risk management behaviour by the northern ireland food consumer

8
186 International Journal of Consumer Studies, 28 , 2, March 2004, pp186–193 © 2004 Blackwell Publishing Ltd Blackwell Science, LtdOxford, UKIJCInternational Journal of Consumer Studies1470-6431Blackwell Publishing Ltd, 200428 2186193 Original Article Risk management behaviourR. Nelson Correspondence Roy Nelson, Loughry Campus, College of Agriculture Food and Rural Enterprise, Cookstown BT80 9AA, UK. E-mail: [email protected] Risk management behaviour by the Northern Ireland food consumer Roy Nelson Loughry Campus, College of Agriculture Food and Rural Enterprise, Cookstown, UK Abstract The main criteria that consumers use during the decision- making process when purchasing food have traditionally been a combination of prices, incomes, taste and social attitudes, with price being seen as the main determinant. However, in the past 10 years risk has become a ‘new’ crite- rion that can affect the consumer’s decision whether or not to purchase a particular food item. The effect of the percep- tion of risk has been observed during the numerous food scares in the last decade and in trends for the demand for foods that connote a health image. This research, carried out in Northern Ireland, looked at how consumers quantified and managed risk. The research involved 202 primary food con- sumers and identified the factors that were perceived to be important from both a societal and a personal perspective. Using principal component analysis techniques, societal food risks were seen as either processing or dietary, and personal food risks were seen as either extrinsic or intrinsic. Further investigation into the attributes of the personal risk using the same techniques revealed a three-factor solution described as fear, involvement and newness. Although these factors cannot be used as predictors of the risk associated with a particular food, they help to describe and explain how the risk may be managed. The relationship between two of the factors, involvement and fear, provide a framework for understanding the way consumers manage their perception of the risk, particularly of high-risk items. Consumer and scientific knowledge of the risk in question, and the degree of control over the risk were seen as important in the man- agement of the risk in question. The conceptualization of the mechanism by which risks are processed and acted upon provides information regarding risk management and com- munication strategies that should be employed by educa- tors, food retailers and government policy makers. Keywords Risk management , risk perception , attributes of risk , consumer decision-making . Introduction Over the last decade there have been considerable changes in food purchasing behaviour. Traditionally the decision on what food to purchase was based upon price and to a lesser extent taste and quality. 1–5 However, now the consumer is very interested in food that is safe and wholesome and in food that promotes a positive health image. 6–9 Evidence of this can be observed in changes in consumer demand for foods such as those marketed as functional foods, 10,11 or for foods providing healthy options to the diet. 12 In addition, there have been dra- matic changes in behaviour during the numerous food scares that have been experienced globally such as those associated with Salmonella in eggs, 13 bovine spongiform encephalopathy (BSE) in cattle 14 and Alar in apples. 15 Consequently how the consumer perceives the concept of food safety will have implications within the whole of the food supply chain, affecting primary producers, processors and manufacturers, retailers and finally the end consumer. Defining risk Safety can be thought of as a subjective assessment of the risk associated with an object or process. There- fore to determine how safe something is requires the level of risk to be estimated. 16 Once that is deter- mined, then the acceptability of the risk level can be ascertained either for individuals, or to society as a whole. Estimates of risk are regularly based upon the premise that the risk level can be objectively deter- mined from calculations based upon data related to the probability of occurrence of a defined risk incident and data related to the magnitude of the conse- quences of that occurrence. As a result risk can be

Upload: roy-nelson

Post on 14-Jul-2016

212 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Risk management behaviour by the Northern Ireland food consumer

186

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

© 2004 Blackwell Publishing Ltd

Blackwell Science, LtdOxford, UKIJCInternational Journal of Consumer Studies1470-6431Blackwell Publishing Ltd, 200428

2186193

Original Article

Risk management

behaviourR. Nelson

Correspondence

Roy Nelson, Loughry Campus, College of Agriculture Food and Rural Enterprise, Cookstown BT80 9AA, UK. E-mail: [email protected]

Risk management behaviour by the Northern Ireland food consumer

Roy Nelson

Loughry Campus, College of Agriculture Food and Rural Enterprise, Cookstown, UK

Abstract

The main criteria that consumers use during the decision-

making process when purchasing food have traditionally

been a combination of prices, incomes, taste and social

attitudes, with price being seen as the main determinant.

However, in the past 10 years risk has become a ‘new’ crite-

rion that can affect the consumer’s decision whether or not

to purchase a particular food item. The effect of the percep-

tion of risk has been observed during the numerous food

scares in the last decade and in trends for the demand for

foods that connote a health image. This research, carried out

in Northern Ireland, looked at how consumers quantified and

managed risk. The research involved 202 primary food con-

sumers and identified the factors that were perceived to be

important from both a societal and a personal perspective.

Using principal component analysis techniques, societal

food risks were seen as either processing or dietary, and

personal food risks were seen as either extrinsic or intrinsic.

Further investigation into the attributes of the personal risk

using the same techniques revealed a three-factor solution

described as fear, involvement and newness. Although these

factors cannot be used as predictors of the risk associated

with a particular food, they help to describe and explain how

the risk may be managed. The relationship between two of

the factors, involvement and fear, provide a framework for

understanding the way consumers manage their perception

of the risk, particularly of high-risk items. Consumer and

scientific knowledge of the risk in question, and the degree

of control over the risk were seen as important in the man-

agement of the risk in question. The conceptualization of the

mechanism by which risks are processed and acted upon

provides information regarding risk management and com-

munication strategies that should be employed by educa-

tors, food retailers and government policy makers.

Keywords

Risk management

,

risk perception

,

attributes of

risk

,

consumer decision-making

.

Introduction

Over the last decade there have been considerablechanges in food purchasing behaviour. Traditionally thedecision on what food to purchase was based upon priceand to a lesser extent taste and quality.

1–5

However, nowthe consumer is very interested in food that is safe andwholesome and in food that promotes a positive healthimage.

6–9

Evidence of this can be observed in changesin consumer demand for foods such as those marketedas functional foods,

10,11

or for foods providing healthyoptions to the diet.

12

In addition, there have been dra-matic changes in behaviour during the numerous foodscares that have been experienced globally such as thoseassociated with Salmonella in eggs,

13

bovine spongiformencephalopathy (BSE) in cattle

14

and Alar in apples.

15

Consequently how the consumer perceives the conceptof food safety will have implications within the wholeof the food supply chain, affecting primary producers,processors and manufacturers, retailers and finally theend consumer.

Defining risk

Safety can be thought of as a subjective assessment ofthe risk associated with an object or process. There-fore to determine how safe something is requires thelevel of risk to be estimated.

16

Once that is deter-mined, then the acceptability of the risk level can beascertained either for individuals, or to society as awhole. Estimates of risk are regularly based upon thepremise that the risk level can be objectively deter-mined from calculations based upon data related tothe probability of occurrence of a defined risk incidentand data related to the magnitude of the conse-quences of that occurrence. As a result risk can be

Page 2: Risk management behaviour by the Northern Ireland food consumer

© 2004 Blackwell Publishing Ltd

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

187

R. Nelson

Risk management behaviour

seen as a measure of probability and severity ofadverse effects. It should therefore be a measure ofempirical fact.

17

Using this as a basis, levels of riskhave been calculated and compared for a wide rangeof objects and processes, such as the risk of smoking,travelling in the car on the motorway, or being hit bya meteorite. The broad definition has also led to anumber of empirical models that have been used tocalculate risk. For example, Sandemann

18

proposed in1986 that when estimating risk scientists should: (1)evaluate toxicity; (2) express exposure to people; (3)estimate the likelihood of harmful effects; and finally(4) calculate the level of hazard. Consequently, scien-tists and scientific advisors, using models such as thesehave been able to provide advice to government andeducational establishments regarding the likelihoodand potential risk associated with a wide range ofproducts and processes that may affect the populationas a whole. However, the risk level predicted or calcu-lated in this way have been seen to be regularly atvariance with public sentiment with respect to manyof the risks in question. In situations such as this, aconflict of opinion forms between the two groups, andwhen this happens scientists have been known to‘believe that the public are either irrational or wilfullyignorant’.

19

There may, however, be a number of reasons whythis may happen and why such an empirical approachcannot be applied to estimate risk. For example theremay be little or no data related to the negative conse-quences associated with the risk and the situationalcontext of the risk may differ from individual to indi-vidual. It has also been demonstrated that individuals(experts and non-experts) are weak at probabilisticthinking.

20,21

People generally rely on other mecha-nisms, such as their subjective assessment of the riskbased upon their own perceptions of that risk. For thisreason an alternative definition of risk may have moreutility where risk may be viewed as ‘a perceptual orsubjective response to an environmental event thatinvolves uncertain danger or the possibility of suf-fering harm or loss’.

22

Using this definition, the sub-jective assessment of an uncertain event, wouldtherefore result in risk being estimated differentlydepending upon the differing perspectives of theindividual.

Perception of risk

There are two ways by which perception can be con-sidered; one is immediate perception, which is totallyphysical, following mechanistic neural pathways and isstrictly physiologically driven, whereas the secondlevel of perception is at the interpretative level wherepsychological processes are implicated in the percep-tual act.

23,24

This is the complete act of perceiving andbetter describes the process affecting decision-making.Perception is therefore a process by which individualsinterpret data and stimuli into usable mental represen-tations of the world

25,26

or, as described by Martin

27

asa ‘process of knowing’, and is a subjective assessmentbased upon beliefs and attitudes. The perceptual pro-cess is related to a specific product or process, which isthe situation regarding food purchase and consump-tion. If a product or process is judged to be risky, riskthen becomes one criterion for deciding what courseof action should be taken. Probabilistic theory doesnot take into account situations where human judge-ment is involved, the context is not specific, and theoutcome is unknown.

20

Under these conditions ofuncertainty, people rely on four general rules knownas heuristics to help reduce complex tasks to simplerones.

Availability

This refers to whether the event can be easily imaginedor recalled, whereby frequently occurring events aregenerally easier to recall than rare events.

28

However,availability is affected by numerous factors unrelatedto frequency such as media coverage, which in manycases leads to rare causes of death being overestimatedwhereas common causes of death tend to be underesti-mated.

29,30

People are also insensitive to the problemsof small sample sizes.

31

Availability is an important cueaffecting how both experts and the layperson perceiverisks.

Overconfidence

People are very confident in the judgements made bythemselves, to the extent that some believe they canexert control over events.

32,33

In many instances this

Page 3: Risk management behaviour by the Northern Ireland food consumer

Risk management behaviour

R. Nelson

188

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

© 2004 Blackwell Publishing Ltd

appears to increase with task complexity but inverselywith knowledge. Training does not seem to reduce thistendency.

34

The rationale for this confidence is that therisk looks very small from the perspective of theindividual.

Desire for certainty

The desire to reduce anxiety is commonly illustrated asdenial, which is analogous to the concept of selectiveattention to evidence.

35

This bias extends also to selec-tive information seeking, discounting, interpretationand testing which in many cases leads people to ‘seekout information that confirms the wisdom of theirdecisions’.

36

It won’t happen to me

Overconfidence can also be observed in that peopletend to consider themselves personally immune to manyhazards whose societal risks they would readilyacknowledge. This can also be expressed as the theoryof optimistic bias.

37

Methodology

The aim of this study was to investigate how the percep-tion of risk affected the decision-making process of theNorthern Ireland food consumer and to determine therisk management strategies that may be employed inorder to reduce the risks perceived. The data collectionmethod developed was based upon the pioneeringworks of Slovic

38

in 1967 and of Starr

39

in 1969. In thisa psychometric paradigm was a central component formeasuring the main dependent variable of the study –the perception of risk associated with food.

In order to meet this aim a questionnaire was devel-oped in order to measure how food risks were estimatedin both a societal and personal context. Respondentswere asked to estimate what in their opinion would bethe consequences of either others or themselves person-ally engaging in the risks presented to them. The firstestimate reflected their societal estimate whereas thelatter reflected their personal estimate. Two lists of riskitems were drawn up. The societal list contained 15items four of which were labelled generic food risks. The

personal list contained 15 items based upon the fourgeneric food items. The risk items to be included in thequestionnaire were based upon a focus group meetingcomprising staff in the Food Technology DevelopmentBranch (FTDB) at the Loughry Campus of the Collegeof Agriculture and reflected issues that were thought tobe of concern to the general public and had receivedcoverage in the media. Details of the respective lists andthe links between them can be seen diagrammatically inFig. 1 below.

To determine societal estimates respondents wereasked to estimate the risk of people dying in NorthernIreland as a result of each of the risk items, whereas inorder to determine personal estimates respondentswere asked to estimate the risk of ill health or death tothemselves as a result of each of the risk items. In orderto collect comparative empirical data, the respondentswere initially asked first to rank each risk from lowestrisk to highest and then assign a score to each risk inquestion. The lowest value they could assign any risk

Figure 1

Diagrammatic representation of the relationship between the societal risk items and the personal food risk items.

Personal risks Societal risks

Diet related

Fat

Salt

Sugar

Temperature control of Fridges andFreezers

Food poisoning organisms

Personal hygiene

BSE

Troubles

Fire

Smoking

Motor cars

AIDS

Pollution

Home appliances

Alcohol

Sunbathing

Bicycles

Nuclear power and weapons

Food additives

Food contaminants

Food poisoning

Angel dust

Nitrates

Foreign bodies

Irradiation of food

Cling film

Pesticides

Food colourings

Food preservatives

Page 4: Risk management behaviour by the Northern Ireland food consumer

© 2004 Blackwell Publishing Ltd

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

189

R. Nelson

Risk management behaviour

was 10, and the respondents were informed that therewas no upper limit to the score, as long as it reflectedtheir own perception of the risk in question. To ensurethat all the respondents had a common ‘base point’ towork from, an additional item of lowest risk wasincluded with a value of 10. The respondents wereasked to relate each risk against this common startpoint.

The lists developed were extensively piloted on twooccasions with over 100 respondents and modificationsmade accordingly. Other sections included in the ques-tionnaire were sections on: (1) demographic and per-sonal information; (2) past experiences of food safetyincidents; (3) recent changes in purchasing behaviour;and (4) the characteristics of risk. The final section wasdeveloped in order to ascertain the factors that areinfluential in framing perceptions. This section askedrespondents to rate each of the 15 food risk itemsagainst nine risk attributes using a seven-point scale.The nine risk attributes were based on those factorshighlighted in the literature review and after discus-sions with members of the FTDB at the LoughryCampus. The nine risk attributes are described inTable 1.

Food purchasers who were decision-makers withinhouseholds in Northern Ireland were selected to com-plete the questionnaire. Groups of people were identi-fied as the most efficient way to collect the data and thequestionnaire was administered in six areas of NorthernIreland. This geographic spread was thought to give a

fair representation of the Northern Ireland population.It was decided to limit the data collection phase to3 weeks in order to make sure that no new food scarewould affect the results. In total, 203 people agreed tocomplete the questionnaire, but 18 were incomplete,giving a sample population of 185. Seventeen staffwithin FTDB at the Loughry Campus who had not beeninvolved in either the focus groups or the pilot study,but with tertiary level qualifications in food also com-pleted the questionnaire, resulting in a sample popula-tion in the study of 202.

For the two risk sections, each risk item was printedon a 12.5 cm

¥

7.5 cm card. Each respondent was askedto study the set of cards relating to the societal risks,and for each item, think of all the ways in which, as aconsequence of the activity or technology, people inNorthern Ireland were at risk of dying as a result of theissue in question. After this the respondents were thenasked to place in order the 15 items, as they perceivedthem from least riskiest to most riskiest. When this wascompleted the interviewees were then asked to assign avalue against each risk in question using the scale aspreviously described. This was repeated for the secondlist only this time they had to consider the risk to them-selves personally. After completion of this section therespondents then scored each of the 15 food risksagainst the risk attributes. In total, the questionnairetook approximately 2 h to complete.

Results

Analyses of the risk estimate scores were based upontheir geometric means. To obtain these, the raw datawere transformed by taking the log

10

of each score, aver-aging the score and then taking the antilog. This enabledstandardization of the scores.

Factors affecting the estimation of food risk

After analysis three factors were found to be influential.

Level of formal education

A total score for the societal and personal risk items wascalculated and it was observed that the estimation ofrisk increased as level of formal education increased.This trend can be observed in Fig. 2.

Table 1

Risk attributes descriptions

Risk attribute Description

Voluntariness A voluntary activityImmediacy Affecting the respondent immediately or some

later timePersonal knowledge Being known by the respondentScientific knowledge Being known by scientistsPersonal control Able to be avoided through personal skill and

diligenceNewness Being new and novelCatastrophic Having the potential to kill or cause illness to a

great number of peopleDread Causing fearSeverity The potential to cause ill health or death

Page 5: Risk management behaviour by the Northern Ireland food consumer

Risk management behaviour

R. Nelson

190

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

© 2004 Blackwell Publishing Ltd

The effect of personal perception of risk on the perception of societal risk

The relationship between these two influences can beseen in Fig. 3 were there is a strong correlation(

r =

0.587) indicating that with an increased perceptionof risk from a personal perspective there is an increasedperception of that risk in society, and vice versa.

The effect of past experience

This was seen as having an effect whereby respondents

who suffered a food safety incident in the past 2 yearsperceived the risk personally to themselves as signifi-cantly higher than those that had not. This effect wasobserved not just affecting the generic food groupwhere this incident took place but across the whole foodchain, a process labelled in the context of this researchas ‘risk inoculation’.

Conceptualization of the risks

Further investigation of the food risk items was carriedout using principal component analysis. This was carriedout for both the societal and personal food risks items.Initially the focus groups had described food risks underfour generic headings. However, the analysis revealed atwo-factor solution for both societal and personal risk.These bivariate constructs represented 90% and 65% ofthe variance, respectively, and Figs 4 and 5 illustrate theresults of the component plots indicating that in bothinstances the consumer simplifies the decision-makingprocess. Figure 4 describes the component plot for soci-etal risk. In this the risks associated with food process-ing, food contaminants, and food poisoning wereperceived as a single construct and was labelled pro-cessing risks. The remaining risk item was perceivedseparately and was labelled dietary risk. Figure 5describes the component plot personal risk. One con-struct was composed of nitrates, irradiation, foodcolourings, food preservatives, pesticides, angel dust,BSE, cling film and foreign bodies. This was labelledextrinsic risk. The second construct was composed offood poisoning, fat, salt, sugar, personal hygiene, andtemperature control of fridges and freezers. This waslabelled intrinsic risk.

Management of risk

As described, respondents scored each of the 15 per-sonal food risk items against nine attributes using aseven-point scale. This data was also factor analysed todetermine the internal structure of these attributes thataffect the expression of the risk in question. Thisrevealed a three-factor solution representing 57% of thevariance and is represented in Fig. 6. These werelabelled as fear, composed of the variables dread, sever-ity, catastrophe and immediacy; involvement, composed

Figure 2

The effect of level of formal education on risk perception.

228 463 529 719887

12241713

1934

2625

4455

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

No Formal(n = 39)

O’ & ‘A’ levels(n = 50)

Cert/Dip(n = 38)

Degree(n = 38)

Postgrad(n = 20)

Formal education level

Geo

met

ric

Mea

n o

f p

erce

pti

on

of

risk

Societal perception Personal perspective

Figure 3

Correlation scatter plot diagram relating personal and societal risk perception.

Personal risk perception of food

706050403020100

Soc

ieta

l ris

k pe

rcep

tion

of fo

od

20

18

16

14

12

10

8

6

4

2

0

Page 6: Risk management behaviour by the Northern Ireland food consumer

© 2004 Blackwell Publishing Ltd

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

191

R. Nelson

Risk management behaviour

of the variables personal and scientific knowledge, con-trol and voluntariness; and newness composed of a sin-gle variable.

Based upon these findings a model of food risk per-ception was drawn up relating these componentstogether and which is illustrated in Fig. 7.

Discussion and conclusions

A number of key issues were seen to be affecting howthe Northern Ireland food consumer perceived and sub-sequently managed food risks. Although conceptuallyfood risks may be thought of more complexly the con-sumer simplifies the decision-making environment bygrouping food risks in a different manner to which maybe initially envisaged by food experts. Instead of havinga situation were the consumer deals with four broadgroups of risk they reduce this to two groups. This wasevident in both the responses to the questions concern-ing food risks perceived in both a societal and personalperspective. Although viewed from either perspectivethere are similarities between the two groups – extrinsicrisks can be equated as processing risks and intrinsicrisks can be equated to dietary risks. The grouping alsoreflects the sources of the risks. Processing and extrinsicrisks can be seen as risk ‘added’ to the food throughintervention and are ‘man-made’ or artificial, whereasdietary and intrinsic risks can be seen as ‘inherent’ andare natural. Using this simplified model of how risks are

Figure 6

Principal component analysis diagram of the nine risk attributes for the 15 food risks.

0.50.0-0.5

0.6

0.4

0.2

0.0

-0.2

-0.4

First factor

Sec

ond

fact

or

Severe

Dread

Catastrophic

New

Control

Scientific knowledge

Personal knowledge

Immediacy

Voluntariness

Figure 7

Model of food risk perception.

Processing Extrinsic risks risks Societal Personal Dietary Intrinsic risks risks Risk perception Purchasing behaviour Education Risk Experience Attributes Fear Involvement Newness

Figure 5

A component plot of the factor analysis of the personal food risk items.

0.0-0.1-0.2-0.3-0.4-0.5-0.6-0.7-0.8-0.9

0.4

0.3

0.2

0.1

0.0

-0.1

-0.2

-0.3

-0.4

-0.5

First factor

Sec

ond

fact

or

PreservativesColour

BSE

T. ControlP. Hygiene

Poisoning

SaltSugar

Fat

F. Bodies

A. DustPesticides

IrradiationNitrates

C-Film

Figure 4

Component plot of the factor analysis of the societal risks.

0.0-0.1-0.2-0.3-0.4-0.5-0.6-0.7-0.8-0.9

0.4

0.2

0.0

-0.2

-0.4

-0.6

First factor

Sec

ond

fact

or

Processing

Poisioning

Diet

Contaminants

Page 7: Risk management behaviour by the Northern Ireland food consumer

Risk management behaviour

R. Nelson

192

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

© 2004 Blackwell Publishing Ltd

perceived two other important findings were revealed.The effect of past experience affected the perception ofrisk. This was as expected. However, the effect was notseen just within the food group the incident occurred,but affected the perception of risk across all the groups.This ‘inoculation’ effect can cause severe distortionswithin the food supply chain through a domino effect.Also if the risk was seen to be low from a personalperspective it was generally seen to be a small risk tosociety and vice versa. This has implications for thoseinvolved in informing others, either in advisory or edu-cational roles as the decisions reached are in the largepart based upon conditions of uncertainty and will relyin many instances upon the risk heuristics previouslydescribed. It should be noted that within this study therewas no significant differences seen between the expertand non-expert groups, except for a small number ofitems. However, it was observed that the experts werevery confident in their own judgements, which underconditions of uncertainty, could lead to incorrect advicebeing given to others. An effect was also seen of increas-ing formal education increasing the perception of riskof food across all the risk items. This does not, however,indicate that those with more years of formal educationare more correct in their perceptions, it is that they areestimating the risk from the perspective of their ownbackgrounds and experiences.

The analysis of the risk attributes, however,revealed a mechanism that the consumers use whendealing with or managing risk associated with food. Asdescribed the nine attributes were labelled fear,involvement and newness. On investigation of howstrong these constructs were across the two foodgroups it was seen that the fear component was simi-lar in each group. This is as expected, as no matterhow you may become ill or subsequently die it is a‘fearful’ event. However, involvement was seen to behigher in the intrinsic group and if involvement fell(for both groups) the fear component increased. Thiswas especially so for the high risk items. Involvementis therefore the key to the management of risk as itdescribes the degree of participation and control therespondents have in the decision-making process. Theinvolvement process represents the knowledge peoplehave about the risk in question, including any possiblerisk aversion strategies. This can be observed, where

lack of involvement is expressed during ‘food scares’and the only mechanism is total avoidance. Newnesswas not seen to be a factor in the management of riskalthough extrinsic risks were generally seen to be thenew risks.

Recommendations

Clear communication with the consumer is essential toavoid severe distortions in the market place. This com-munication should contain the message of how to ‘deal’with the risk in question and provide coping strategies.If the consumer is involved in the process and providedinformation in a transparent manner then the risk ismore likely to be managed and dealt with on a personallevel.

Those who are in the position of offering advice toothers should be aware of the effect of their own per-ceptions and biases affecting this advice and be asimpartial as possible at all times. This is especially sowhen the risk in question has characteristics that requiredecisions to be made under conditions of uncertainty.Consequently uncertainty should be acknowledged inany risk communication.

All sectors of the food supply chain should beaware of the effects that food safety incidents haveupon their industry. Past experience can sensitize indi-viduals to the potential of food risks and create adomino effect whereby the perception of risk in onesector is transferred to all other sectors in the supplychain. This heightened awareness of risk by the gen-eral public should be acknowledged and sectors notinitially involved in the risk incident should not becomplacent of the effects that this may have upontheir own businesses.

References

1. Ellis, K. & Uncles, M.D. (1991) How private labels affect consumer choice.

British Food Journal

,

93

, 41–49.2. Wheelock, J.V. (1986) Coping with change in the food

business.

Food Marketing

,

2

, 20–45.3. Slattery, J. (1986) Current diet and health regulations –

the response of the food industry.

Food Marketing

,

2

, 180–186.

4. Gafton, I. & Ness, M. (1991) Meaty horrors and poisoned delights.

Times Higher Education Supplement

,

8

, 16.

Page 8: Risk management behaviour by the Northern Ireland food consumer

© 2004 Blackwell Publishing Ltd

International Journal of Consumer Studies,

28

, 2, March 2004, pp186–193

193

R. Nelson

Risk management behaviour

5. Woodward, J. (1988)

Consumer Attitudes to Meat and Meat Products.

Food Policy Research Unit, University of Bradford. Horton Publishing Limited, Bradford, Yorkshire.

6. Johnson, M. & Mitchell, A. (1989) Crisis in food.

Marketing

,

October

, 24–29.7. Bartlam, M.J. (1993) Health and light foods.

Understanding the consumer.

British Food Journal

,

95

, 3–11.8. Spencer, C. (1991) Flesh loses its allure.

Weekend Guardian

, 5–6 October.9. Beardsworth, A.D. & Keil, E.T. (1991) Vegetarianism,

veganism and meat avoidance: recent trends and findings.

British Food Journal

,

93

, 19–24.10. Heasman, M. & Mellentin, J. (2001)

The Functional Food Revolution – Healthy People, Healthy Profits?

Part 1. A New Era in Food Health. Earthscan Publications Ltd, London. pp. 3–19.

11. Hasler, C. (1996) Functional foods: the western perspective.

Nutritional Reviews

,

54

, 6–10.12. Hunt, J.R. (1994) Nutritional products for specific health

benefits: food, pharmaceuticals, or something in between?

Journal of the American Dietetics Association

,

94

,

151–153.13. Hornsby, M. (1989) Sales of eggs after edwina currie.

The Times

, 7 April.14. Tilston, C.H., Sear, R., Neale, J.P. & Gregson, K. (1992)

The effect of BSE: consumer perceptions and beef purchasing behaviour.

British Food Journal

,

94

, 23–26.15. Begley, S. & Hager, M. (1992) A guide to the grocery.

Newsweek

, March 27, p. 20.16. Royal Commission on Environmental Pollution (1998)

Getting Environmental Standards.

21st Report. Chapter 4, pp. 51–62. ISO.

17. Moseley, B. (1990)

Food Safety: Perception and Research.

In

Royal College of General Practitioners Members Handbook.

pp. 346–351. RCGP, London.18. Sandemann, P.M. (1986) Risk communication: facing

public outrage.

Environmental Protection Agency

,

13

, 21.19. Blake, E.R. (1995) Understanding outrage: how scientists

can help bridge the risk perception gap.

Environmental Health Perspectives

,

103

(Suppl. 6), 123–125.20. Freudenburg, W.R. (1988) Perceived risk, real risk: social

science and the art of probabilistic risk assessment.

Science

,

242

, 44–49.21. Krantz, D.H., Fong, G.T. & Nisbett, R.E. (1983)

Formal Training Improves the Application of Statistical Heuristics to Everyday Problems

. Institute for Social Research, University of Michigan, Ann Arbor, MI.

22. Millburn, T.W. & Billings, R.S. (1976) Decision making perspectives from psychology.

American Behavioural Scientist

,

20

, 111–126.23. Gregory, R.L. (1983) Visual illusions.

Scientific American,

219

, 66–67.24. Gross, R.A. (1992)

Psychology: the Science of Mind and Behaviour

, 2nd edn. Hodder & Stoughton, London.25. Gregory, B.L. (1966) Eye and Brain. Wiedenfeld &

Nicolson, London.26. Coon, D. (1983) Introduction to Psychology, 3rd edn. West

Publishing Co, St. Paul, MN.27. Martin, R.S. (1969) The Psychology of Consumer

Behaviour. Prentice–Hall, Englewood Cliffs, NJ.28. Tversky, A. & Kahneman, D. (1971) The belief in the law

of small numbers. Psychological Bulletin, 76, 105–110.29. Henson, S. & Traill, B. (1993) The Demand for Food

Safety, Market Imperfections and the Role of Government. Food Policy, 18, 152.

30. Fischhoff, B., Lichtenstein, S., Slovic, P., Derby, S. & Keeney, R. (1981) Acceptable Risk. Cambridge University Press, Cambridge.

31. Edwards, W. (1982) Conservatism in human information processing . In Judgement Under Uncertainty: Heuristics and Biases (ed. by D. Kahneman, P. Slovic & A. Tversky), pp. 359–369. Cambridge University Press, Cambridge.

32. Strickland, L.H., Lewicki, R.J. & Katz, A.M. (1966) Temporal orientation and perceived control as determinants of risk taking. Journal of Experimental Psychology, 2, 143–151.

33. Ross, M. & Fletcher, G.J. (1985) Attribution and social perception. In Handbook of Social Psychology, 3rd edn, (ed. by G. Lindzey & E. Aronson), Vol. II, Random House, New York, NY.

34. Lichtenstein, S. & Fischhoff, B. (1980) Training for calibration. Organisational Behaviour and Human Performance, 26, 149–171.

35. Lord, C., Lepper, M.R. & Ross, L. (1979) Biased assimilation and attitude polarisation: the effects of prior theories on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098–2110.

36. Nisbett, R.E. & Ross, L. (1980) Human Inference: Strategies and Shortcoming of Social Judgement. Prentice Hall, Englewood Cliffs, NJ.

37. Weinstein, N.D. (1989) Optimistic biases about personal risks. Science, 244, 1232–1233.

38. Slovic, P. (1967) Perception of risk. Science, 236, 280–285.39. Starr, C. (1969) Social benefit versus technological risk.

Science, 165, 1232–1238.