complication for platers (and probably most lay people). Experts in the risk assessment process identify hazards, how hazards might arise, determine consequences and probabilities of effect and then design risk reduction measures. Platers think in terms of problems and how to deal with them, and lastly consider how the effects might be caused. This mirrors the distinction between safety communication and risk communication identified in Chapter 1.1. While experts ranked nasal septum perforation as the most risky outcome from plating this was not the effect that most platers mentioned first, although it was recognised as being a potential outcome. Platers tended to talk about burns first, whereas experts ranked these 6th in terms of risk, presumably drawing distinctions based on seriousness of effects and possibly reversibility of effects rather than probability. Platers, drawing on experience, talked about burns as probable and therefore serious. It was noticeable that occupational asthma was not mentioned in the first phase interviews by a single plater and only by a few in the second phase when the platers had been exposed to the new risk messages. However, the experts ranked this as the third most risky outcome, with the occupational health specialists particularly noting this as a significant risk. It is worth noting at this point that there were a number of differences between the responses of the inspectors compared to the occupational health/medicine experts. These have not been analysed in detail because of the small number of the latter included in the expert group (4 out of 13). However, those differences that were identified are sufficient to raise questions about the definition of expertise and also the potential value of studying the mental models of experts in themselves. The fact that there could be differences between these expert groups due to training, responsibilities, experience of the industrial sector and technical knowledge is evident. The fact

50

there could be differences within these groups – particularly potentially factory inspectors – for the same reasons, and due to personal experience and character, also cannot be ignored. The greatest difference between plater and expert mental models was evident in the knowledge of technical terms, such as ‘respiratory carcinogen’, ‘sensitisation by skin contact’, ‘category 1 carcinogen’ (Table 3.5). The questionnaire responses suggested some considerable confusion in that some platers could identify that chromic acid was carcinogenic or might cause sensitisation by skin contact but then were not able to define these terms – perhaps they were remembering having seen the phrases in Safety Data Sheets or other literature on chromium? Other platers could choose the right answer out of four definitional options, but did not know that the phrase related to chromic acid – perhaps they had never seen a Safety Data Sheet (19% of the platers had never used a Safety Data Sheet)? or were simply good at looking at words and choosing the right answer based on analogy but did not really know what they meant? The questionnaire itself could have confused people (a point discussed further in 6.8). However, the conclusion relating to the low relevance of these terms to the mental models of many platers cannot be avoided. The difficulty experts had in reaching a consensus on an appropriate lay definition for sensitisation is not surprising. When any new risk message is composed a considerable amount of expert iteration is required to arrive at a lay statement that all can agree with, and particularly to eliminate any element that could potentially be misinterpreted. Indeed the research team experienced first-hand this problem when it had to agree amongst its five members the lay terminology for the new risk messages. In the short space of time the experts had to respond to the questionnaire, and without face-to-face iteration, it is not surprising that expert consensus could not be reached. Nevertheless, the difficulties faced by the experts serve to stress that these technical terms are not part of everyday language but are an expert/professional fabrication, in some cases - like category 1 carcinogen – a form of expert shorthand. Safety Data Sheets, in adopting such expert shorthand, not least to protect potential supplier liabilities, cannot be expected to influence beliefs if the primary language is unintelligible to the reader. It is important to note that 49% of platers had no educational qualifications (including GCSEs or O levels). Another study for the HSE has identified a low reading age (63% below 12 years 5 months reading age) amongst workers in firms in a number of sectors including electroplating (Biggs & Crumbie, 2000). There was evidence in both phases of the project that platers had primary reading/literacy problems not just understanding deficiencies.

6.3 SOURCES OF INFORMATION

6.3.1 Value of Official Sources Platers may not understand all of the terminology in Safety Data Sheets but they do believe that they are likely to tell them something they may not know. 88% believed that they needed to know not only how to handle a chemical safely, but also what the hazard was (Table 3.6). While there was undoubtedly some over-reporting in the answers to this section, with people potentially concerned not to downplay the importance of official information, nevertheless the understanding of the risks involved in the plating was evidently leading to a belief that formal information is important. Platers’ claims to use labels, often greeted with considerably scepticism by the experts, could be interpreted as over-reporting if it is questioned why people with such a length of experience in a job would need to read labels frequently. However, a more appropriate interpretation of the responses in both phases may be simply that warning labels on equipment and chemical

51

containers are part and–parcel of everyday experience in the plating shop: legally required, impossible to miss, familiar, people being continually aware of them. However, the importance of familiarity with warning signs in industries such as plating was something to be built upon in the design of the new messages. A pictorial image can be powerful when it is instantly recognisable. The evident confusion amongst some platers as to whether all sections in the Safety Data Sheets contained information useful to them undoubtedly in part reflected their unfamiliarity with what might be included under a heading such as ‘Toxicological Information’. As Safety Data Sheets are not the primary source of information for platers (particularly non-managers) and are not referred to frequently, it is unlikely that they are aware of the layout of the material (not least because Data Sheets do not have to conform to a standard layout, although the section headings are standardised). We gathered examples of Safety Data Sheets from some of the companies visited and there was little consistency of presentation and layout. Data Sheets are designed as a starting point for a risk assessment not as a day-to-day usable information resource for workers. Furthermore, they are substance-specific rather than activity/process specific. A few of the larger companies had collated a company-specific information pack with the primary objective of translating complex, technical information into a more user-friendly and company-specific resource. This was used particularly for training new staff, the interviews suggesting that few experienced platers referred to the literature. However, this is not to say that they did not the know that the information resource existed and could be accessed if required. In companies where in excess of 20 chemicals may be in use it could not be expected that platers would have instant recall of information relating to each one, particularly in the event of an emergency or spill where there would be a need to check the appropriate steps to take.

6.3.2 Importance of the Oral Tradition Platers learn from one another rather than from written information Plating shops are small in terms of numbers of people involved in the process – while one company employing 80 people had 55 platers, this was unusual. Another employing 60 people only had 3 platers. In many very small firms most people worked in the plating process, including the manager. A tight-knit workforce focused on one activity inevitably draws on individuals as a source of knowledge and experience (either directly or indirectly). This oral tradition focuses on the translation of messages relating to the need to be careful of certain chemicals (not necessarily why) and what to do to protect yourself from exposure, which combines with ‘learning from experience’ of getting it wrong. This was reflected in the platers’ identification of the most important advice for a new plater being to always wear the protective equipment. The oral tradition is practical and pragmatic. The importance of the oral tradition within companies combined with the literacy issues account for the fact that most platers did not verbalise the hazards and risks in the same manner as an expert. For example an expert might say “if the extraction fan is not working then chromic acid mist will collect over the tanks with the danger of inhalation at an unacceptable level”. A plater is more likely to say – “the extraction fan is important”. While an expert might say “prolonged exposure to high levels of chromic acid in air could cause cancer”, a plater is more likely to say “it can cause trouble with your lungs”. It would be wrong to automatically equate difficulties of articulation with lack of understanding and this is why the method of talking to people and identifying their knowledge and beliefs is important. The purpose of the mental models approach is to determine what people need to know. If a plater knows that an extraction fan is important and that if chromic acid builds up in the air this could cause lung problems in the future, this is what they need to know. It is more

52

important than the same person being able to articulate that chromic acid is carcinogenic and what the MEL is. However, there was some evidence in the interviews that people did not think of how exposure could be caused, relatively rarely articulating a chain of events and outcomes. The importance of the oral tradition has potential in terms of new risk messages. Experts who could engage platers face-to-face to describe the hazards and risks, as opposed to merely providing written information, seem important. However, who would be a credible expert would have to be linked with who a plating firm may be prepared to have in the factory. Much small business research has stressed the potential for firms not to seek information from regulators for fear of the latter identifying something which is being done wrong. However, in relation to chemical risk messages the potential for the chemical supplier to take a face-to-face role is evident and in phase 2 was viewed as useful by the majority of respondents. Only 1 of the experts considered this a good method of communicating chemical risks, emphasising instead the importance of written as opposed to verbal information. The supplier may be regarded as a credible source, which is essential in risk communication. In phase 2 some platers did identify the potential role for the HSE in communicating risks face-to-face although it was not clear whether the HSE was trusted.

6.4 PROTECTION AND CONTROLS Experts think in terms of a hierarchy of risk-reduction starting from the possibility of changing to less hazardous processes, particularly chemical substitution, and finishing with PPE as the last line of defence for people who have to work with hazardous substances. This matches the legal requirements of COSHH. By contrast, platers, especially if they are employees rather than managers, think of PPE as the first control to which they have access, and possibly the only one over which they have any influence. But even then they may see themselves as powerless. Although many companies provide full equipment, inspection is encouraged and replacements are freely available, this is not always the case. Platers may not know what equipment is recommended by the Safety Data Sheets or may not feel able to insist that management provide it. Experienced platers are aware that control of the mists over the plating bath is vital to protecting them against health problems. However, they tend to think less of a theoretical hierarchy of controls compared to the experts. Asked about the relative importance of different controls they simply say that all are important. It is accepted that perceived risk may not influence precautionary behaviour, not least when job-related barriers are taken into account: for example, the heat and conditions of the plating shop and need for dexterity which may lead to PPE being removed. Peer influence is also important. New workers in plating learn from more senior platers – on-the-job. Indeed, the importance of this process of personal, as opposed to ‘classroom’, training was stressed by managers. Therefore, practical working and the experience of how to do this while still protecting against the evident short-term risks is important. It is arguably more important than a correct and full understanding of the risks. However, experts agreed that failure of training was resulting in failures of control in terms of poor work practices and poor working conditions, particularly manual addition of chemicals to plating baths and the failure to maintain extraction systems. Platers seem to be reassured that there is no risks since the air levels were monitored.

6.5 ETHNICITY The project specifically included Asian platers, though it did not undertake a comparative study of the mental models of English and Asian workers: this would not have been possible without a

53

larger sample. The purpose was to ensure that those from the ethnic minorities, who do represent an important component of the workforce particularly in the West Midlands, were not excluded. Only a relatively small number of Asian workers were respondents: 4 out of 22 of the initial interviews and 16 out of 84 – 19% of the questionnaire responses, 13 of these 16 not regarding English as their first language. The small sample size reflected the fact that the research was dependent upon companies agreeing to participate and those that did did not have a significant percentage of Asian platers, despite attempts to identify relevant companies. One company employing 36 people, 27 of whom were platers had 90% of its workforce whose first language was not English. The phase 2 element of the project did provide an opportunity to respond to language issues in the production of the new messages in Punjabi (discussed in 6.7). However, this element of the work must be viewed as a pilot study. There is still an urgent need to understand cultural differences of risk knowledge and beliefs. Asian platers tended to have higher educational qualifications and were younger. Importantly they had spent fewer years in plating, in particular 11 of the 16 had spent less than 5 years in electroplating (69%) compared to 15% of the English workers. Given the importance of experience in the development of mental models this might lead to more misunderstandings of the risks. However, in many other respects their responses, particularly in the questionnaires, cannot be distinguished with any degree of significance from those of English platers. It is evident that where platers do not have English as a first language this can present problems of understanding, as well as ability to participate in research. The fact that in one company the questionnaire was completed as part of an English class is indicative of the difficulties of involvement and the returned questionnaires actually included a large number of ‘Don’t Know’ responses particularly to the questions relating to the characteristics of chromic acid and the definitions of key technical terms. The general responses from platers and experts identified concern that Asian workers may have particular safety problems because of language difficulties and 15 of the 16 Asian workers agreed with this. However, where deficiencies of knowledge of technical terms was evident (see 6.2.2) the point has already been made that low literacy skills are also a problem for English workers. Asian workers had problems with the same technical terms as English workers. The interviews did reveal some potential cultural effects. There was some evidence that in the workplace Asian workers relate more to one another than to English workers and that where an Asian worker had particularly good command of English and/or some technical training (3 respondents had degrees) and a longer period of experience that they became an important point of reference and also felt that it was their ‘duty’ to help other Asian platers. This suggests an enhancement of the oral tradition amongst Asian workers.

6.6 DESIGNING THE NEW MESSAGES

6.6.1 Logic of the Design As discussed in Chapter 4, the new messages were designed to bridge the gap between platers’ knowledge and beliefs and that of experts. This gap was particularly evident in terms of (i) the awareness of the chronic risks and understanding of how these might be caused (e.g. from a single high dose exposure or from continuous low-level exposure), (ii) all of the means by which exposure to chromic acid is caused, and (iii) what to do to minimise the risks. The focus

54

was on practical and pragmatic guidance responsive to platers’ experience and reflective of the oral tradition - i.e. using lay language. The focus was relevant knowledge. The messages were designed to support, not replace, the Safety Data Sheets. However, it was not presumed that platers would read Data Sheets. Therefore, the new messages were designed to include elements of the key information within the Data Sheets and to be capable of providing stand-alone information. The aim was not to be inclusive in terms of types of information (for example, no information on ecological hazards or disposal of chromic acid was included), but to test a style of information and to focus on the platers’ ‘need to know’. Risk information should not contradict workers’ experience: as one manager explained “risks have to be presented in context”. If information is not, it is unlikely to be heeded. Thus, if chromic acid is presented as a significant cancer risk it is unlikely to have sufficient salience to be adopted as knowledge. While it is important that platers are made aware of the danger of breathing in chromic acid mist, this should not be done by inflating the cancer risk. This conclusion supports other work relating to warning signs in the workplace which stresses the need to provide examples of the outcome/consequences of non-compliance or poor working practices, as opposed to simply alerting people to the hazards (Wolgartner & Laughery, 1996). The design in focusing on one chemical could be relatively easily replicated for others. The audience is primarily the person working in the plating shop, although as the research did not find significantly better knowledge amongst managers the plater should not be the sole audience. In large companies resources might be available to prepare different information to a different level of technical detail for different people. Small companies do not have either the resources or in general such a need (although this is not to neglect the fact that some information such as toxicological data is relevant to the person doing a risk assessment). The last two pages of the new messages were designed to explain in lay language the hazard phrases common to a Safety Data Sheet on chromic acid so that the new messages enhance the role of other formal information. These explanations have been enhanced further through the use of warning signs which were identified as being important to people – i.e. the use of a picture which is readily recognised. The Workshop, organised in June 2000 as part of the research project, identified that there has been European discussion of the potential value of moving the hazard warning signs which currently appear on the last page of a Safety Data Sheet forward to the front so that the reader, when faced with a large number of Sheets, can immediately recognise the most serious hazards. Some of the participating experts considered that there was no point in including information relating to things outside of a platers’ control, for example relating to the importance of the extraction system working. However, the decision was taken to keep in such information as even if it is not a plater's responsibility to check extraction arguably he or she has a right to know, knowledge potentially empowering them to ask for the system to be check if concerned. This also responds to the psychological literature which stresses that if a hazard is known to lead to certain outcomes telling people about these will help them to actually look to see if there is a problem and to do something about it. Information about detection of hazards is best presented as a ‘loss-frame’ - i.e. if you do not do it this could be the outcome (Rothman & Salovey, 1997). Using adequate protection against the acute or short-term risks can provide some protection against the long-term. Therefore, although there was evidence of a reasonable understanding of the short-term risks these were still included, not least as a means of relating to platers’ experience and potentially enhancing the credibility of the messages relating to the risks about which they lacked experience.

55

The new risk messages have not been subject to expert design, in terms of layout, colour, size of font, etc , although there was some understanding within the research team of the ergonomics of instructions and warnings. It is evident that design is important in terms of gaining and holding attention and assisting comprehension. The font size on the A4 sheets was enlarged for phase 2 in response to comments from platers. In addition the decision was take to produce an A2 poster showing just the hazard phrases and signs on the back pages of the A4 sheets so as to provide more visible and possibly more readily accessible information in the plating shop. It was assumed that the A4 sheets may not be readily available to platers all of the time, although by producing them as laminated sheets this might encourage their use within the plating shop as opposed to being left in the manger’s officer as ‘paper copy’. The poster also sought to build on the understanding that labels were important sources of information to platers. Thus, by replicating these on the poster with lay explanations of the terms it was hoped to attract attention, or at least to provide an additional ‘warning’ source within the plating shop. The phase 1 questionnaire asked which other languages people spoke and Punjabi was the most common. However, it did not ask which other languages were read. Young Asians born or entirely educated in England may not read other languages, or may not read them well although they may have a good command of the spoken word. It is common in some Government departments (such as Health) and also in local authorities to prepare key information leaflets in other languages. HSE has not traditionally done this. Advice was taken from the company with a significant number of Asian platers about the value of translation of the new messages into Punjabi. After discussion within the company there was a consensus that it would be worth trying. Advice from a local community translation organisation suggested that putting information onto tape may overcome problems with reading abilities. Both these elements represented significant ‘firsts’ in the presentation of health and safety information relevant to electroplating. We understand that the audio tape was given to groups of workers to play while they were travelling to and from work by car.

6.6.2 Difficulty of Measuring Effect The potential difficulty of measuring any effect from the new messages was identified before the project commenced. The most significant weakness of the mental models approach as applied to occupational risks is the fact that it does not address organisational culture impacts on knowledge and beliefs (see 6.8). Therefore, measuring the effect of the new messages in a specific culture is not possible. The pilot test of the new messages in phase 1 faced a number of problems: (i) there was only a short time available for the interviews with people just stopping what

they were doing in the plating shop to talk to the researcher, therefore the potential for people to develop considered responses was limited;

(ii) the new messages were not being tested in their real context – i.e. day-to-day use; (iii) the interview, although valuable, is still an artificial situation – people do not usually

talk about risk and safety information in this manner, indeed they may rarely have to articulate risk knowledge in everyday life, the research had already identified that people find it difficult to articulate the hazards and risks – often only replying with one or two words rather than sentences;

(iv) the language required to explain risk issues is not a natural ‘everyday’ language; (v) the Safety Data Sheet was presented in a privileged format - i.e. it was made legible

(when often Sheets are rather poor photocopies); the relevant sections to look at were identified for the reader so that possibly (perceived) extraneous sections were ignored and the reading experience was made easy;

56

(vi) the test did not allow for within-subject comparison of the Safety Data Sheet and the quiz;

(vii) there were small sample sizes - although it seems unlikely that the number of initial points made would have been significantly higher with more subjects; and

(viii) although the decision was taken not to have identical pre and post reading tests to minimise sensitivity to responses, the verbal protocol method whereby people were asked to add to the pre-reading statements they had made may have encouraged over-optimism about prior knowledge to minimise concerns that they might be considered to lack knowledge.

The qualitative clues about the new messages from the pilot test were more useful. In particular: (i) managers’ comments were mainly positive; (ii) even when critical, most managers took a copy of the material (the sincerest form of

flattery?); (iii) platers’ comments were extremely positive; (iv) despite some support for Safety Data Sheets there was a recurrent theme relating to

their complexity; (v) a number of constructive comments were made about how the new risk messages could

be enhanced further, suggesting that people engaged with the material, felt in control of it, and found it congenial. By contrast the only comment made about the Safety Data Sheet was the need to put it into plain English.

Therefore, the design of the second phase of the project to provide for a more considered test of the new messages was considered to require evaluation after a period of use of the new information in day-to-day working. In practice this was a period of about 4-5 months. Perhaps not long enough to provide for strong evidence of training potential for new platers. However, given the relatively low turnover of platers it is likely that a much longer period of time would be required before such as full test could be undertaken. The existing access to companies and their interest suggested that a shorter time period before they were revisited was important. Phase 2 combined closed and open-ended methods so as to extend the size of response in terms of numbers of platers engaged as well as replicating elements of the pilot test. The extended number of respondents achieved in phase 2 – 43 respondents to the questionnaire and 18 open-ended interviews provided a good level of response from 20 companies in total. There was opportunity to compare responses of people who had, and had not, read the new information. The decision was taken to drop the comparison with Safety Data Sheets so as to allow more detailed exploration of use of the new information as well as knowledge. The pilot had already provided a clear lack of support for Safety Data Sheets as useful sources of information for platers. 6.7 IMPACT OF THE NEW INFORMATION

6.7.1 Long-term health effects Given the clear evidence of divergence between platers’ and expert understanding of the potential long-term health effects of exposure to chromic acid and the focus on conveying this information in the new messages, the success recorded in increasing understanding of these effects can be regarded as an important outcome of the project. The fact that in the open-ended interviews platers offered this information without prompting and alongside the immediate effects that were known to be most significant to them, demonstrated that the new messages had succeeded in conveying the cancer risk in particular and in raising confidence about knowledge. Platers actually identified the cancer risk information as ‘new knowledge’. The fact that 75%

57

understood the term category 1 carcinogen compared with 56% in phase 1, confirms the increased understanding. However, overall, the interview evidence was more convincing than the questionnaire, as only 25% of those who had read the new information agreed that “Working in plating causes no long-term health effects” compared with only 9% of those who had not read it (Table 5.5) (in phase 1 31% of platers had agreed with this). The questionnaire method raises questions about understanding of written terminology (discussed below).The fact that the phase ‘respiratory cancer’ still confused people is continuing evidence of the difficulties of unfamiliar terminology although there was less confusion amongst readers of the new information (Table 5.3). There was an evident remaining lack of understanding of some of the other long-term effects, such as occupational asthma. This suggests that the cancer messages had succeeded in tapping into basic concerns and resonating with fundamental health risk fears, whereas the other messages were still proving difficult for platers to equate with experience of health effects. The majority of respondents expressed concern about the effects of chromic acid on their health in the long-term, although there was a marginally more positive response amongst those who had read the new information (Table 5.5). While platers are increasing their knowledge of the longer-term risks experiential knowledge is embedded in the electroplating culture and thus change may take time. It is interesting that platers specifically recognised the impact of culture or ‘way of working’ in plating and the difficulty of changing this. There is still an emphasis on ‘seeing and hearing’ about health effects in order to know about them. Thus, although information on cancer resonated with people, the terminology – such as ‘respiratory’ and ‘carcinogen’ – still had the potential to confuse. Overall, however, those who had read the new information concluded that it was effective in terms of providing information on chromic health risks, with 78% saying that it was better than the information they had had previously. Importantly no one thought that it was less useful.

6.7.2 Turning messages into practice Platers who had not read the new information were just as likely to understand the practical precautions to take when working with chromic acid as those who had. However, in relation to the longer-term effects – or at least those not so immediately experienced – the new information does seem to have extended understanding of the effectiveness of precautions. For example, understanding that gloves can become contaminated and therefore may not give full protection. Also those who had read the new information were more likely to know that if extraction over a plating tank is not working this could be dangerous (91% compared with 73% of those who had not read the information). This combines with the general increase in people in the interviews who mentioned the importance of extraction and ventilation without prompting. The new information has led some platers to make, or attempt to make, practical changes to the way that they work and even where people did not feel that they needed to do anything else the information had boosted their awareness and confidence in their knowledge and practice. The evidence from the interviews of discussion amongst platers of the risks confirms this new confidence and interest. Platers were confident that health and safety information influences the way that they work (Table 5.5). However, confidence in barrier methods of control was also still evident. Indeed, whereas in the first phase 32% of platers believed that wearing protective clothing prevents you

58

from being harmed (section 3.8), in the second phase this had risen to 44% of those who had read the new information and 55% who had not (Table 5.5). The analysis of responses to working in plating suggest that those who had read the new information did not have significantly different perceptions of their work environments and sources of information than those who had not. This is not surprising given the length of time most people had worked in plating, familiarity with the chemicals and basic understanding that this is a potentially risky working environment. In phase 1, 78% of platers supported Safety Data Sheets as sources of information and understanding over and above ‘experience’ (Table 3.6). However, in phase 2 a slight rewording of this question suggests that Safety Data Sheets do not provide all of the information platers need in their workplace. Interestingly of those who had read the new information there was evidence that they did not believe that work hazards had to be ‘part and parcel’ of working in plating, i.e. they believe that there is an onus on them to ensure safety. Those who had not read the information appeared to be less likely to think this (Table 5.5). All of the other evidence suggests that platers are aware of the risks. Therefore, it is possible that those who take the trouble to read information are also those who believe that they have a personal responsibility in the workplace. People who do not bother to read information (or to ask for it if it is not immediately available) may not have such a basic belief about responsibility. The significant difference in response between those who had read the new information and those who had not to the use of labels compared to Safety Data Sheets (Table 5.5) also could be interpreted as reflecting a different type of person who is willing to read information. In general those who had read the new information were slightly more positive about the value of Safety Data Sheets, slightly less likely to believe that health and safety information is only useful if you have work-related health problems, and more likely to find time to read information than those who had not (Table 5.5). 6.7.3 Use and accessibility of the new information It was not possible to determine whether those who had not read the new information had simply not bothered to read something that was actually accessible to them or whether the information had simply not been available. However, the visibility of the poster in some plating shops when the researcher called suggests that some people simply do not read information even when in front of them. It is possible that in the plating shop warning labels and signs have simply become too familiar. The fact that some companies appear to have either lost or thrown out the information indicates the difficulties in actually getting new material into firms. Managers are inundated with information through the post on a daily basis. Ensuring that they read and retain health and safety information is important: i.e. that they understand that this type of information should be retained. This problem emphasises relevance, however, it also emphasises the need for information to be available through multiple modes – written, verbal, audio-tapes, video, web-based etc. – so as to ensure that it is accessed by some means. The oral tradition of learning was as strongly evident in the second as in the first phase of the project. The interviews revealed the potential importance of practical, verbal dissemination alongside the new ways of presenting written information. The potential role of the chemical suppliers and also of HSE in providing verbal information to back written sources was confirmed.

59

The role of visual images as well as narrative was confirmed in this second phase, particularly amongst those who had read the new information and poster. The preference for written information with pictures as in the poster was backed by support for a video. The variation in approach to dissemination within just 20 companies, from those who had held a meeting with all platers to introduce the information, to those who had given a copy to each plater, to those who had merely filed it in the manager’s office suggests that there is a need to encourage a means of interactive delivery. Written information even when presented in a meaningful language and style still has a degree of formality and potential remoteness from practice. The results relating to platers’ perceptions of the value of the new information arising from its ‘accessibility to lay people’ suggests that it is possible that the new information will give platers more confidence in reading other material, particularly Safety Data Sheets. What is certain is that the new information cannot stand alone as an information source. It is interesting that some people thought the information ‘frightening’ – presumably because of the simplicity and hence vividness of language being used. Safety Data Sheets are conveying the same risks but in a language that is obscure and therefore lacks personal relevance. The new information is designed to be directly relevant to individual platers. ‘Frightening’ information is effective if it makes people change their way of working. However, it is ineffective if it stops managers from passing it on. The problems being faced with the minority of such managers is that their own personal level of understanding may be as poor as that of the platers (as suggested by the phase 1 data). They may be concerned that the information could prompt platers to demand new or further protection placing demands upon themselves as managers. However, the overwhelming conclusion must be that the style of the information in terms of being comprehensible to platers and prompting new knowledge is important. If presented as support to Safety Data Sheets then the fact that both are conveying the same risk information should serve to stress that the information is not designed to be frightening, merely that the risks are potentially serious. The new information has already been used as training material and the quiz style of presentation was viewed as useful in this respect, even if it was not being used in a ‘formal’ question and answer mode between a trainer and trainee. As identified earlier the level of reading skills and literacy were not measured in this study, although other studies have indicated a low level. As in phase 1 there was evidence of difficulties with the questionnaire for some platers, particularly where terminology was not understood by them. The decision to translate the information into Punjabi was considered useful by the companies who accessed it, although it was evident that a simple translation of written information cannot provide for the different levels of literacy. There were few Asian platers who could both speak and read Punjabi. Many who have been educated in England primarily only speak Punjabi within their families. There are some who can speak Punjabi but read English poorly, for whom the audio-tape was considered particularly useful. The tape was also accessible in their own time. There is no doubt that the Asian platers valued the fact that an attempt had been made to present information in an manner useful to them. This supports the argument for a user-centred approach to the design of risk messages.

60

6.8 VALUE OF THE MENTAL MODELS APPROACH

6.8.1 Conceptual Problems In the pre-project Workshop (Anon, 1999), while the potential of the mental models approach was stressed, a number of criticisms and problems were summarised from the literature: (i) it is unclear how the approach deals with situations where there is limited scientific

understanding or agreement; (ii) it might be unduly restrictive in conceptualising risk understanding in terms of

technocractic and scientific understandings; (iii) therefore, it might undervalue lay knowledge and expertise; (iv) it ignores social and cultural impacts on both cognition and behaviour; (v) where occupational experience is at odds with expert understanding, official advice may

be undermined through not being credible; (vi) it is not possible to derive a true expert model only a representation of expert

understanding at a point in time; (vii) defining the expert will impact on the model defined, however for some risks there may

be multiple expertise and multiple models all of which may be relevant and ‘accurate’ (viii) directly comparing the expert and lay model is rarely easy, and (ix) measuring the effect of the new messages is difficult. The mental-models approach can be considered to be ‘pre-modern’ in that it assumes that experts make rational decisions, but that lay people do not. In seeing expert opinion as the baseline against which other beliefs should be judged, it focuses on what experts’ perceive, not how they arrived at these opinions. Differences between experts will reflect culturally determined value judgements which are not only influenced by experience (as discussed in 6.2.1) but different scientific backgrounds (e.g. engineer or medicine or chemistry), and the different cultures of the organisation they work in (compare a regulatory agency with a commercial consultancy, for example). The mental models approach is only one tool to assist in understanding lay and expert knowledge. To expect too much of the approach is counterproductive. As already discussed this research project experienced several of the problems identified above, particularly relating to (iii), (vi), (vii), (viii), and (ix).

6.8.2 The User-Centred Approach The greatest strength of the mental-models approach is that it is (or should be) user-centred. It has allowed for more detailed understanding of the mental models of platers. The user-centred approach requires direct contact with individuals to determine what they know and believe. It focuses on the user of information and messages to determine what they need to know. This is different to the way in which many risk messages are designed and communicated. The typical process tends to stress design based on expert judgement of what non-experts need to know and communication which is judged as successful by the number of copies of a leaflet, etc. distributed or purchased. The mental-models approach has the potential to over-state expert knowledge and beliefs as being correct. In doing so it may underestimate the value of lay knowledge derived more from experience than books and translated into ability to minimise risks through appropriate action. This potential can be guarded against in its design and execution.

61

The face-to-face interview is particularly important in this regard. What the interview can achieve, which the questionnaire tends not to, is the elicitation of beliefs without putting new concepts into people’s minds, prompting responses not based on the concept of ‘this must be the right answer’, nor leaving existing views unstated. The questionnaire is useful in confirming that the views expressed in the interviews are similar to the views of a larger number of people in the sector being studied. However, designing a questionnaire which does not ‘prompt’ certain answers or produce over-confident answers (e.g. the guaranteed positive response by the majority to the statement ‘I have a good understanding of the risks involved in my work’) is difficult. It cannot be assumed that people completed the questionnaires in either phase 1 or 2 without reference to colleagues. Completing a questionnaire during work time, particularly one which appears to be ‘official’ and to be testing knowledge, is bound to engender unease among some people and to promote some collaborative responses (despite the request that people complete it by themselves). There was evident confusion over some of the multiple choice questions. However, overall the questionnaire can be useful in confirming areas where knowledge is less certain (i.e. where there is greatest difference with that of experts) and in identifying any differences within the group. It has a secondary role to the interviews. This study supports the views of other mental models researchers (e.g. Bostrom et al, 1994) that open-ended interviews come much closer to simulating the conditions in which information is read than group survey methods such as focus groups. However, group methods are more likely to be valuable in identifying the important organisational culture and climate issues which impact on mental models, as has been the case in relation to research relating to the attitudes of people in small firms to the environment and environmental compliance (Petts et al, 1999). This organisational element still needs to be tackled. The interviews for this project identified many organisational culture and climate issues which are relevant, including relating to: (i) the fundamental culture that says that plating is a risky activity and that management

systems are important; (ii) ownership of firms - not least the influence of the owner-manager who is also a plater; (iii) the impact of stability of firms and of employment in creating inertia in terms of

changes to practice but on the other hand enhancing the impact of experience on good practice;

(iv) the horizontal structures which encourage communication from management to shop-floor and vice versa;

(v) the impact of risk assessment obligations upon the generation of risk information within companies;

(vi) the lack of technical and human resources which could be applied to the production of relevant risk messages for platers not least in the small firms.

The interview, or probably more aptly termed ‘conversation with a purpose’ (as the purpose is not necessarily to test people or to convey a position of superiority), has the most potential for replication. However, the skills required to conduct a ‘conversation with a purpose’ are considerable. These are interviewing skills not necessarily technical knowledge skills. Indeed the latter could be a hindrance to the process as there could be the potential for the interviewer to direct or bias the discussion to certain technical issues which he or she identified as either being under or over stated by the interviewee. While the latter needs to have some confidence in the credibility of the interviewer if they are to feel comfortable taking part and that it is worthwhile, they are likely to feel uncomfortable if they think that the interviewer is an expert, testing their knowledge. Part of the credibility of the interview is a sympathetic understanding of the practicalities of the workplace. This not only suggests that the interviews should be

62

conducted in the workplace, but reinforces the potential problem if people perceive that the interviewer is testing them on abstract rather than working knowledge. What is evident in the interviews in both phase 1 and 2 of this study is the difficulty people have in articulating what they know and believe such that it is possible to draw direct comparison with experts. When people reply to questions with only one or two words (a problem which can also be apparent in focus groups involving people from lower socio-economic groups or young people) then considerable skill is required to tease out what they actually mean by the statement. The effective involvement of experts demands close attention to who is a relevant expert. In this case a broad view was taken encompassing chemical, medical and operational backgrounds. This proved to be an important decision as even in the relatively small group there was evident disagreement and different mental models which reflected different backgrounds, training and experience. A ‘consensus’ view was often based on 8-9 people out of 13 agreeing. Application of the mental models approach is difficult in small companies. The approach has been widely used in relation to the public who are relatively easy to identify and recruit for research purposes. This is not the case with small firms and particularly micro (<10 employee) firms. As discussed in 2.5, access is hampered by (i) lack of good records which ease identification of companies (this project being successful largely because of the assistance of the SEA), (ii) the large number of firms which have to be accessed in order to derive a reasonable sample of platers given the small number of platers in most companies, and (iii) the primary problems of gaining support and access when companies are under significant business pressures and cannot devote time to non-production activities. Difficulties of access will be different for different industrial sectors. However, the issue of company willingness to participate is similar across sectors. As discussed in 6.5, ethnicity does need to be addressed. The application of the approach in this context raises some practical issues such as potential need to use translators to conduct interviews or to assist people in the completion of questionnaires. Certainly in phase 2 the latter did create problems in the company that had previously had translation assistance during an English class. With the second questionnaire a plater had to assist colleagues with translation of the questions.

63

7.CONCLUSIONS

This study has successfully applied the mental-models approach to gain some understanding of platers’ knowledge and beliefs about the chemical risks of electroplating. Plating has provided a good ‘indicator’ sector for the test of the mental-models approach because of the significant, well-known risks which it presents and the relatively large amount of information and guidance which is available, particularly on chromic acid. Although the focus has been individuals’ beliefs and knowledge, the conduct of the study and its findings have confirmed the importance of the complex mix of information processes, institutional structures, social group behaviour and individual responses which shape the experience of, and responses to, risk. The mental-models approach only provides a partial explanation of knowledge and beliefs. Most particularly it undervalues lay knowledge and experience and over-emphasises assumed expert rationality. Compared to expert beliefs it is apparent that platers place a greater emphasis on the short-term rather than the long-term risks of the key chemicals used in plating – chromic acid and nickel. The primary source of their knowledge is direct, practical experience, which considering that the majority have worked in plating for more than 10 years is significant. The day-to-day familiarity of direct contact with hazardous chemicals and of the precautions which are necessary, combined with the experience of learning to work effectively to meet production targets while also managing the risks have an important impact on beliefs. This is an inherently risky business and people have rationalised these risks to enable themselves to feel comfortable about working in it and to do the job effectively. Platers have difficulty articulating the effects of exposure to the chemicals and how this exposure might occur. It is wrong to equate the difficulties of articulation with lack of knowledge. The difficulty partly reflects unfamiliarity with talking about the hazards and risks. However, it also reflects their focus on outcomes and the important precautions to be taken. They do not distinguish between hazards and risks and it is arguable whether they need to. Experts place a greater emphasis on the long-term risks particularly occupational asthma (not recognised by any platers) and cancer. A significant minority of platers do not understand key technical terms and hazard phrases which are common to official information on chromic acid – particularly in Safety Data Sheets. There is some evidence of fundamental literacy problems which affect understanding, including amongst those whose first language is not English. Safety Data Sheets, in adopting expert short-hand, not least to protect supplier liabilities, cannot be expected to influence beliefs if the primary language is unintelligible. Experience and learning from others are the primary means by which platers find out about the risks, although they do stress the need to also have written information, albeit that this is not used often. Platers stress the value of labels as a source of information. However, this could be over-reporting of the importance of labels as they are more likely to simply be part of everyday experience in the plating shop. The importance of the oral tradition has potential in terms of risk messages. Experts who could engage platers face-to-face to describe the hazards and risks seem important. A potential role for chemical suppliers is evident in this regard. New risk messages relating to chromic acid but readily adapted for other chemicals have been designed and tested. The rationale for the design is the need to use the language which platers understand, to answer key questions which they need to know – i.e. What can it do to me? How

64

might this happen? What should I do to reduce the risk?, and to emphasise both the short-term and long-term risks. The focus is on pragmatic and practical guidance responsive to platers’ experience and reflective of the oral tradition and also to spoken language. The new messages appear to have had a positive impact in terms of raising understanding of the long-term effects of chromic acid amongst those who have read them. They have also had an impact in terms of raising awareness of practical precautions to take in the plating shop that were not based in direct experience. The focus on practical and pragmatic guidance appears to have been successful. However, written information even in a more relevant form than Safety Data Sheets is no replacement for experience. The new information appears to be particularly useful for training although there has not been sufficient opportunity to test this. The testing of new information in a primary other language and direction to issues of ethnicity has been important but also requires considerably more work. Methodologically the project has faced a number of challenges that impact on the potential replicability of the mental-models approach across other sectors. These challenges include: (i) the difficulties of identifying and accessing small firms, (ii) the skills required to conduct effective face-to-face interviews with experienced people who may not be used to talking about the risks and who may be wary of exposing a lack of knowledge, (iii) the difficulties of designing a questionnaire which does not prompt correct or overconfident responses, (iv) the need to access divergent expert knowledge but to identify areas of expert agreement as well as disagreement, (v) the effective design of new messages particularly incorporating lay language effectively without leading to the possibility of misinterpretation, (vi) the need to consider the visual presentation of risk messages, and (vii) the difficulty of measuring the effectiveness of the new messages particularly with limited time. While the project has proactively sought to include those from the ethnic minorities it has not been possible to draw any specific comparison between Asian and English workers. This element of the work must be viewed as a pilot. The methodological difficulties combined with the inherent weaknesses of the mental models approach suggest that a more flexible approach to the design of new messages, but one which is based on a set of core principles may be useful. The suggested principles together with outstanding research requirements are presented in the next chapter.

65

8. RECOMMENDATIONS

8.1 GENERIC PRINCIPLES FOR THE DESIGN OF NEW RISK MESSAGES The mental-models approach does help in the identification of individuals’ knowledge and beliefs and particularly of some of the things people do not know. It has been able to be applied in the electroplating sector. The method was supported by the participants, and it has enabled an understanding of platers’ knowledge and beliefs to be gained. However, the time, effort and multiple skills required to this effect have been significant. While the mental-models approach has value we do not consider it to be appropriate to conclude that it must be replicated, in the form used in this project, across industrial sectors. The resources required for replication make such a recommendation impracticable. The approach’s significant deficiencies are such that methodological replication without questioning may lead to only a partial improvement in understanding of individuals’ mental models. In particular, the approach’s support of the expert view to the detriment of ‘lay’ views could be counterproductive in the workplace, undervaluing experience and practical ‘expertise’. Therefore, the report’s authors believe that a more useful way forward is to conceive of a set of generic principles derived from the approach which should underpin the design and testing of new risk messages. These principles are: (i) People who have to use risk information should have a say about the content and be

involved in its design – i.e. the user-centred approach. (ii) Managers’ and workers’ experience, beliefs and knowledge and their information

requirements should be understood primarily through face-to-face discussion with them in the workplace – i.e. the ‘conversation with a purpose’.

(iii) The discussion should aim to identify deficiencies in beliefs and knowledge so as to identify what new or different messages people need to receive to enable them to make appropriate risk reduction decisions.

(iv) The discussion needs to focus not only on what people believe, but why. (v) Divergence of views due to age, length of experience in a job, experience of

accidents/ill-health, culture (ethnicity), and possibly disability, should be expected. The user-centred approach should be inclusive to ensure that these differences are revealed.

(vi) A person’s inability to articulate the hazards and risks of specific chemicals cannot be assumed to equate to lack of ability to take appropriate actions. Therefore, the user-centred approach must focus upon what people do in the workplace to reduce risks as well as what they say they know.

(vii) The impact of organisational culture on beliefs, knowledge and behaviour must be recognised. Therefore, new risk messages must be tested not only in terms of whether they change/improve understanding and meet the criterion of ‘need to know’ but also whether they change behaviour.

(viii) It should not be assumed that there is a single expert mental model. Identification of the various experts relevant to chemical risks and understanding of their different mental models is important. In particular there is a need to consider the differences between occupational health experts and factory inspectors.

(ix) It should not be assumed that the expert mental model, particularly in terms of how hazards might be realised and what are effective risk reduction measures, is any better than that of a worker in a particular industrial sector.

(x) It should not be assumed that the expert is devoid of cultural and experiential influences

66

(xi) Risk information is most useful when it is presented in a style and format that answers key questions – i.e. what can it do to me? how might this happen? how can I protect myself?

(xii) Risk information should build on what people already know and believe and offer the opportunity for appropriate generalisation to other products/processes.

(xiii) Risk information presented in lay language is more likely to be understood. Many expert terms – including hazard phrases – are incomprehensible to people.

(xiv) Different styles, forms and language of presentation may be relevant for people whose first language is not English.

(xv) The design and layout of information, including the use of visual images, is as important as the terms and words used. Effective design requires specialist design input.

(xvi) Key industrial sectors where the user-centered approach to the design of risk messages may be appropriate include those where there is expert disagreement over the risks (and hence potential for ‘lay’ misunderstanding); where the potential for serious consequences is high, and where there is a high incidence of accidents/ill-health.

8.2 RESEARCH NEEDS This project has identified a number of research requirements related to the mental models of experts and workers in electroplating. It is not considered that these are sector-specific needs, although the electroplating sector does warrant further research. Small firms (<50 employees), and particularly micro firms (<10 employees), do warrant further attention. These research requirements are: (i) Investigation of the impact of knowledge and beliefs on behaviour, including

identification of the factors which impact on behaviour, those which lead to guidance being ignored and those which lead to change of behaviour.

(ii) Identification of the use made of printed information sources within companies and by employees (i.e. the actual ways that sources of information are consulted, interpreted and acted upon).

(iii) Analysis of the impact of the incorporation of information into training and of training material on practical experience and accidents.

(iv) Linked to (i) and (ii), analysis of the extent to which mistaken beliefs about the properties of chemicals in use exist, make a real or potential difference to the way that employees behave, and can be corrected.

(v) Identification of methods for evaluating the effectiveness of information and training. (vi) Comparison of the chemical risk perceptions of different experts – inspectors,

consultants, occupational health specialists, doctors – i.e. a ‘within’ expert group mental-models survey.

(vii) Analysis of differences of beliefs, knowledge and behaviour amongst the ethnic minorities.

(viii) Analysis of differences of beliefs, knowledge and behaviour amongst the old and the young, the able and the disabled.

(ix) Analysis of different mental models about specific chemicals between industrial sectors.

67

ACKNOWLEDGEMENTS

The research team wishes to thank all of the companies and individual managers and platers who contributed to both phases of this research. Particular thanks go to the Surface Engineering Association who provided valuable company contacts and supported the project. The help of the HSE staff in the Midlands regional office in Birmingham, including their input to the expert questionnaire, together with independent occupational health consultants has been invaluable. Finally, thanks are due to the HSE Project Board for their enthusiasm and detailed attention to the project.

68

REFERENCES

ACSNI (Advisory Committee on Safety in Nuclear Installations (1993) Third Report, Organising for Safety, Human Factors Study Group of ACSNI. HSE Books, Suffolk Anon (1999) Report on Proceedings of the Health Directorate Workshop on Risk Perception and Risk Communication – Gatwick, Jaunary 1999. EWP/99/20. Health & Safety Laboratory, HSE. Biggs, D. & Crumbie, N. (2000) Characteristics of Individuals who use Chemicals in Small Enterprises. HSE, unpublished report. Bostrom, A., Atman, C.J., Fischhoff, B. & Morgan, G.M. (1994) Evaluating risk communications: completing and correcting mental models of hazard processes, part II. Risk Analysis, 14(5), 789-798 Covello, V.T. (1992) Trust and credibility in risk communication. Health and Environmental Digest, 6(1): 1-3 Cox, S. and Flin, R. (1998) Safety culture: philosopher's stone or man of straw. Work and Stress, 12(3): 189-201 Cox, S. and Tait, R. (1998) Safety, Reliability and Risk Management. 2nd Edition. Butterworth Heinemann, Oxford Cox, S., Thomas, J.M., Cheyne, A and Oliver, A (1997) Safety Culture: The Prediction of Commitment to Safety in Manufacturing Industry. British Academy of Management Annual Conference, pp. 121-128 Craik, K. (1943) The Nature of Explanation. Cambridge University Press, Cambridge. Ferguson, E., Cox, T., Farnsworth, W. and Irving, K. (1994) Nurses anxieties about biohazards as a function of context and knowledge. Journal of Applied Social Psychology, 24(10), 926-940 DTi (Department of Trade and Industry) (1989) Guidelines for Better Instructions and Safety Information for Consumer Products. HMSO, London Eiser, J.R. (1998) Communication and interpretation of risk. British Medical Bulletin, 54: 779-790 Epstein, S. 91994) Integration of the cognitive and the psychodynamic unconscious. American Psychologist, August, 709-724 Fischhoff, B. (1995) Risk perception and communication unplugged: Twenty years of process. Risk Analysis, 15, 137-145. Fischhoff, B., Bostrom. A., Jacobs, B. and Quadrel, M. (1997) Risk Perception and Communication: Oxford textbook of Public Health. Oxford University Press, Oxford. Glaser, B and Strauss, A (1967) The Discovery of Grounded Theory: Strategies for Qualitative Research. Aldine. Chicago Glendon, A.I. and McKenna, E.F. (1995) Human Safety and Risk Management. Chapman Hall, London ILGRA (Inter-Departmental Liaison Group on Risk Assessment) (1998) Risk Communication: A Guide for Regulatory Practice. Johnson-Laird, P.N. (1983) Mental Models. Harvard University Press, Cambridge, Mass. Jungermann, H., Schutz, H & Thuring, M. (1988) Mental models in risk assessment; informing people about drugs. Risk Analysis, 8(1), 147-155 Lasswell, H (1948) The structure and function of communication in society. In L Bryson (ed) The Communication of Ideas. Institute for Religious and Social Studies, New York Lee, T.R. (1995) The role of attitudes to safety culture and how to change them. Paper presented to the Conference on Understanding Risk Perception, February, The Robert Gordon University, Aberdeen Lee, T.R. (1987) Action on Attitudes to Risk in Department of Health and Social Security Strategies for Accident Prevention. Report of a Colloquium for the Medical Royal Colleges of the Uk on 26 March , 1987. HMSO, London

69

Leiss, W (1996) Three phases in the evolution of risk communication practice. Annals of the American Academy of Political and Social Science, Vol.545, pp85-94. Löfstedt, R E and Renn, O (1997) The Brent Spar controversy: An example of risk communication gone wrong, Risk Analysis, 17(2), 131-136. MacGregor, D.G (1989) Inferences about product risks; a mental modelling approach to evaluating warnings. Journal of Product Liability, 12, 75-91 MacGregor, D.G. and Fleming, R. (1996) Risk perception and symptom reporting. Risk Analysis, 16(6), 773-783 MacGregor, D.G., Slovic, P. and Morgan, M.G. (1994) Perception of risks from electromagnetic fields: a psychometric evaluation and risk communication approach. Risk Analysis, 14, 815-818 Mearns, K., Flin, R., Gordon, R., and Fleming, M. (1998) Measuring safety climate on offshore installations. Work and Stress, 12(3), 238-254 Morgan, M.G., Fischhoff, B., Bostrom, A., Lave, L and Atman, C.J. (1992) Communciating risk to the public: first learn waht people know and believe. Environmental Science and Technology, 26 (11), 2048-2056 Nicholson, P.J. (2000) Communicating occupational and environmental issues. Occupational Medicine, 50(4):226-230. NRC (National Research Council) (1989) Improving Risk Communication. National Academy Press, Washington DC Petts, J (1997) The public-expert interface in local waste management decisions: expertise, credibility and process. Public Understanding of Science, 6, pp. 359-381 Petts, J., Herd, A., and O'hEocha, M. (1998) Environmental responsiveness, individuals and organisational learning: SME experience. Journal of Environmental Planning and Management, 41(6), pp 711-730 Petts, J., Herd, A., Gerrard, S. & Horne, C. (1999) The climate and culture of environmental compliance within SMEs. Business Strategy and the Environment, 8(1), 14-30 Renn, O (1991) Risk communication and the social amplification of risk, In Kasperson, R.E. & Stallen, P.M. (eds) Communicating Risks to the public: International Perspectives, Kluwer, Dordrecht. Renn, O. (1992) Risk communication: towards a rational discourse with the public. Journal of Hazardous Materials, 29, 465-519 Renn, O. (1998) Three decades of risk research: accomplishments and new challenges. Journal of Risk Research, 1(1), 49-71. Rothman, A. & Salovey, P. (1997) Shaping perceptions to motivate healthy behaviour: the role of message framing. Psychological Bulletin, 121: 3-19 Rundmo, T, (1992) Risk perception and safety on offshore petroleum platforms - Part 1: Perception of Risk. Safety Science, 14, 39-52 Rundmo, T. & Sjoberg, L., (1998) Risk perception by offshore oil personnel during bad weather conditions. Risk Analysis, 18(1), 111-118 Rundmo, T., Hestad, H., Ulleberg, P. (1998) Organisational factors, safety attitudes and workload among offshore oil personnel. Safety Science, 29, 75-87 Schulte, P.A., Boal, W.L., Friedland, J.M., Walker, J.T., et al. (1993) Methodologic issues in risk communications to workers. American Journal of Industrial Medicine, 23, 3-9 Sorahan, T & Harrington, M. (2000) lung cancer in Yorkshire chrome platers, 1972-97. Occupational Environmental Medicine, 57, 385-389 Weyman, A.K. & Scobbie, A. (1998) Perceptions and understandings of risk in small dry-cleaning businesses. Health and Safety Laboratory, HSE, unpublished Wogalter, M.S. , DeJoy, D.M. & Laughery, K.R. (1999) Warnings and Risk Communication. Taylor Francis, London Wogalter, M.S. & Laughery. K.R. (1996) WARNING: sign and label effectiveness. Current Directions in Psychology, 5, 33-37

APPENDIX 1

PLATERS' QUESTIONNAIRE (PHASE 1)

71

72

73

74

75

76

77

78

APPENDIX 2

EXPERTS' QUESTIONNAIRE

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

APPENDIX 3

PLATERS' QUESTIONNAIRE (PHASE 2)

97

98

99

100

101

102

103

APPENDIX 4

NEW RISK MESSAGES

104

105

106

107

108

109

Printed and published by the Health and Safety ExecutiveC30 1/98

Printed and published by the Health and Safety ExecutiveC1.25 01/02

CRR 400

£20.00 9 780717 622450

ISBN 0-7176-2245-2