empathy, resilience and distress in norwegian and english

Empathy, resilience and distress in Norwegian and English investigative interviewers.

Thesis submitted in partial fulfilment of the degree of

Doctorate in Clinical Psychology

(DClinPsy)

University of Leicester

By

Stine Iversen

October 2020

Declaration

I confirm that this thesis and the research reported within it compromises my own work. It

was written and submitted for the degree of Doctorate of Clinical Psychology (DClinPsy). It

has not been submitted for any other academic award. The reported thesis has been checked

for errors prior to submission.

Stine Iversen

2

Empathy, resilience and distress in Norwegian and English investigative interviewers.By Stine Iversen

Thesis abstractThere has been increased scrutiny of occupations who less obviously experience adverseimpacts from indirect exposure to trauma. This thesis aimed to understand occupationaldistress in non-traditional professions, who might similarly experience secondary traumathrough exposure to traumatogenic material; legal profession and law enforcement.Literature ReviewA systematic literature review explored the prevalence and predictors of secondary trauma inlegal professions. Repeated exposure to clients who have experienced trauma comprises asignificant part of lawyers’, attorneys’, solicitors’ and judges’ roles. The search elicited tenarticles for review. Quality appraisal showed significant methodological limitations in thestudies. Still, results showed a high prevalence of secondary trauma, compared to populationgroups and other professional groups. Predictors included gender, work experience, personaltrauma and level of exposure. Findings and conclusions are discussed and indicate that thesignificant variability between studies suggest a need for conceptually consistent andscientifically robust studies in the future.Research ReportThe research report examined secondary trauma prevalence and predictors in English andNorwegian investigative interviewers, as well as the potential role of empathy in thedevelopment of secondary trauma. Police officers whose work task involved investigativeinterviewing were invited to complete an online survey, consisting of questionnairesmeasuring psychological distress, secondary trauma, empathy and perceived organisationalsupport. The results indicated moderate levels of distress and secondary trauma, which werefound to be higher than reported national and international norms. Secondary trauma wasfound to correlate with wider psychological morbidity, and predictive models of distress andresilience were suggested. Findings emphasised the need to implement organisational supportstructures and resilience-building schemes to sustain police officers in key roles to ait thevulnerable and secure prosecutions.

3

Acknowledgements

I would like to acknowledge various people who have supported me through this journey in

recent years as I have worked on this thesis. First, I owe an enormous gratitude for the endless

moral support extended with love, by my parents and grandparents, whose financial support

and passionate encouragement made it possible for me to complete this project. I wish to

extend a special thanks to my dad who let me brainstorm ideas and listened to my worries and

thought processes, without fully understanding exactly what I was saying.

A sincere thanks to my supervisor Professor Noelle Robertson, without whose

inspiration, help and support, this project could not have happened. I appreciate your time,

your guidance and your thoughtful sharing of knowledge and reassurance. I also wish to thank

my secondary supervisor in Norway, Professor Eva Oddrun Langvik, who agreed to cooperate

on my research and contribute towards participant collection, despite the long distance and

many challenges along the way.

To my wonderful cohort, who has shared this experience with me over the last three

years. Thank you for the support, the laughs and the encouragements. I truly appreciate the

love and friendships I have gained, and I am so proud of all of us.

Finally, a dear thank you to my friends in Norway, Kjersti, Henriette and Ingjerd who

always have my back. Thank you for spending time with me during this process, planning and

dreaming of our future and travels. I truly appreciate how you always welcome me back as if

no time has passed.

4

Word CountsSection Word count

Thesis Abstract 291

Literature review 7027

Abstract 160

Full text (excluding reference list, figures and tabulated data) 5569

Main research report 8760

Abstract 186

Full text (excluding reference list, figures and tabulated data) 6522

Appendices* (excluding figures and tabulated data) 2540

Total word count (excluding thesis abstract, tabulated data, figures,

references and mandatory appendices) 18327

*Word count for appendices excludes mandatory appendices as specified in the Coursework

Guidelines and Assessment Regulations Handbook.

5

Table of contentsDeclaration..................................................................................................................................2Thesis abstract.............................................................................................................................3Acknowledgements.....................................................................................................................4Word Counts............................................................................................................................... 5Table of contents.........................................................................................................................6List of Tables.............................................................................................................................. 8List of Figures............................................................................................................................. 9List of Appendices.................................................................................................................... 10List of Abbreviations................................................................................................................ 11Section One: Literature Review................................................................................................12

Abstract................................................................................................................................. 13Introduction........................................................................................................................... 14

Occupational distress in the legal profession.................................................................... 14Conceptual clarification.....................................................................................................15Why examine secondary trauma in legal professionals?...................................................16

Method.................................................................................................................................. 16Study design...................................................................................................................... 16Search strategy...................................................................................................................16Selection criteria for studies.............................................................................................. 17Data extraction...................................................................................................................19Quality appraisal................................................................................................................19

Results................................................................................................................................... 19Elicited studies’ characteristics......................................................................................... 19Methodological features.................................................................................................... 20Methodological quality of included studies.......................................................................21Prevalence of secondary trauma........................................................................................ 23Predictive factors............................................................................................................... 23

Discussion............................................................................................................................. 26Prevalence and measurement.............................................................................................26Predictors and associated factors.......................................................................................27Methodological limitations................................................................................................27Implications....................................................................................................................... 28Strengths and limitations................................................................................................... 29Future research.................................................................................................................. 29Conclusions....................................................................................................................... 30

References............................................................................................................................. 31

6

Section Two: Main research report...........................................................................................35Abstract................................................................................................................................. 36Introduction........................................................................................................................... 37

Investigative interviewing................................................................................................. 37Conceptual framing of occupational distress.....................................................................38Resilience...........................................................................................................................39Empathy and distress......................................................................................................... 39Aims...................................................................................................................................40

Method.................................................................................................................................. 41Research design................................................................................................................. 41Ethical approvals............................................................................................................... 41Participants........................................................................................................................ 41Recruitment....................................................................................................................... 41Procedure........................................................................................................................... 42Measures............................................................................................................................42Data analysis......................................................................................................................44Missing data.......................................................................................................................45Prospective Power and sample size analysis..................................................................... 45

Results................................................................................................................................... 46Respondents.......................................................................................................................46Test of reliability............................................................................................................... 46Demographics....................................................................................................................46What levels of distress and resilience (CF/CS) are reported by investigative interviewersin Norway and the UK?..................................................................................................... 48What factors are associated with reported distress and resilience?................................... 52To what extent do noted correlates predict reported distress and resilience?................... 54Does reported empathy moderate/mediate the relationship with CS/CF and proposedpredictor variables in investigative interviewers?............................................................. 54

Discussion............................................................................................................................. 56Summary of findings......................................................................................................... 56Strengths and limitations................................................................................................... 58Implications....................................................................................................................... 58Future research.................................................................................................................. 59Conclusion......................................................................................................................... 60

References............................................................................................................................. 61Appendices:...............................................................................................................................66

7

List of TablesList of Tables Page number

Section One: Literature Review

Table 1. PICOS-statement informing inclusion and exclusioncriteria.

17

Table 2. Quality Appraisal scores using Appraisal Tool forCross-Sectional Studies

22

Section Two: Research Report

Table 3. Reliability analysis with Cronbach’s alpha values for

each of the measures and subtests.

46

Table 4. Sample characteristics of the UK and the Norwegian

populations.

47

Table 5. One-sample t-tests comparing ProQOL sample means

with norm data

48

Table 6. One-sample t-tests for HADS anxiety and depression

scores compared to UK and Norwegian population

norms.

49

Table 7. Means and Standard deviation on measures for each

participant group, with chosen statistical analysis, effect

sizes and p-values.

50

Table 8. Percentages of distribution of level ranking for each of

the measures, between Norwegian and UK participant

group.

51

Table 9 Bivariate non-parametric correlations between ten

variables

53

Table 10 Simple regression model for STS on anxiety. 55

Table 11 Multiple linear regression model of predictors of CS and

burnout.

55

Appendices

Table 12 Study characteristics and results relating to researchquestions.

77

Table 13 Test of Normality and assumptions statistics 110

8

List of FiguresList of Figures Page number


Figure 1 PRISMA chart of shortlisting process 18

Appendices

Figure 2 SPSS output of Normal Q-Q Plot for ProQOL

Compassion Satisfaction

110

Figure 3 SPSS output of Normal Q-Q Plot for ProQOL burnout 111


Secondary Traumatic Stress

111

Figure 5 SPSS output of Normal Q-Q Plot for HADS Depression 112

Figure 6 SPSS output of Normal Q-Q Plot for HADS Anxiety 112


Cambridge Behaviour Scale – Empathy Quotient

113

Figure 8 SPSS output of Normal Q-Q Plot for Survey of

Perceived Organisational Support.

113

9

List of AppendicesMandatory appendices are marked with an asterisk (*)

Page number

Appendix A* Guidelines to authors for journal targeted for literature

review

67

Appendix B Expanded details on search methodology 73

Appendix C Data extraction form 74

Appendix D* AXIS Quality appraisal tool items 75

Appendix E Table of findings for literature review 77

Appendix F* Guidelines to authors for journal targeted for research

project

87

Appendix G* Ethics letter of approval from UK, UOL 90

Appendix H* Norwegian ethics approval from Regional committee for

medical and health related research ethics (written in

Norwegian)

92

Appendix I* Participant information sheet 94

Appendix J* Online consent form 96

Appendix K* Complete online questionnaires, with public measures 97

Appendix L Exploration of missing data 108

Appendix M Evaluation of assumptions 109

Appendix N Q-Q plots with normality tests 110

Appendix O* Checklist assuring confidentiality and anonymity of

participants.

114

Appendix P* Chronology of research 116

Appendix Q* Statement of epistemological position taken 117

10

List of Abbreviations.AUDIT World Health Organisation’s Alcohol Use Disorders Identification Test

AXIS Appraisal tool for cross-sectional studies

BPNSW Basic Psychological Needs Satisfaction at Work

CBI Copenhagen Burnout Inventory

CES-D Center for Epidemiologic Studies Depression Scale

CF Compassion fatigue

CS Compassion satisfaction

DASS-21 The Depression, Anxiety and Stress Scale

EQ Cambridge Behaviour Scale – Empathy Quotient

HADS Hospital Anxiety and Depression Scale

HSCL-25 The Hopkins Symptoms Checklist-25

IDLS International Depression Literacy Survey

K-10 Kessler Psychological Distress Scale

LEC-5 Life Events Checklist for DSM-5

MBI-GS Maslach’s Burnout Inventory – General survey

MHP Mental Health Practitioners

MRA Multiple regression analysis

PCL-5 PTSD Checklist for DSM-5

POS Perceived organisational support

ProQOL Professional Quality of life Scale

PTSD Post-Traumatic Stress Disorder

SDS Sheehan Disability Scale

SPOS Survey of Perceived Organisational Support

STS Secondary traumatic stress

STSS Secondary Traumatic Stress Scale

TABS Trauma and Attachment Belief Scale

TIPI Ten Item Personality Measure

VT Vicarious Trauma

VTS Vicarious Trauma scale

WHOQOL-

BREF

World Health Organisation’s Quality Of Life –BREF

11


Prevalence and predictors of secondary trauma in the legal profession: A systematic review

Target journal: Psychiatry, Psychology and Law

See Appendix A for author guidelines

Presented here using BPS Style guidelines.

12

AbstractThere has been increased scrutiny of occupations who less obviously experience adverse

impacts from indirect exposure to trauma. In legal professions, repeated exposure to clients

who have experienced trauma comprises a significant part in the role of lawyers’, attorneys’,

solicitors’ and judges’. The current review aimed to explore the prevalence and risk factors of

secondary trauma reported by such legal professionals. A systematic search using seven

psychological and legal databases elicited ten articles for review. Quality appraisal revealed

several methodological frailties in the reviewed articles, and therefore the findings should be

interpreted with caution. Results indicated a higher prevalence of secondary trauma in legal

professionals, compared to both the general population and other professional groups.

Synthesis also highlighted predictors and correlates of secondary trauma, notably gender,

work experience, personal trauma and level of exposure. The review emphasised significant

variability between the studies making systematic comparisons challenging, as well as the

need for further conceptually consistent and scientifically robust studies.

13

IntroductionRepeated exposure to clients who have encountered trauma is a dimension of working within

legal occupations. Indeed, legal professionals, particularly those employed in criminal law,

will experience work with vulnerable victims of crime. This exposure implies contact with

traumatogenic material, such as graphic, injurious photographic evidence or witness accounts,

and narratives of traumatic events accompanied by heightened levels of emotions and distress

from clients. Work tasks may involve direct and indirect experience of strong emotions

through consultation with clients pursuing custody and liability battles, imprisonment,

managing court proceedings or issuing sentences, which may also imply trauma exposure.

Given the frequency of such work-related exposure, legal professionals may themselves be

vulnerable to adverse psychological impacts (James, 2020).

In 2013, the Diagnostic and Statistical Manual of Mental Disorders (DSM-5;

American Psychiatric Association, 2013) included indirect occupational exposure as a viable

criterion for the Post-Traumatic Stress Disorder (PTSD) diagnosis. Symptoms of PTSD

include intrusions, avoidance, negative alterations in cognition/mood, and alterations in

arousal and reactivity. The reported criterion related to indirect exposure specifies possible at-

risk professions, such as first responders and health professionals (Greinacher et al., 2019).

In recent years there has been greater awareness and research examining the potential

adverse effects of indirect trauma, particularly on health and social care professionals; notably

nurses (Yang & Kim, 2012), mental health professionals (Collins & Long, 2003), therapists

(Canfield, 2005) and social workers (McFadden et al., 2014). Recently research has

broadened its scope to other occupations that may also potentially experience adverse impacts

from indirect exposure to trauma.

Occupational distress in the legal professionFor this review, the term “legal profession” encapsulates several different roles: barristers;

attorneys; solicitors; lawyers; and judges. These professionals, often associated with high

status and high earning, also report significant indices of psychological morbidity, such as

high levels of clinical depression, stress (Parker, 2014) and excessive alcohol use (Krill et al.,

2016).

Furthermore, a growing body of research shows that other groups involved in legal

processes experience distress. Up to 50 per cent of criminal jurors reported trauma-related

symptoms following their involvement in criminal court (Lonergan et al., 2016), known as

14

juror stress (Miller & Bornstein, 2004). It follows that other legal professionals exposed to

similar traumatic processes and materials may be similarly affected.

Additionally, an expanding evidence base suggests that other professions working

within the legal and justice system exhibit adverse effects from indirect exposure to trauma;

notably police officers have been extensively researched compared to other legal professions.

Reviews of adverse impacts for police officers following exposure to traumatised individuals

indicated that gender and personal trauma history were potential risk factors, whilst

peer/social support, use of humour and organisational recognition of distress were potential

resilience factors (Greinacher et al., 2019, MacEachern et al., 2019). Related professions

might show similar predictors from indirect exposure to trauma.

Conceptual clarificationThe adverse impacts of exposure to other’s distress was first identified amongst those working

in human services and constructed as ‘burnout’. Initially conceived by Maslach (Maslach &

Jackson, 1981; Maslach et al., 1996), burnout encompasses exhaustion, depersonalisation,

reduced personal accomplishment and cynicism due to work requiring intense emotional

involvement with clients over time, culminating in occupational exhaustion.

Emerging from psychotherapeutic encounters, vicarious trauma (VT) was described by

Pearlman and Saakvitne (1995). They noted changes in therapists’ inner schemata about their

views of the world and their safety within it, via exposure to traumatised clients. Discriminant

validity proposed that VT is distinct from other concepts of secondary trauma due to the

transformative cognitive processes that occur (Baird & Kracen, 2006).

At much the same time, Figley (1995) noted PTSD-like symptoms in social workers

who had been exposed to traumatised clients, including symptoms such as hyperarousal,

flashbacks and avoidance. This distress with trauma-like features was termed secondary

traumatic stress (STS), arising via four mechanisms: empathetic engagement; personal trauma

history; unresolved trauma; and working with childhood trauma. Figley also coined the term

Compassion Fatigue (CF), using the term interchangeably with STS. In 2010, Stamm

redefined the term CF given the valence of the word fatigue, instead suggesting both STS and

burnout as two constituent elements forming CF. Further refinement of the construct was

conducted by Nolte et al. (2017), whose review offered a model explaining trigger factors,

physical and emotional symptoms. Symptoms included fatigue, physical pain, disturbed sleep,

withdraw and hopelessness.

15

Thus indirect trauma exposure can contribute to forms of occupational distress which

are variably and inconsistently described as burnout, VT, STS, and CF. Attempts have been

made to provide clarification, however overlap in descriptions and domains remains, core

features are contested (Elwood et al., 2011), and diverse operationalisation and measurement

has generated a cadre of research under differing conceptual frameworks. For the purposes of

this review, the collective term of secondary trauma will be used to encapsulate the

phenomena as mentioned earlier (PTSD, burnout, VT, STS and CF), thought to arise from

effects of indirect trauma exposure.

Why examine secondary trauma in legal professionals?Given that legal professionals appear regularly exposed to traumatogenic material, a critical

appraisal and synthesis of the extent of any secondary trauma and potential risk factors seems

warranted. A recent review (James, 2020) examined the history of trauma-informed legal

practise and recent studies of lawyer’s wellbeing, while discussing the recognition of

secondary trauma within the legal profession. This article also puts great emphasis on the

relative paucity of studies so far and the need for organisational recognition and structural

trauma-informed change. To date, there has been no systematic attempt to summarise,

appraise and synthesise published research. While such work-related distress will be captured

variously, examining quantitative research focusing on prevalence and risk factors under the

rubric of secondary trauma should summarise the extent of reported difficulties and factors

related to them. Potentially, this will contribute to identifying gaps in the research base, offer

suggestions to mitigate distress and opportunities for further research. The current review thus

has the following aims:

1) To examine the prevalence of secondary trauma in legal professionals and the

dominant means of measuring the phenomena.

2) To understand what predictors and associated factors within the legal profession

appear to contribute to secondary trauma, including how those suggested factors

explain levels of variance in developed secondary trauma.

MethodStudy designA study protocol, and its reporting, was informed by PRISMA recommendations (Moher et

al., 2009) and the Centre for Reviews and Dissemination (CRD, 2009). The current review

was not registered with PROSPERO (Booth et al., 2012), but interrogation of the database

revealed no registered reviews on this topic.

16

Search strategyThe search strategy was developed with a senior university librarian. Initial scoping searches

identified no current literature reviews on indirect trauma exposure in the legal professions,

but an examination of grey literature via Google scholar revealed a growing evidence base

regarding indirect traumatisation and occupational distress across professional groups.

A systematic search strategy was undertaken in October 2019, informed by PICOS

(Table 1). Seven databases (PsychINFO, PILOTS, PubMed, Web of Science, Westlaw UK,

Business Source Premier and Criminal Justice Abstracts) were utilised, to ensure coverage of

health research, social sciences and legal subjects. The search terms employed included;

“burnout” OR “compassion fatigue” OR “secondary trauma” OR vicarious trauma” OR

“PTSD”, with combinations of; “solicitor” OR “barrister” OR “lawyer” OR “attorney” OR

“judge” OR “legal”, with all variations and synonyms of the terms. Boolean operation and

truncation methods were carried out in individual databases. A complete list of the search

terms and search strategy is provided in Appendix B.

Selection criteria for studiesTable 1PICOS-statement informing inclusion and exclusion criteria.

Inclusion criteria Exclusion criteriaP-population

Legal professions; barristers,attorneys, solicitors, lawyers andjudges.

Other professions, including lawenforcement and policing.

I-intervention

N/A Aim to measure coping strategies

C-comparison

Including both comparative andnon-comparative studies.

N/A

O-outcome

Prevalence or predictors ofsecondary trauma.

Study aimed to measure directtrauma.

S-study type

Quantitative and mixed methodswith independently discerniblequantitative data.

Qualitative studies, editorials andreviews.

The filters applied to the databases limited results to “peer-reviewed publications” and

“English language”. The author determined study eligibility for inclusion. No date restrictions

were made given no literature reviews were found during initial scoping. All results from the

databases were exported to EndNote X9, where all duplicates were removed automatically, as

well as manually.

17

Eligibility criteria for selected studies were guided by the PICOS method. The

inclusion criteria, were employed during the examination of titles and abstracts, with

reference mining employed to ensure all salient articles were included. Thirty-four articles

met the inclusion criteria and were read in full. After reading, 24 articles were excluded.

18

Figure 1. PRISMA chart of shortlisting process

19

Initial search of PsychINFO, PILOTS, PubMed, Web of science, Westlaw UK,Business Source Premier and Criminal Justice Abstracts databases using

specified search terms and applied filters.

Search results n = 3289PsychINFO = 1396, PILOTS = 458, PubMed = 615, Web of science = 461,

Westlaw UK = 8, Business source premier = 140, Criminal justice abstracts =211

Articles exported to EndNote X9n = 3289

Titles and abstracts screened forinclusion criterian = 2303

Articles read in full and examinedfor specific inclusion criterian = 34

Total articles included for reviewn = 10

Duplicates removed n = 986

Articles excluded n = 2269

Articles excluded n = 24Focus on direct trauma (n=14);Review (n=6);Focus on coping strategies (n=1),Focus on conceptual meanings(n=2) Using qualitative method(n=1).

Data extractionA specially constructed form was developed to extract salient information (Appendix C).

Essential information extracted comprised: author and publication date; aims; sample

demographics; sampling method; study design; measures; analysis (reliability and validity);

results; conclusions; implications; and limitations. To clarify how secondary trauma had been

conceived and applied, the definition adopted in each paper was included as well as

psychometric properties of measures, method of statistical analysis and effect sizes.

Quality appraisalAfter paper selection, articles were assessed for quality using Downes et al. (2016) Appraisal

Tool for Cross-sectional Studies (AXIS). AXIS was developed to address a need for

specificity, generalisability and transferability often absent in other tools appraising cross-

sectional studies. Its development included a Delphi panel, adding to its robustness by

applying expertise and evidence from several different disciplines. Its use in this review is to

assess and compare study quality rather than for exclusion of studies.

The AXIS tool comprises twenty questions (Appendix D) across three domains:

“study design” (seven questions); “reporting” (seven questions); and “introduction of bias”

(six questions). The responses to the questions are categorical (“yes”, “no, “do not know”),

rather than numerically linear, to avoid inter-scale incongruity. Each question was

individually appraised as either present (“yes”) or not present in the study (“no” or “do not

know”). The author created a sum out of 20 items, where ≥16 indicated high-quality studies

(Kiss et al., 2018; Wong et al., 2018).

All studies were quality assessed by two individual and independent raters using the

same appraisal criteria to assess 50 per cent of the studies, which were randomly selected.

Inter-rater reliability showed a strong level of agreement, κ=.84, p<.001.

ResultsElicited studies’ characteristicsTen articles satisfied inclusion criteria and were included in the current review. All papers

were published between 2003 and 2019, were cross-sectional in design and employed survey

methodology. Most studies were solely quantitative, but one study reported mixed-methods.

Five of the studies purposively surveyed their target professional group, and the rest used

comparative between-groups designs. One article reported comparison data on mental health

professionals’ (MHP) and lawyers’ occupational distress, without data on sample size or

characteristics (Levin & Greisberg, 2003). The remaining nine articles reported data from

20

1593 participants, 1454 of whom were legal professionals (the remainder from comparator

professions). Legal professionals represented in the papers comprised lawyers (N=907, 62.4

per cent) and judges (N=547, 37.6 per cent), working in different legal domains: immigration

(Lustig et al., 2008); asylum (Piwowaczyk et al., 2009); criminal/non-criminal law (Vrelevski

& Franklin, 2008). In comparative studies, legal professionals’ distress was variously

compared to other professions, such as social workers, MHPs, psychologists (Levin &

Greisberg, 2003; Maguire & Byrne, 2017), or administrative support staff (Levin et al., 2011).

Across the nine studies, 52.6 per cent were female, and 46.3 per cent were male, which

is consistent with distribution in Australia and Canada (FLSE, 2017; LSNSW, 2016). The

gender distribution in this review is different from reported statistics in the USA (Male: 65 per

cent; ABA, 2018). Age was reported both in mean and in percentages per age group, making

direct comparison challenging. All studies were completed in English speaking countries

(Australia, Canada, and the USA). Only two studies offered data on ethnicity, both reporting

that a majority of the participants were reported as either Caucasian or Anglo-Saxon (Leclerc

et al., 2019; Vrelevski & Franklin, 2008). Sample characteristics are presented in detail in

Appendix E.

Methodological featuresAll studies employed survey design and convenience sampling. None of the studies reported

justification for the chosen sample size, which may reflect the sampling method or the

exploratory nature of some of the studies. Two articles (Jaffe et al., 2003; Schrever et al.,

2019) made use of author-designed questionnaire items. In between-group design studies, the

aim of the comparison group was rarely justified or clarified further, mainly methodological

limitations inherent to comparing two different professions.

All studies measured elements of indirect exposure to trauma, either through validated

measures, or through development of their questionnaire items. The studies operationalised

secondary trauma in various ways: VT (Jaffe et al., 2003; Levin & Greisberg, 2003; Maguire

& Byrne, 2017; Vrelevski & Franklin, 2008); PTSD (Leclerc et al., 2019); STS (Lustig et al.,

2008; Piwowarczyk et al., 2009; Schrever et al., 2019); and CF (Levin et al., 2011; Miller et

al., 2018).

Measures focused predominantly on elements of secondary trauma (VTS, IES-r, MBI-

GS, STSS, LEC-5, CBI, PCL-5, ProQOL) and psychological distress (DASS-21, K-10, CES-

D, IDLS, HSCL-25). Furthermore, studies made use of other specific measures related to

alcohol use (AUDIT), personality (TIPI), functional impairment (SDS) and trauma/attachment

21

beliefs (TABS). Only one study explicitly reported instructing participants to focus on work-

related trauma and indirect trauma exposure from clients (Leclerc et al., 2019), making it

questionable whether reported secondary trauma scores were a result of direct or indirect

exposure. Variability in constructs used, and hence measures, precluded direct comparisons

between studies.

Methodological quality of included studiesThe quality scores of the included studies ranged from five to 17 criteria met, with one article

meeting the threshold for high quality (>16). A summary of the studies’ methodological

quality according to specific AXIS items is provided in Table 2, with the complete AXIS

items in Appendix D.

The “quality of reporting” category (seven items) had the highest criteria met (M=5.6;

range=1-7). Aims, target population and methodology were described across most of the

studies, as were basic data and limitations. However, four studies failed to report statistical

significance or precision estimates appropriately.

The “study design” category (seven items) revealed fewer scores in criteria met (M=4;

range=2-5). Most studies showed appropriate design relevant to the proposed aims, measured

relevant outcome variables and linked their results to the discussion. However, none of the

studies presented power calculations to justify their sample size, and only three studies

reported on conflicts of interest. Three studies failed to report on ethical approval or consent

procedures.

The “risk of bias” category (six items) showed fewest criteria met (M=2.9; range=0-5).

Most studies used previously published measures. Sampling methodology was primarily

informed by convenience sampling, which has inherent limitations and difficulties in

addressing non-response bias.

22

Table 2.Quality Appraisal scores using Appraisal Tool for Cross-Sectional Studies (AXIS; Downes et al., 2016).

Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q20

Author andyear

RQ SD SD RQ SD RB RB SD RB RQ RQ RQ RB RB RB RQ SD RQ SD SDTotal(≤20)

Jaffe et al.(2003) − − − − 10Leclerc et al.(2019) 17Levin &Greisberg(2003)

− − − − − 5

Levin et al.(2011) − − 15Lustig et al.(2008) − − 13Maguire &Byrne (2017) − − 14Miller et al.(2018) − − 14Piwowarczyket al. (2009) − − − − − − 8Schrever etal. (2019) 15Vrelevski &Franklin(2008)

− − 14

Appraisal category: RQ=Reporting Quality; SD=Study Design; RB=Risk of Bias (Kiss et al., 2018)Ratings: =criteria met; =Criteria not met; −=Do not know/partially met.

24

Prevalence of secondary traumaAll but one reviewed study (Miller et al., 2018), which reported perceived general stress,

reported the prevalence of secondary trauma. Prevalence was reported in percentages and

through statistical comparison between occupations. Two studies examining judges revealed

that the majority were experiencing one or more symptoms of secondary trauma (63 per cent,

Jaffe et al., 2003; 83.6%, Schrever et al., 2019). When using PTSD screening tools, 9 per cent

(Leclerc et al., 2019) and 11 per cent (Levin et al., 2011) of lawyers met the criteria for

PTSD. By interpreting STSS scores, Schrever et al. (2019) reported that 30.4 per cent of the

judicial officers qualified for a PTSD diagnosis. Piwowarczyk et al. (2009) found that 9 per

cent of their participants scored higher than 30 on the STSS scale, without stating a cut-off

score or providing inferences based on this score.

In studies comparing legal professionals’ secondary trauma to that in other

occupations, PTSD-symptoms and psychological distress were consistently reported as higher

for legal professions: attorneys compared to social workers and psychologists (Levin &

Greisberg, 2003); lawyers and MHPs (Maguire & Byrne, 2017); lawyers and administrative

support staff (Levin et al., 2011); immigration judges and hospital or prison staff (Lustig et

al., 2008). Additionally, criminal lawyers reported significantly higher secondary trauma

scores compared to non-criminal lawyers (Vrelevski & Franklin, 2008), and judges reported

significantly higher secondary trauma scores compared to barristers (Schrever et al., 2019).

Most studies did not report on effect sizes (Appendix E). However, Maguire and Byrne

(2017) comparing lawyers and MHPs reported a large effect size (d=.90) for vicarious trauma

on the VTS and a medium effect size (r=.40) for impact of trauma on IES-r. Levin et al.

(2011) also found large effect sizes from the PROQOL on burnout (d=.98) and STS (d=.78)

when comparing lawyers and administrator staff.

Consequently, nine of the ten papers report elevated levels of secondary trauma in the

legal profession. Furthermore, all comparative studies reported significantly higher scores in

the legal profession, compared to other professions, though not all studies reported on effect

sizes. Prevalence of secondary trauma was primarily reported as the presence of secondary

trauma symptoms or as meeting criteria for a PTSD diagnosis.

Predictive factorsAs demonstrated in Appendix E, all ten studies collected information regarding diverse

predictors. The author will aim to present examined risk factors and highlight similarities or

differences in key findings. The review considered both associated factors emergent from the

26

studies and results from predictive regression analyses. Factors elicited were: gender; age;

work experience; personal trauma; trauma exposure; and personality traits. In total, only three

studies made use of linear regression analysis examining different variables (Levin et al.,

2011; Maguire & Byrne, 2018; Miller et al., 2011), which limits extensive examination of the

level of variance explained by the predictors.

Gender

Six included articles reported on comparative gender data as an influencing factor of

secondary trauma in the legal profession, revealing equivocal results and no reported effect

sizes. Some reported that females were more likely than males to report STS symptoms (Jaffe

et al., 2003), meet a probable PTSD diagnosis (Leclerc et al., 2019) or experience higher

levels of burnout and stress (Lustig et al., 2008). Two studies reported no effect of gender on

STS (Levin et al., 2011; Maguire & Byrne, 2017). Miller et al. (2018) found that gender was a

mediator between social support and outcome variables (burnout, perceived stress and job

performance), such that high levels of social support related to lower levels of burnout, but

only for males. Furthermore, reduced social support in females showed adverse effects on job

performance.

Age

Three studies examining age as a predictor of secondary trauma reported no significant effect

on reported secondary trauma (Levin et al., 2011; Lustig et al., 2008; Maguire & Bryne,

2017).

Personality traits

In the sole paper examining personality, Maguire and Byrne (2018) examined personality

factors in secondary trauma between professions (lawyers and MHPs). Through hierarchical

multiple regression analysis of secondary trauma, profession accounted for 15 per cent of the

variance. Next, the personality trait Emotional Stability (neuroticism) increased explained

variance by 20 per cent. Four further personality traits (Extroversion, Openness,

Agreeableness, and Conscientiousness) explained variance by a further 18 per cent, and

personal trauma history added 6 per cent to the explained variance. The authors framed their

findings where neuroticism/agreeableness were described as potential risk factors, and

conscientiousness/openness potential resilience factors for secondary trauma.

Direct personal trauma

Three of the studies suggested a personal history of trauma correlated with secondary trauma

symptoms. Maguire and Byrne (2017) found that although direct trauma did not correlate

27

significantly with secondary trauma, it contributed to 6 per cent of the variance. Vrelevski and

Franklin (2008) found that direct trauma impacted on secondary trauma, irrespective of the

level of trauma exposure at work. Leclerc et al. (2019) also reported that prior trauma was a

significant distal factor of distress; women, who were reporting prior trauma and elevated

weekly hours at work, had an increased risk of developing depression and anxiety symptoms.

Work experience/ work hours.

Contribution of work experience showed mixed results, either showing no effect of years of

experience psychological distress (Leclerc et al., 2019; Levin et al., 2011) or conversely that

longer work experience predicted higher levels of STS (Jaffe et al., 2003) and burnout (Miller

et al., 2018). Leclerc et al. (2019), whose study suggested a cumulative effect of trauma

exposure, noted increased weekly hours of work heightened the risk of developing PTSD

symptoms by 1.89 times. Levin et al. (2011) also reported that weekly work hours acted as a

significant mediator in developing STS symptoms, when comparing attorneys and

administrative support staff.

Levin et al. (2011) examined the explained variance of work hours and trauma

exposure with secondary trauma between professions (attorneys and administrative staff). A

direct association model showed that attorneys had higher high levels of PTSD symptoms,

and that profession explained 7 per cent of the variance. When work-related exposure was

added, the model explained 14 per cent of the variance. Next, they found that attorneys were

significantly associated with high levels of STS and burnout, which explained 20 per cent of

the variance. When including work hours and trauma-exposure as mediators, the model

explained 32 per cent of the variance.

Level of indirect trauma exposure

Two studies reported within-profession comparisons to control for level of trauma exposure.

Leclerc et al. (2019) collected information on the percentage of respondents’ caseload

containing clients experiencing trauma in the past year, finding the level of trauma exposure

increased the risk of developing PTSD, higher levels of psychological distress and lower

quality of life. Effect sizes between levels of trauma exposure were: no exposure to moderate

exposure (medium; d=.40); no exposure and high exposure (large; d=.70); and moderate and

high exposure (small to medium).

Vrelevski and Franklin’s (2008) comparison of criminal and non-criminal lawyers

reported that criminal lawyers (arguably exposed to more traumatogenic material) were more

likely to develop secondary trauma symptoms and psychological distress, than non-criminal

28

peers; they also disclosed more difficulties relating to self-safety, other-safety and other-

intimacy attachment beliefs.

DiscussionThe current study is the first systematic review to examine the prevalence and measurements

of secondary trauma in legal professionals and associative and predictive factors of secondary

trauma. The review aimed to understand the extent of any difficulties and what factors might

be amenable for mitigation or intervention. After a systematic process of identifying,

interrogating and appraising available research findings, ten articles examining legal

professionals were chosen for review.

Prevalence and measurementAppraisal and synthesis of the ten elicited studies all revealed elevated reported levels of

secondary trauma. The prevalence rate of one or more symptom of secondary trauma ranged

from 63 per cent to 83.6 per cent. A higher percentage of the studies populations met the

DSM criteria for a PTSD diagnosis, ranging from 9 per cent to 30.4 per cent. The prevalence

was significantly higher than reported numbers from The World Health Organisation world

mental health survey (Koenen et al., 2017) reporting cross-national lifetime prevalence of

PTSD at 3.9 per cent. As some studies did not indicate if they instructed participants to only

consider secondary trauma when completing questionnaires, it is unclear whether reported

secondary trauma scores were a result of direct or indirect exposure, or a combination of the

two. This is further complicated by the studies’ chosen prevalence score, as PTSD levels

could potentially include participant’s own past trauma experiences.

Comparative studies also revealed that legal professionals reported significantly higher

levels of secondary trauma compared to MHPs, social workers, psychologists, administrative,

hospital and prison staff. Comparisons were also made within the legal professions, where

criminal lawyers had higher levels of secondary trauma compared to non-criminal lawyers,

and judges were found to have higher levels compared to barristers. The two studies reported

high effect sizes when comparing lawyers with MHPs (Maguire & Byrne, 2017) or with

administrative staff (Levin et al., 2011). Despite these findings, the results should be

interpreted with caution due to the poor scientific methodology in most of the chosen studies.

The disparate studies reflected the conceptual confusion in the area of secondary

trauma, and the lack of reported effect sizes and prevalence rates across key findings made it

more challenging to draw conclusion. As such, prevalence was more distinctly reported in this

review as meeting a diagnosis of PTSD, rather than other typical secondary trauma terms.

29

Moreover, the studies used a mixture of non-validated explorative questions and validated

instruments on trauma (IESr, PCL-r, LEC), burnout (MBI, CBI), secondary trauma (VTS,

STSS, ProQOL) and general psychological distress.

Predictors and associated factorsAcross the ten studies, only two consistent findings were found; age was unrelated to

secondary trauma, and levels of trauma exposure consistently correlated with secondary

trauma levels. The results from the remaining proposed variables were inconsistently found

across studies, notably gender, work-experience, history of personal trauma and personality

traits. These findings were inconsistent with results presented in a review on secondary

trauma in first responders (Greinacher et al., 2019), which found effects of age, gender and

previous trauma. Mediators reported in the studies, which contributed to the variance of

secondary trauma, included work hours, trauma exposure, personality traits and personal

trauma history. Again, these findings should be interpreted with caution.

The current review only reported on factors presented in the ten articles. This does not

exclude the possibility of several other potential factors, which were not presented in the

current review. Interestingly, organisational information was not collected in any of the ten

articles. The lack of consideration around organisational contexts could promote blame on

individuals rather than acknowledging how the organisation could play a part in occupational

distress. One could speculate that an individualistic culture could limit resilience-building

schemes, which are protective of secondary trauma – such as peer support or supervision

(James, 2020).

Methodological limitationsAnalysis of the chosen articles for review revealed multiple methodological flaws and

considerable variation in quality and scientific rigour. All studies made use of self-report

measures and convenience sampling, which elicits several limitations such as social-

desirability bias, confirmation bias and selection bias. Moreover, as several studies were

exploratory, they made use of questionnaire items that had not been validated. It is hard to

assume that the studies were measuring what they intended, especially considering the current

conceptual difficulties. It is unclear if participants were asked to report specifically on work

related indirect trauma, and as such the PTSD measures would likely have included elements

of direct and indirect exposure to traumatic content.

In the studies that made use of control groups, the substantial differences in occupation

could have contributed to numerous confounding variables, and the large reported effect sizes.

30

Furthermore, when comparing the effects of reported trauma exposure with secondary trauma

measures, one could argue that they measure very similar constructs on which the measures

were based. In that regard, it is unclear whether the results in those cases depict significant

effects or if it is merely confirming construct validity in the secondary trauma measures. Most

of the studies did not adequately control for trauma exposure to indicate a valid effect of

secondary trauma.

Consistent results relating to the high prevalence of secondary trauma could indicate a

presence of publication bias, specifically as all presented articles were published in peer-

reviewed journal articles. Among the risk factors, the only observed consistent results were

age and trauma exposure. There seems to be a high need for further examination of the

inconsistent factors: gender; work experience; and personal trauma history.

Given the numerous methodological limitations and variability in quality appraisal

scores, the results should be interpreted with caution. Overall, interpretations of the findings

were challenging due to the methodological and conceptual limitations inherent in the

literature. Nevertheless, the current review of the chosen articles contributed to some

noteworthy findings that might prove useful in informing future research.

ImplicationsDespite the relatively few studies, the disparate measurement and some equivocal findings,

the results from this review signify the importance of understanding secondary trauma in

occupations other than health and social work. Whilst the findings from this review should be

interpreted with caution, they could have implications for diagnostic manuals that currently

focus on first responders and health professions. Interestingly, by opening the diagnostic

criteria to other professions, their occupational distress could be taken seriously in academia

and medical environments. However, implementing professions within the PTSD-diagnosis

might also contribute to a risk of pathologising normal responses to adverse traumatogenic

exposure.

Though the legal profession might not be traditionally viewed as high risk of trauma

exposure, the results suggest that is not the case. Addressing elevated secondary trauma levels

are not only essential to promote professional wellbeing and sustainability of their careers, but

also given their need to argue for delivery of best evidence, judicial reviews and enacting the

law. As health-professionals receive supervision and trauma-training, they might be better

equipped in dealing with indirect trauma-exposure. Findings from the current review could

emphasise the need for training and support available for legal professionals with high trauma

31

exposure through their work. However, the distinct individualistic focus within the studies

prohibits any examination of the development of resilience, or perceived organisational

support. Perhaps the recognised absence of organisational data could have implications for

reviewing legal professionals’ individualistic culture.

Strengths and limitationsThe current study is the first systematic review to examine secondary trauma prevalence,

predictors and associated factors in the legal profession. The inclusion criteria allowed for a

range of legal professions to be included. However, despite this breadth, the search only

elicited a small number of articles fitting the full inclusion criteria. This indicates that the

discipline is still in its infancy. The current review emphasises a great need to improve

research quality and quantity within this area of study. Additionally, this review usefully

highlights that professions other than health and social work might be affected by indirect

trauma exposure.

However, the present review is also subject to several limitations. Given the

considerable variability in conceptual understanding and methodology, this inhibited the

opportunity for adequate comparison of effect sizes. Although the review has contributed to

an understanding of the research aims, they may not have been addressed in a way that

provided a definite answer. Moreover, this study solely focused on quantitative research,

where qualitative sources could have provided a richer comprehension of the experience and

meaning-making with the presented concept and population.

This review focused solely on the psychological construct of secondary trauma. It is

important to note that the legal profession might experience other elements of occupational

and psychological distress. Secondary trauma is merely one possible difficulty within a

greater context.

Future researchAs this is a relatively understudied area, the possibilities for future research are enormous.

Future research would benefit from greater consensus on what constitutes professional

distress, where conceptual clarification would have an impact beyond exploration of legal

professions. As it follows, psychometric instruments would also benefit from comparative

analyses as well as conceptual clarification and target groups.

Clear articles with conventional research structures could contribute to academic

professionals giving greater attention to this vital subject. Research would also benefit from

the development of more scientifically robust methods, especially in consideration of

32

vocational control groups. Addressing the aforementioned methodological weaknesses could

contribute to more scientifically precise findings. Partly, this would include diversifying

sampling methodology and addressing non-responders. Moreover, the sources of participants

are only recruited in Australia and Northern America, in countries subject to an adversarial

jurisdictional system. Potentially examining European inquisitorial systems would elucidate

the topic further.

The ten selected articles comprised several professions with differences that could

benefit from further consideration of the topic of secondary trauma. Lawyers and judges in

criminal, asylum and immigration law might be more exposed to traumatised individuals

compared to other areas of the law. Also, it would be useful to examine how judges’

responsibilities of making rulings and decisions, seemingly without the influence of emotions,

relationships or other factors, are potentially influenced by moral distress and secondary

trauma.

Given an over-emphasis on potential predictive and associated factors located in

individual vulnerability, future developments could focus on factors underpinning resilience

as well as deficits, organisational and training contexts, which feature in understanding

professional distress in other professions (Greinacher et al., 2019; MacEachern et al., 2011).

Within this, there are several possibilities of experimental designs examining the

implementation of organisational changes, peer support or coping strategies.

ConclusionsThis review is the first to provide collated information from a systematic literature search

regarding secondary trauma prevalence, predictors and associated factors in the legal

profession. The results indicated a high prevalence of secondary trauma within this

population. It also emphasised the importance of further exploration regarding predictors of

secondary trauma and suggested improvements for future research. As such, this article

highlights the potential risk of indirect exposure to trauma in legal professions. However,

considering the small number of articles and poor quality of research the findings are subject

to great scrutiny and should be interpreted with caution. Therefore, this review emphasises the

need for more scientifically robust research on this topic.

33

ReferencesReviewed articles are marked with an asterisk (*)

ABA, American Bar Association (2018). A current glance at women in the law, Chicago:

American Bar Association. Available via

https://www.nyipla.org/images/nyipla/Programs/2018December6/

acurrentglanceatwomeninthelawjan2018.pdf Retrieved 23 March 2020

American Psychiatric Association (2013). Diagnostic and statistical manual of mental

disorders (DSM-5®). American Psychiatric Pub.

Baird & Kracen (2006). Vicarious traumatization and secondary traumatic stress: A research

synthesis. Counselling Psychology Quarterly, 19(2), 181-188

Booth, A., Clarke, M., Dooley, G., et al. (2012). The nuts and bolts of PROSPERO: an

international prospective register of systematic reviews. Systematic

Reviews, 1(1), 2.

Canfield, J. (2005). Secondary traumatization, burnout, and vicarious traumatization: A

review of the literature as it relates to therapists who treat trauma. Smith College

Studies in Social Work, 75(2), 81-101.

Collins, S., & Long, A. (2003). Working with the psychological effects of trauma:

consequences for mental health‐care workers–a literature review. Journal of

Psychiatric and Mental Health Nursing, 10(4), 417-424.

Downes, M. J., Brennan, M. L., Williams, H. C. & Dean, R. S. (2016). Development of a

critical appraisal tool to assess the quality of cross-sectional studies (axis). BMJ

Open, 6(12). 10.1136/bmjopen-2016-011458

Elwood, L. S., Mott, J., Lohr, J. M., & Galovski, T. E. (2011). Secondary traumasymptoms in

clinicians: A critical review of the construct, specificity, andbimplications for

trauma focused treatment. Clinical Psychology Review, 31(1), 25-36.

Figley (1995) Compassion Fatigue: Coping With Secondary Traumatic Stress Disorder In

Those Who Treat The Traumatized. Routledge, New York NY

FLSC; Federation of Law Societies of Canada (2017). Membership 2017 Statistical

report of the Federation of Law Societies of Canada. Ottawa: FLSC Available via

https://flsc.ca/wp-content/uploads/2019/04/2017-Stats-Report.pdf Retrieved 23 March

2020

Greinacher, A., Derezza-Greeven, C., Herzog, W., & Nikendei, C. (2019). Secondary

traumatization in first responders: a systematic review. European Journal of

Psychotraumatology, 10(1), 1562840.

34

*Jaffe, P., Crooks, C., Dunford-Jackson, B., & Town, J. (2003). Vicarious trauma in

judges:The personal challenge of dispensing justice. Juvenile & Family Court

Journal, 54(4), 1-9.

James, C. (2020). Towards trauma-informed legal practice: a review. Psychiatry,Psychology

and Law, 1-25.

Kiss, R., Schedler, S. & Muehlbauer, T. (2018). Associations between types of balance

performance in healthy individuals across the lifespan: A systematic review and meta-

analysis. Frontiers in Physiology, 9, (1366-1366). 10.3389/fphys.2018.01366

Koenen, K. C., Ratanatharathorn, A., Ng, L., et al. (2017). Posttraumatic stress disorder in the

world mental health surveys. Psychological Medicine, 47(13), 2260-2274.

Krill, P. R., Johnson, R., & Albert, L. (2016). The prevalence of substance use and other

mental health concerns among American attorneys. Journal of Addiction

Medicine, 10(1), 46.

*Leclerc, M. E., Wemmers, J. A., & Brunet, A. (2019). The unseen cost of justice: post-

traumatic stress symptoms in Canadian lawyers. Psychology, Crime & Law,

21(1).

*Levin, A. & Greisberg, S. (2003) Vicarious trauma in attorneys. Pace Law

Review, 24(11), 245-252.

Lonergan, M., Leclerc, M. È., Descamps, et al. (2016). Prevalence and severity of

trauma and stressor-related symptoms among jurors: A review. Journal of

Criminal Justice, 47, 51-61.

*Lustig, S., Delucchi, K., Tennakoon, L. et al. (2008). Burnout and stress among United

States immigration judges. Benders Immigration Bulletin, 13, 22-30.

MacEachern, A. D., Jindal-Snape, D., & Jackson, S. (2011). Child abuse investigation:

police officers and secondary traumatic stress. International Journal of

Occupational Safety and Ergonomics, 17(4), 329-339.

*Maguire, G., & Byrne, M. K. (2017). The law is not as blind as it seems: Relative rates of

vicarious trauma among lawyers and mental health professionals. Psychiatry,

Psychology and Law, 24(2), 233-243.

Maslach, C., & Jackson, S. E. (1981). The measurement of experienced burnout. Journal of

Organizational Behavior, 2(2), 99-113.

Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout Inventory. 3rd ed. Palo

Alto, CA: Consulting Psychologists Press.

35

McFadden, P., Campbell, A., & Taylor, B. (2014). Resilience and burnout in child

protection social work: Individual and organisational themes from a systematic

literature review. The British Journal of Social Work, 45(5), 1546-1563.

Miller, M. K., & Bornstein, B. H. (2004). Juror stress: Causes and interventions. T.

Marshall L. Rev., 30, 237.

*Miller, M. K., Reichert, J., Bornstein, B. H., & Shulman, G. (2018). Judicial stress: the roles

of gender and social support. Psychiatry, Psychology and Law, 25(4), 602- 618.

Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for

systematic reviews and meta-analyses: the PRISMA statement. Annals of Internal

Medicine, 151(4), 264-269

Najjar, N., Davis, L W., Beck-Coon, K. & Doebbeling, C.C. (2009). Compassion

fatigue: A Review of the Research to Date and Relevance to Cancer-CareProviders.

Journal of Health Psychology, 14(2), 267-277

Nolte, A. G., Downing, C., Temane, A., & Hastings‐Tolsma, M. (2017). Compassion

fatigue in nurses: A metasynthesis. Journal of Clinical Nursing, 26(23-24), 4364-4378.

LSNSW; the Law Society of New South Wales (2016) NSW Profile of solicitors 2016 Final

report. Sydney: Urbis Pty Ltd Available via: https://www.lawsociety.com.au/sites

/default/files/201804/ NSW%20PROFILE%20OF%20SOLICITORS%202016%

20FINAL%20REPORT.pdf Retrieved 23 March 2020

Parker, C. (2014). The Moral Panic over Psychological Wellbeing in the Legal Profession: A

Personal or Political Ethical Response. UNSW Law Journal, 37, 1103.

Pearlman, L. A., & Saakvitne, K. (1995). Trauma and the Therapist: Countertransference

and vicarious traumatization in psychotherapy with incest survivors. New York: W.W.

Norton

*Piwowarczyk, L., Ignatius, S., Crosby, S. et al. (2009). Secondary trauma in asylum

lawyers. Benders Immigration Bulletin, 14(5), 1-9.

Sabo, B. M. (2006). Compassion fatigue and nursing work: can we accurately capture

The consequences of caring work? International Journal of Nursing Practice.

12(3), 136-142.

Santos, C. M. D. C., Pimenta, C. A. D. M., & Nobre, M. R. C. (2007). The PICOstrategy for

the research question construction and evidence search. Revista Latino-Americana de

Enfermagem, 15(3), 508-511.

36

*Schrever, D., Sourdin, T. & Hulbert, C.A. (2019). The psychological impact of Judicial

Work: Australia’s first empirical research measuring judicial stress and wellbeing.

Journal of Judicial Administration, 28(3)

Stamm, B. H. (2010). The concise ProQOL manual, 2nd ed. Pocatello, ID: ProQOL.org.

*Vrklevski, L. P., & Franklin, J. (2008). Vicarious trauma: The impact on solicitors of

exposure to traumatic material. Traumatology, 14(1), 106-118.

Wong, J. N., McAuley, E. & Trinh, L. (2018). Physical activity programming and counselling

preferences among cancer survivors: A systematic review. International Journal of

Behavioral Nutrition and Physical Activity, 15(1), 48. 10.1186/s12966-018-0680-6

Yang, Y. H. & Kim, J. K. (2012). A literature review of compassion fatigue in

nursing. Korean Journal of Adult Nursing, 24(1), 38-51.

37

Section Two: Main research report

Empathy, resilience and distress in Norwegian and English investigative interviewers.

Target journal: Investigative Interviewing: Research and Practice (II-RP)

See Appendix F for author guidelines

Presented here using BPS Style guidelines.

38

AbstractBackground: Police officers working in investigative interviewing can be exposed to

traumatogenic material through their work. Indirect trauma through occupational exposure is

presented here as compassion fatigue, which proposes adverse effects through empathetic

engagement with traumatised individuals. The current study aims to examine associated

factors, predictors and levels of compassion fatigue and satisfaction. The study also explored

the moderating/mediating relationship of empathy between suggested variables and

compassion fatigue.

Methods: The study used a cross-sectional survey-design of UK and Norwegian police forces.

Results: Levels of distress were reported as low/moderate in the both the UK and Norwegian

participant group, but were also comparatively higher than reported community norms.

Compassion fatigue correlated with measures of psychological distress but did not show a

relationship with empathy.

Discussion: Empathy was not found to be a mediating factor in compassion fatigue, and

further research is required to understand the role of empathy, factors influencing wellbeing,

and strategies to counter distress. Findings suggest that supportive organisational strategies

should be offered to police interviewers, in order to maintain resilience to sustain them in key

roles to aid the vulnerable and secure prosecutions.

39

IntroductionOver the last two decades, there has been an increasing awareness of the potentially adverse

impacts of working with vulnerable clients in social and healthcare professions. This has been

variously described under the rubric of “the cost of caring” (Figley, 1995), and is argued to

occur as a result of listening to another person’s traumatic experiences. The effects of such

indirect exposure to trauma appears to be related to emotional engagement with affect-laden

content and client distress.

Certain professions routinely exposed to others’ trauma and suffering have been

identified as particularly vulnerable to occupational distress, notably nurses (Yang & Kim,

2012), mental health professionals (Collins & Long, 2003), therapists (Canfield, 2005) and

social workers (McFadden et al., 2014). Staff in non-therapeutic professions have been

relatively understudied. However, more recently, those undertaking policing roles have been

identified as a similarly vulnerable professional group (Andersen & Papazoglou, 2015).

Whilst police themselves would accept that stress is an inevitable part of the work for which

they are responsible, and that the distress they experience may be manageable and indeed

forge resilience (Kronenberg et al., 2008), it has been increasingly recognised that front line

police officers can be adversely by their work (MacEachern et al., 2010).

Police officers exercising the law through diverse duties are frequently exposed to

traumatogenic events, material and discourse; they may experience and be witness to actual,

or consequences of violent crime, exploitation of the vulnerable, and fatalities (Cross &

Ashley, 2004; Karlsson & Christianson, 2003). They may also experience moral injury if a

lack of resources in the wider judicial system means they are unable to affect legal processes

(Papazoglou et al., 2019).

Akin to therapeutic engagement with traumatised clients, building rapport is an

essential yet often unexamined element of police work, which might engender indirect

trauma. Through the process of investigation, police are regularly and repeatedly tasked with

collecting information from distressed and traumatised victims and witnesses, through

specific methods known as investigative interviewing.

Investigative interviewingInvestigative interviewing is a core duty for police officers in the UK and in other European

countries who investigate alleged crime – in roles such as ‘detective’ and ‘specialist

interviewer’. Investigative interviewing is arguably the most crucial aspect of any criminal

investigation as it seeks to elicit information from witnesses, victims, and suspects that is

40

essential to successful case resolution (Milne & Bull, 1999; Lassiter, 2004). The purpose of

these interviews is to provide an ethical, research-based and non-coercive method of

questioning witnesses and suspects. Through specific methods, such as cognitive

interviewing, the investigators aim to maximise relevant and detailed information, while

minimising bias and contamination of evidence (Bull, 2014; College of Policing, 2019). This

is achieved through adhering to specific protocols including: planning and preparation;

engaging and explaining; account clarification and challenging; closure; and evaluation.

These elements feature in the PEACE model, used by UK investigative interviewers, and

adopted by the other countries’ police forces (McGurk et al., 1993; Milne & Bull, 1999).

Notably, the Norwegian Police University College (NPUC) invited English police staff and

researchers to share their expertise on the PEACE-model, which is presently being utilised

across Norwegian police districts under the acronym KREATIV (Bruusgard et al., 2013).

Critical to extracting detailed evidence through the interview is the need to build

rapport and engage empathically with an interviewee. The purpose of investigative

interviewing is not to mitigate trauma necessarily, but to aim for accuracy and reliability.

Specific guidance is provided for sensitivity training when interviewing witnesses and

suspects. Additional interview training is required for interviewing vulnerable adults or

children, with an emphasis on engagement with the interviewee and eliciting information

(College of Policing, 2019). However, these guidelines do not appear to consider interviewer

needs, with little attention given to the potential effects of engagement with distressed

interviewees and distressing testimony through the interview process.

Conceptual framing of occupational distressUnderstanding how investigative interviewing may potentiate indirect trauma requires

consideration of the constructs which currently describe such phenomena; variously termed as

Post-traumatic Stress Disorder (PTSD), Burnout, Vicarious Trauma, Secondary Traumatic

Stress (STS) and Compassion Fatigue (CF). Najjar et al. (2009) stated that these terminologies

have significant similarities, emerging through empathic emotional engagement with clients,

though there are also noteworthy differences. Due to the difficulties defining and measuring

the different concepts adequately, the literature can be inconsistent in application of the terms

(Elwood et al., 2011). Currently, CF seems to be the most utilised and appropriate concept in

research into the impact of working in law enforcement as a result of exposure to others’

distress.

41

CF, the most recently developed term, describes the potential for experiencing trauma

through learning about the traumatising experiences of others and manifests as a form of

emotional and physical exhaustion (Figley, 1995). The term was used interchangeably with

STS, until it was redefined to include both STS and elements of burnout (Stamm, 2010),

making CF the most inclusive of the different concepts. STS itself encompasses PTSD like

symptoms (Figley, 1995) including avoidance, arousal and intrusion, as well as fatigue,

physical pain, withdrawal and disturbed sleep (Nolte et al., 2017). Burnout, a form of

occupational exhaustion from sustained intense emotional engagement (Maslach & Jackson,

1981; Maslach et al., 1996), presents through exhaustion, depersonalisation, reduced personal

accomplishment and cynicism. Within Stamm’s model, it follows that indirect exposure to

traumatic content might lead to physical and emotional exhaustion and reduced empathetic

capacity (Elwood et al., 2011; Evces, 2015).

ResilienceAlthough police interviewers remain in their post for sustained periods (Crawford et al.,

2017), their responses to their work also appear to be positive. Constructs such as compassion

satisfaction (CS; Stamm, 2002), posttraumatic growth (Arnold et al., 2005) or resilience

(Hernández et al., 2007) may be helpful to understand their experiences. CS (Stamm, 2002)

describes the gaining of pleasure or satisfaction accumulated through working in helping

professions. As a model, CF-CS provides a way of conceptualising occupational distress and

resilience together.

Resilience in policing is relatively underexplored, but both individual and

organisational measures have been proposed to build resistance. Supportive occupational

provision such as supervision, peer support and counselling (Figley, 1995) are argued to

mitigate the combination of high psychological demands, low control in decision-making and

limited social support that are associated with exhaustion (Evans et al., 2006).

Empathy and distressInvestigative interviewing requires professional skills which, along with empathetic

sensitivity, can ease and facilitate an interview. The capacity to understand or feel another’s

experience (Goubert et al., 2005) is needed to navigate a witness’ emotional presentation and

build an effective, trusting relationship (Mathieu, 2007; Risan et al., 2016). This is

challenging for interviewers in a culture where outward resilience is cultivated, emotional

responses are often diminished and caseloads are high (Hanway & Akehurst, 2018); little is

known about the impact of sustained empathic engagement in this role.

42

Empathy appears central in regulating adverse indirect effects of trauma exposure

(Figley, 1995); however, its precise role is unclear. Empathic concern has predicted elevated

CF in oncology nurses (Duarte et al., 2016), which is consistent with Figley’s suggestion that

sensibility to others’ suffering may be a vulnerability in the development of CF. Hunt et al.’s

(2017) review of burnout, STS and empathy in nursing suggested that the relationship

between these constructs are complicated, and an ability to self-regulate emotions during

empathic engagement may reduce the risk of burnout. In only one published study of police

officers, specialist officers working with those alleging rape (Turgoose et al., 2017) revealed

no relationship between empathy and CF or STS, but that trait empathy predicted elevated

burnout. The role of empathy appears to warrant further investigation. There is recognised

need to examine the effects of investigative interviewing on the interviewer as well as

examining the levels of support police officers have available to them (Hanway & Akehurst,

2018).

AimsGiven the nature of the work undertaken by investigative interviewers and likely exposure to

traumatogenic material, this study aimed to describe emotional responses to work (CF-CS) for

police officers undertaking this role. Since investigative interviewing principles are generally

agreed upon internationally, but implementing them has proven to be problematic, it was

decided that officers’ responses would be examined in two nations (the UK and Norway) in

which investigative interviewers adopt similar roles and principles in interviewing. Given

known predictors of CF, and an aim to examine the role of empathic engagement in

investigative interviewing, this study hoped to investigate the relationship between empathy,

distress and resilience within these populations alongside demographic factors associated with

occupational distress in other professional groups. This report aimed to provide a particular

focus on the mediating/moderating role of empathy. Since the developed terminology on

potentially distressing facets of work are variously defined, this study used the CF-CS model,

which includes STS, burnout and CS (Stamm, 2010), to capture the phenomenon in

investigative interviewers, as a term of occupational distress and resilience.

Research questions:

1. What are the levels of distress and resilience (CF/CS) reported by investigative

interviewers in Norway and the UK?

2. What factors are associated with reported distress and resilience?

3. To what extent do associated factors predict reported distress and resilience?

43

4. Does reported empathy moderate/mediate the relationship with CS/CF and proposed

predictor variables in investigative interviewers?

MethodResearch designThe current study utilised a quantitative cross-sectional survey design to examine reported

levels of distress and resilience in investigative interviewers from police forces in the UK and

Norway. Furthermore, the study examined factors and predictors of CS and CF, as well the

level to which empathy was related to the variables. The specific dependent variables

comprised CS, STS and burnout, with the independent variables being gender, age, tenure in

post, empathy, psychological distress (anxiety and depression) and perceived organisational

support (POS).

Ethical approvalsThe University of Leicester Ethics Sub-committee of Psychology granted ethical approval in

May 2019, following both peer and service user reviews (Appendix G). In Norway, ethics

approval was gained from the Regional Committees for Medical and Health Research Ethics

(REC) in September 2019 (Appendix H), and the National Police Directorate (subordinate to

the Ministry of Justice and Public Security).

ParticipantsInclusion criteria for participants were similar for recruitment in the UK and Norway. All

participants were currently employed in the public service as detectives or specialist

interviewers, undertaking the role of investigative interviewing. The roles encompassed

similar tasks in both countries, with specialist interviewers needing additional training for

interviewing vulnerable populations. Having specialist training, or additional interview

training did not exclude participants. Through their respective Colleges of Policing, both UK

and Norwegian participants had received training on completing witness assessment during

investigative interviewing, and cognitive interviewing techniques. Recruitment took place

from police constabularies of varying sizes, and from districts in various geographical areas in

UK and Norway.

RecruitmentThe method of recruitment was convenience sampling, over a period of six months (October

2019 to April 2020). All recruitment information specified that participants should be working

in the role of ‘investigative interviewer’. In the UK, police forces of differing sizes who

44

employed investigative interviewers were approached via gatekeepers in national

investigative resourcing who publicised access to an online questionnaire in the UK. In

Norway, recruitment was coordinated with a researcher from the Norwegian University of

Science and Technology (NTNU) and the STrESOB project, who advertised the project

during a national police conference and distributed printed paper questionnaires to the police

districts who agreed to participate. The STrESOB project involves Norwegian national

research aimed at preventing CF in detectives and specialist interviewers who work with child

sexual abuse.

ProcedureIn the UK, gatekeepers distributed e-mail invitations to complete an online questionnaire. The

e-mail included a participant information sheet (Appendix I), a questionnaire link and

password to access the questionnaire. Whilst the researcher had no direct contact with

participants, all potential participants were provided with the principal investigator’s contact

details should they have additional questions or wanted a summary of findings. Consent was

collected within the online questionnaire, in addition to the assumed consent from

questionnaire completion.

In Norway, questionnaires were sent to districts who had agreed to participate in the

study. Each printed questionnaire also provided stamped and addressed envelopes ready to be

sent to NTNU, participant information sheets and consent forms. The associate Norwegian

researcher had no direct contact with participants.

MeasuresThe questionnaire was prefaced by a privacy/consent declaration (Appendix J). To capture

key variables of interest the questionnaire comprised questions relating to demographic

variables (gender, age, ethnicity, work tenure and level of interview training). Four validated

tools were included (Appendix K) to assess psychological wellbeing, empathy, POS, and CF-

CS. These were selected because of suitability to the research questions, their statistical

properties and availability in both English and Norwegian.

CF-CS

The Professional Quality of Life Compassion and Satisfaction Subscale v5 (ProQOL; Stamm,

2010) is a 30-item questionnaire assessing CS, STS and burnout, and has previously been

used with police officers (Greinacher et al., 2019). It employs a five-point Likert scale

measuring frequency from “never” to “very often”, creating scores of 1-5 for each of ten

questions relating to the assessed variables. Each item which included the identifier as

45

“helper” was changed to consider the investigative interviewing role, fitting more with the

identification of the participant group. Scores are calculated individually for each domain,

with higher scores suggesting higher levels of CS, STS or burnout. Respondents’ scores were

categorised via the self-score methods offered within the ProQOL manual, such that STS,

burnout and CS were categorised as low, average or high.

Stamm (2010) argues the measure shows good construct validity given the number of

published articles utilising the measure. Moreover, within the CF scale, inter-scale

correlations indicated two per cent shared variance (r=-.23) with STS and five per cent shared

variance (r=-.14) with burnout. In 2005, Internal reliability appears high; Cronbach’s α in CS

(α=.87), burnout (α=.72) and STS (α=.80; Stamm, 2010). Watts and Robertson (2015)

suggested supporting the measure with reliable and valid psychometrics, such as the Hospital

Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983).

Psychological wellbeing

The Hospital Anxiety and Depression Scale (HADS; Zigmond & Snaith, 1983) is a 14-item

questionnaire assessing anxiety and depression reported over the preceding seven days. Both

dimensions are assessed through seven questions (14 in total) via a Likert scale scored 0-3.

Higher scores indicate greater levels of distress, with a maximum score of 21 for each

dimension. A dimension score of 11 or above is considered ‘abnormal’ scores according to

the authors, and indicated clinically elevated depression or anxiety, also called clinical

caseness. Bjelland et al. (2002) demonstrated good construct and convergent validity, and

concluded that the HADS performed well when assessing severity and caseness for both

anxiety and depression in the general population. The scores do not automatically infer an

official diagnosis when it is considered without a psychiatric opinion (Silverstone, 1994).

A Norwegian validation review (Leiknes et al., 2016) reported high internal validity

(α>.70). Construct validity was shown in several factor analysis studies, most of them

supporting the two categories. Convergent validity was demonstrated when compared with

the Montgomery-Aasberg Depression Rating Scale, correlating at moderate to high level on

emotional content items. Correlation between GHQ and HADS-D was moderate (Cohen’s

k=0.37, Pearson’s r=0.69, p>.001) and HADS-A was moderate to good (Cohen’s k=0.41,

Pearson’s r=0.69). HADS was chosen as a mental health questionnaire that showed great

statistical properties in both languages and has been recommended by Watts and Robertson

(2010) to support the ProQOL measure.

Empathy

46

The Cambridge Behaviour Scale – Empathy Quotient (EQ; Baron-Cohen & Wheelwright,

2004) short version is a 40-item self-report questionnaire containing items assessing both

cognitive and emotional components of empathy. The questionnaire offers four potential

responses to items (two of which are scored as 0) calculating scores from a range of 0-2, with

higher scores indicating higher levels of empathy. Scores of 0-32 indicate lower than average

empathy, 33-52 average, 53-63 above average, and 64-80 very high empathy.

The EQ has demonstrated face validity (Baron-Cohen & Wheelwright, 2004) and

concurrent validity between the subscales “empathetic concern” and “perspective-taking”

within the Interpersonal Reactivity Index, and they also showed good psychometric properties

of internal and test–re-test reliability (Lawrence et al., 2004).

Perceived Organisational Support

The Survey of Perceived Organisational Support (SPOS; Eisenberger et al., 1986;

Eisenberger, 2019) is a 36-item questionnaire assessing POS. It employs a seven-point Likert

scale capturing agreement with statements (strongly disagree 0 to strongly agree 6). Higher

scores indicate higher POS. Sixteen items, eight-item and three-item scales are also available

for a shorter form versions. The survey is unidimensional, examining elements of POS. The

present survey used the eight-item scale.

Shore and Tetrick (1991) demonstrated the construct validity of the SPOS, showing

that the survey was distinguishable from affective and continuance organisational

commitment. The results also supported the claim of the SPOS as a unidimensional scale. A

reliability analysis indicated good internal consistency (α=0.81) and item-total correlations

ranging from 0.64-0.67 (Worley et al., 2009).

Data analysisIBM/SPSSS Statistics version 26 was used for statistical analyses, analysis of missing data

(Appendix L) and evaluation of assumptions for each of the analyses (Appendix M and N).

To examine levels of distress and resilience between the Norwegian and the UK sample, five

independent t-tests were completed to compare means between groups for the CS, STS,

burnout, empathy, and POS. Following evaluation of assumptions, burnout and CS failed to

meet assumptions, and therefore non-parametric Mann-Whitney U analyses were conducted.

Given the multiple analyses undertaken, p-values were set and interpreted more stringently at

p<.01 (Perneger, 1998).

Bootstrapped single sample t-tests were conducted to compare sample means with

population norm data. Normative studies were used for the ProQOL (De La Rosa et al., 2018)

47

and the HADS for English (Crawford et al., 2001) and Norwegian (Leiknes et al., 2016)

population data. To examine the differences in caseness for anxiety and depression between

the Norwegian and UK participant group, two chi-square analyses were performed. Caseness

was indicated by number of participants who scored over 11 for the anxiety and depression

scores in the HADS.

Spearman’s Rho was applied to the whole sample (N=92) to examine what variables

were significantly associated with occupational distress and resilience. Ten variables were

explores using a correlation analysis, and as such a more stringent p-value was used a p<.01.

Moreover, p-values are also presented with Bonferroni corrections at p<.005.

Regression analyses were performed with the significant correlates as predictors for

the three outcome variables; CS, STS and burnout. Anticipated mediation analyses to explore

the role of empathy were not undertaken, based on the small sample size, the results from the

regression analyses and correlation results.

Missing dataThe sample (N=92) had a low rate of missing data (0.29 per cent) with <5 per cent of data

missing for any individual case, except for gender, missing 6.5 per cent of the items. A total

of 18 data points were missing for outcome variables and seven data points were missing for

demographic items, mainly gender (N=6). The Little’s Missing Completely at Random

(MCAR) test showed that data for each variable were missing randomly χ2(1192, N=92) =

1117.379, p = .939. To address missing categorical gender data, this was treated as a separate

category for analysis (Appendix L).

The remaining outcome data comprised fewer than two percent of items missing

within less than 15 per cent of the data. All missing items in the outcome variables were

imputed using the median score of each questionnaire item. Imputed data was used in all

subsequent analyses. This method was chosen over complete case analysis to preserve

variability of data and statistical power. Given the small proportions of missing data and

random distribution of missing scores it is unlikely that the imputation will significantly affect

variance or statistical power.

Prospective Power and sample size analysisG*Power 3.1 software (Faul et al., 2007) was used to conduct a power analysis for a linear

multiple regression fixed model, with an estimated medium effect size (f2= .15). The results

indicated that a sample of 103 participants were required for a regression analysis with

possible seven predictors (power=0.80, α=.05).

48

ResultsRespondentsThe survey software identified that 49 UK respondents clicked on the hyperlink. From these

respondents, 32 completed the survey (65 per cent), indicating that the hyperlink was pressed

17 times without the participant proceeding to complete the questionnaire. No participants

were excluded as all fit the inclusion criteria, and all participants checked all items on the

privacy and consent form. A final convenience sample of 32 participants were eligible. Sixty

respondents were obtained from the Norwegian police services, from 110 questionnaires

issued, indicating a response rate of 54.5 per cent.

Test of reliabilityThree variables were shown to have acceptable levels and two of the variables had good

levels of internal consistency (Table 3). One measure showed poor internal consistency values

of α=.57 (SPOS), suggesting that the items in the questionnaire may not be measuring the

same construct.

Table 3.

Reliability analysis with Cronbach’s alpha values for each of the measures, and subtests.

Measure Internal consistency Cronbach’s alpha

ProQOL (30 items) Acceptable α=.678.

CS (10 items) Good α=.883

STS (10 items) Good α=.804

Burnout (10 items) Acceptable α=.616

HADS (14 items) Acceptable α=.724

Anxiety (7 items) Acceptable α=.631

Depression (7items) Acceptable α=.671

EQ (40 items) Good α=.853

SPOS (8 items) Poor α=.57

DemographicsStatistical analyses were performed on descriptive data between the UK and Norwegian

participant group (Table 4). Chi-square tests of independence compared UK and Norwegian

samples for gender revealing no significant difference in gender distribution (X2(1, N=85)

=1.312, p=.252, ϕ =.12). The missing gender data did not significantly differ from the data

reported as male or female (Appendix L). Furthermore, the Norwegian participants were

49

younger, when comparing those over 40, with those under 40 (X2(1, N=92) =5.6135, p=.018,

ϕ =.25). A further chi-square indicated that the UK population reported a significantly greater

attainment of further training to work with vulnerable clients (X2(1, N=91) =10.743, p=.001, ϕ

=.34). A Mann-Whitney U test indicated that the UK population (Mdn=16.00) had

significantly lengthier tenure at work, compared to the Norwegian population (Mdn=9.00),

U=467.000, p<.001.

Table 4.

Sample characteristics of the UK and the Norwegian populations.

UK (N=32) Norway (N=60)

Gender

N % N %

Female 13 40.6 17 28.3

Male 17 53.1 38 63.3

Prefer not to say 1 3.1 - -

Missing data 1 3.1 5 8.3

Age

N % N %

18-29 2 6.3 23 38.3

30-39 10 31.3 15 25

40-49 13 40.6 16 26.7

50-59 7 21.9 6 10

60+ - - - -

Interview training for vulnerable populations

N % N %

Yes 26 81.3 27 45

No 6 18.8 32 53.3

Missing - - 1 1.7

Tenure in post (years)

M SD M SD

18 7.18 10.87 8.90

What levels of distress and resilience (CF/CS) are reported by investigative interviewersin Norway and the UK?

50

With regard to STS, 43.8 per cent of UK interviewers and 35 per cent of Norwegian

interviewers’ scores reported moderate levels of STS. Neither groups reported high levels. For

burnout, the majority of UK participants respondents (71.9 per cent) reported moderate levels,

as did Norwegian participants’ (98.3 per cent). For CS the majority of UK participants

reported moderate scores (87.5 per cent), and the majority of Norwegian interviewers reported

moderate levels (96.7 per cent). The mean scores of the variables are presented in Table 5.

A normative study for the ProQOL (De La Rosa et al., 2018), mainly including health

professionals in an international sample (N=5612), indicated means and standard deviations

for CS (M=37.7, SD=6.5), STS (M=16.7, SD=5.7) and burnout (M=22.8, SD=5.4). Normative

scores were significantly lower for STS and burnout, and higher for CS compared to the

sample mean, indicating higher distress and lower resilience in the sample population. The

effect size for these statistically significant differences are large, as Cohen’s d indicates

almost one standard deviation in differences (Table 5).

Table 5

One-Sample t-tests comparing ProQOL sample means with norm data

M SD df t p d

CS 31.42 5.34 91 -11.274 .001* -1.18

Burnout 27.89 4.66 91 10.488 .001* 1.09

STS 21.16 5.42 91 7.902 .001* 0.82

*Denotes significant findings, with p<.05

Depression scores showed that 3.1 per cent disclosed caseness for the UK group and

8.3 per cent of Norwegian participants scored above the cut-off. Normative data for the

HADS in a UK non-clinical sample (N=1792) showed a mean score of 3.68 (SD=3.07) for

depression and a prevalence of 3.6 per cent for depression (Crawford et al., 2001). The

prevalence of depression in the UK was reported to be 0.5 per cent lower than reported

community norms. Norwegian normative general population data (N=39573) for the HADS

showed a mean score of 3.32 (SD=2.89) for depression (Leiknes et al., 2016). Mean

depression scores were reported as higher than the norm in both the Norwegian and UK

sample (Table 6). Effect sizes for the norm comparisons were large for the Norwegian

population and medium for the UK respondents.

51

Among the UK participants, there was a 21.9 per cent caseness for anxiety prevalence

and the Norwegian participants scored 16.7 per cent for anxiety caseness. Anxiety normative

data (Crawford et al., 2001) for the UK (N=1792) showed a mean score of 6.14 (SD= 3.76),

and a caseness prevalence of 12.6 per cent for anxiety. Norwegian normative general

population data (N=39277) for the HADS showed a mean score of 4.02 (SD=3.29) for

anxiety, and a prevalence rate of 12.6 per cent was noted in the community norms (Leiknes et

al., 2016). The Norwegian sample anxiety mean were significantly higher than the population

norms. This difference demonstrated a large effect size (Table 6). UK anxiety scores showed

no significant difference between the sample and population means.

Table 6.

One sample t-test for HADS anxiety and depression sample scores compared to UK and

Norwegian population norms

M SD df t p d

Norwegian sample

Anxiety 8.45 2.11 59 16.257 .001* 2.10

Depression 5.09 3.47 59 28.105 .001* 3.63

UK sample

Anxiety 7.16 4.54 31 1.265 .205 0.22

Depression 9.42 1.68 31 2.306 .035* 0.41


Five between group analyses were performed to examine statistical difference of the

means, with a more stringent alpha-value set at p=.01 (Table 7). Independent samples t-test

were performed for STS, EQ and SPOS as they met assumptions for parametric tests. Neither

STS (t(90=.920, p=.360) nor EQ variables (t(90)=-2.364, p=.020) showed a statistical

significant difference between means the Norwegian and English participant groups (Table 5),

when using the adjusted alpha value. However, Norwegian participants (N=60) reported

significantly higher levels of POS (M=28.77, SD=10.58, t(90)= -2.855, p=.005) than the UK

participant group (N=32, M=22.13, SD=10.72).

Mann-Whitney U test were performed for burnout and CS, as the variables did not

meet assumptions for parametric tests. Comparison between Norwegian (Mdn=47.92) and UK

participants (Mdn=43.84) indicated no statistical difference in CS (U=875.000, p=.485).

52

However, the Norwegian participants reported significantly higher levels of burnout (Mdn:

52.42), compared to the UK participants (Mdn=35.41), U=605.000, p=.003.

Chi-square tests of independence compared UK and Norwegian samples for caseness

of anxiety and depression (Table 6), revealing no significant difference caseness in anxiety

(X2(1, N=92) = 0.2758, p=.540). However, Norwegian participants were more likely to report

case-level depression (X2(1, N=92) =6.9511, p=.008).

Table 7.

Means and Standard deviation on measures for each participant group, with chosen

statistical analysis, effect sizes and p-values.

UK sample Norwegian

sample

Statistical analysis Effect

sizes

p-

values

M SD M SD

STS 21.88 5.82 20.78 5.20 Independent t-test d=.20 .360

Burnout 25.88 5.94 28.97 3.39 Mann-Whitney U r=.89 .003*

CS 30.66 6.24 31.83 4.80 Mann-Whitney U r=.05 .458

Depression 5.09 3.47 9.42 1.68 Chi-square ϕ =.28 .008*

Anxiety 7.16 4.54 8.45 2.11 Chi-Square ϕ =.05 .540

Empathy 43.00 12.70 48.47 9.25 Independent t-test d=.49 .020

SPOS 22.13 10.72 28.77 10.58 Independent t-test d=.62 .005*


53

Table 8.

Percentages of distribution of level ranking for each of the measured variables, between

Norwegian and UK participant group.

United Kingdom Norway

Category levels N % N %

STS Low 18 56.2 39 65

Moderate 14 43.8 21 35

High - - - -

Burnout Low 9 28.1 1 1.7

Moderate 23 71.9 59 98.3

High - - - -

CS Low 3 9.4 2 3.3

Moderate 28 87.5 58 96.7

High 1 3.1 - -

Depression Normal 24 75 15 25

Borderline 7 21.9 40 66.7

Abnormal 1 3.1 5 8.3

Anxiety Normal 18 56.3 18 30

Borderline 7 21.9 32 53.3

Abnormal 7 21.9 10 16.7

Empathy Low 8 25 3 5

Average 18 56.2 38 63.3

Above average 3 9.4 16 26.7

High 3 9.4 3 5

SPOS Low 8 25 8 13.3

Moderate 19 59.4 28 46.7

High 5 15.6 24 40

54

What factors are associated with reported distress and resilience?Relationships between ten variables were examined for the entire sample (N=92).

Correlational results for scores on CS, STS, burnout, anxiety, depression, empathy, POS and

participant demographics are presented in Table 9. Gender was found to have no significant

relationship with any of the proposed variables. Age was unsurprisingly significantly related

tenure (r=.90, p<.005) as they are similar constructs.

The three main constructs (CS, STS and burnout) showed some significant

correlations with each other. CS correlated negatively with burnout (r=-.40), but not with

STS. Burnout and STS revealed a strong positive correlation with each other (r=.51, p<.005),

supporting the construct of CF.

When examining how the three main constructs related to other variables, the results

showed that CS also correlated with anxiety (r=-.29, p<.01), empathy (r=.28, p<.01) and POS

(r=.49, p<.005). Burnout significantly correlated with depression (r=.42, p<.005) and anxiety

(r=.47, p<.005), in addition to STS. STS correlated strongly with anxiety (r=.47, p<.005).

Empathy showed no relationship with the CF-constructs, but appeared significantly correlated

with CS (r=.28, p<.01) with a low effect size. Empathy was signiEmpaEmpathy also

correlated significantly with Empathy was also significantly correlatied with POS (r

55

Table 9.

Bivariate non-parametric correlations between ten variables

Gender Age Tenure CS Burnout STS Depression Anxiety Empathy POS

Gender 1

Age .061 1

Tenure .011 .895** 1

CS -.084 .227 .098 1

Burnout .125 -.218 -.246 -.404** 1

STS .017 .002 .020 -.201 .512** 1

Depression -.083 -.203 -.260 -.139 .419** .208 1

Anxiety -.085 -.157 -.123 -.290* .467** .597** -.439** 1

Empathy .132 .010 -.065 .281* .520 -.133 .240 -.047 1

POS .126 -.165 -.278 .491** -.208 -.260 .095 -.241 .244* 1

**Denotes significant correlations at the .005 level (2-tailed).

*Denotes significant correlations at the .01 level (2-tailed).

56

To what extent do noted correlates predict reported distress and resilience?To address this, simultaneous forced entry multiple regression analyses (MRA) and a single

regression analysis were conducted, for each of the dependent variables; STS burnout and CS

with variables previously shown to have a relationship added simultaneously to the MRA.

Results are presented in Table 10 and 11. For STS, a simple linear regression was conducted

to examine the degree to which anxiety predicts STS, revealing a significant equation (F(1,90)

= 52.731, p<.001), with an R2 of .369.

For burnout, a forced entry MRA with anxiety and depression as predictors, revealed a

significant effect of the model (F(2, 89) = 29.769, p<.001) with an R2 of .401. The coefficient

analyses indicated that anxiety (t=4.158, p<.001) and depression (t=3.073, p=.003) were both

significant predictors in the model. Anxiety had the largest unique contribution of the

predictor variables in this model.

For CS a forced entry MRA of anxiety, EQ and POS revealed a significant model

(F(3, 88) =13.318, p<.001), with an R2 of .312. POS (t=4.699, p<.001) was the only

significant predictor in the model.

Does reported empathy moderate/mediate the relationship with CS/CF and proposedpredictor variables in investigative interviewers?Comparative examination of reported empathy (at 75 per cent of the UK sample and 95 per

cent of Norwegian sample reporting average or higher categories on the EQ-scale) revealed

no significant difference. Whilst empathy showed significant positive correlations with CS

(r=.28), no significant relationship was evidenced with either STS or burnout. Moreover,

MRA showed that empathy was not a significant contributor in the CS regression model.

Given this, and a small, underpowered study, model construction via mediation analysis was

not pursued.

57

Table 10.

Simple regression model for STS on anxiety.

Variable Predictor B SEb P R2 change F P R square Adjusted

R2

STS .369 52.731 .000 .369 .362

Anxiety 1.026 .141 .000

Table 11.

Multiple linear regression model of predictors of CS and burnout.

Model Independent

variables

B SEb b’ t P R2 change F P R

square

Adjusted

R2

CS .312 13.318 .000 .312 .289

Anxiety -.293 .153 -.176 -1.914 .059

EQ .066 .045 .133 1.475 .144

SPOS .214 .046 .442 4.699 .000

Burnout .401 29.769 .000 .401 .387

Anxiety .596 .144 .304 3.073 .003

Depression .458 .153 .313 3.001 .004

58

DiscussionThis exploratory study sought to assess the extent of work-related distress and resilience

(measured via STS, burnout and CS) reported by police investigative interviewers

operating in similar systems in Europe (the UK and Norway). The study also examined

significant correlates and predictors of distress and considered empathy as a possible

mediating variable between independent and dependent variables.

Summary of findingsReported levels of CF (STS and burnout) fell within moderate and low categories for

both UK and Norwegian interviewers, with CS also largely reported as moderate. The

findings reveal CF at levels (low to moderate) similar to those reported by police

officers working with rape victims (Turgoose et al., 2017) and first responders

(Greinacher et al., 2019). Nevertheless, comparisons with community norms suggest

that both the UK and Norwegian respondents are disclosing elevated distress, with

higher mean scores in anxiety, depression, STS and burnout. Moreover, levels of

reported CS appeared lower than community norms. Differences showed large effect

sizes when compared to the norms, for the three main constructs and for the Norwegian

participant group’s anxiety and depression scores, compared to the norm population.

Examination between UK and Norwegian interviewers found significant

differences in burnout, caseness for depression and POS. Surprisingly, Norwegian

participants reported higher scores for burnout, despite being a younger group. Burnout

is often understood to arise from cumulative effect, although the group reported mean

time in post exceeding 10 years. However, they also appeared to be younger and have

fewer years in post than their UK counterparts, so this distress may be a manifestation

of less experience or emotional resource to manage the challenge of interviewing.

Although appearing inconsistent with the fact that Norwegian interviewers also reported

higher POS, it may be that whilst Norwegian police officers have supportive measures

and additional training during early years of employment, this is does not necessarily

mitigate distress or impact on individual resilience.

Regarding relationships with distress measures, burnout was significantly related

to wider psychological morbidity, such that higher anxiety and depression correlated

positively. STS was also correlated with anxiety, which is unsurprising given the role of

anxiety in conceptualisation of STS as a trauma reaction. Correlations between burnout

and STS, which show similar correlations with other studies on police (Turgoose et al.,

59

2017), could indicate construct confirmation of CF. Burnout was negatively associated

with CS, indicating that higher burnout levels was associated with lower reported levels

of CS. Empathy and POS were both positive correlates of CS, which was observed in

studies with social work staff (Thomas & Otis, 2010). Contradicting Greinacher et al.

(2019), the current study found no effects of gender. Moreover, this study found no

correlations between CF and CS, which have been suggested to be opposite but related

processes (Stamm, 2010). Perhaps their separateness may help understand why empathy

correlates with CS and not with CF. Linley and Joseph (2007) suggested that for

therapists, CS is not the obverse of CF and they are related to different processes.

Multiple regression analysis significantly explained 31.2 per cent of variance in

CS with three predictors: anxiety; empathy; and POS. POS was the only significant

contributors to variance in the model. In a simple regression model, anxiety was found

to predict 36.9 per cent of STS scores significantly. Finally, a multiple regression

analysis was conducted for burnout, where depression and anxiety were predictors. This

model indicated that 40.1 per cent explained variance in the model. Though the

regression models account for a significant part of variability, the prediction levels also

indicate presence of other explanatory variables. Potentially this could include social

support or alcohol use (Greinacher et al., 2019), different elements of empathy or

mindfulness coping strategies (Thomas & Otis, 2010).

Empathy only showed significant correlation with CS, which within this

underpowered study, deterred pursuing mediation analysis. However, it is notable that

as with the study examining empathy in police interviewers of rape victims (Turgoose et

al. (2017), no relationship was evident with CF. This differs from studies of caring

professions, in which empathy has shown significant association with CF (Duarte et al.,

2016; Wagaman et al., 2015). No significant relationship was found between empathy

and work-related distress, which may reflect the non-therapeutic role of investigative

interviews or that the pragmatically selected measure of empathy (for language reasons)

was insufficient to capture the specific elements of empathy needed to build rapport or

relate to interviewees. As empathy levels seem high within this group, the results may

reflect engagement in emotional suppression against distress, or the measure lacks a

focus on empathy related to the specific skills in investigative interviewing. Potentially,

the methodology of cognitive interviewing techniques and the PEACE model inhibits

the emotional engagement with the work. As Turgoose et al. (2017) has noted, the

60

conceptual model Figley (1995) advances is not clear about what facets of empathy

potentiate CF (Sabo, 2011), so further consideration of how empathy is measured

towards vulnerable witnesses is needed.

Strengths and limitationsThe current study is the first to examine the extent and predictors of work-related

distress and resilience in investigative interviewers, using robust measures amenable to

exploration in two comparable national police services. Indeed selected questionnaires

were given availability and validity for both UK and Norwegian populations.

Despite endorsement and the use of national gatekeepers such as the UK College

of Policing, recruitment in the timeframe was insufficient to power the study fully, and

caution should therefore be exercised in data interpretation. In both UK and Norway

legislatures budgetary cuts have increased pressure on forces such that this study may

have been viewed as of low priority whilst significant job stressors, increased workloads

and limited resources are evident (HMIC, 2012). Low response rate may also be

associated with interviewers’ concerns about social desirability, fear of job loss

(Kronenberg et al., 2008) and reluctance to disclose distress given known covert

cultural pressures to use emotional suppression to cope. Response bias may have been

created by officers’ reluctance to endorse symptoms of a psychological burden, and

electing not to participate. Bias may also have been created through those officers who

felt most distressed by their work having additional motivation to respond to a study

with this focus.

This study did not control for impacts of direct trauma, and as such the levels of

distress might have been impacted by variables unbeknownst to the principle researcher.

However, recruitment methods made sure to target those police officers who are in the

main role of interviewers, and the CF-CS measure was specified to consider the

investigative interviewing role in particular.

ImplicationsPsychological morbidity, in both UK and Norwegian samples, exceeded community

norms for both anxiety and depression, with particularly large effect sizes for the

comparisons with the Norwegian sample. On dimensions of work-related distress, a

considerable proportion of investigative interviewers reported moderate levels of both

STS and burnout, in the UK and Norway; again above community norms and with large

61

effect sizes. There appeared a need to acknowledge that there are significant

psychological challenges inherent in undertaking interviewing of this type, as well as

greater organisational awareness and communication to normalise emotional challenges

and risks of distress from work in this area. There has been awareness that multiple

interviews of this type can be traumatising to witnesses and victims, and addressed via

initiatives such as Barnahus (Haldorsson, 2017). It seems timely to consider parallel

innovations to support staff and to ensure that service to victims through interview

quality is maintained.

Whilst there were some differences between UK and Norwegian samples there

were also consistent findings. The study suggested benefits in examining impacts of

work in similar legislatures offering similar support to witnesses and victims.

Collaborative examination of distress and resilience in policing between nationalities

can facilitate mutual understanding of strengths and weaknesses in procedures, in a

world where international police cooperation is growing.

In addition, this research raises questions regarding organisational resilience

measures and their efficacy, as well as specific coping mechanisms and possible

inherent resilience in those who have chosen a policing career. The results might be

supportive of successful measures aimed to promote good mental health and wellbeing,

though the comparison to community norms still show a higher level of distress in the

sample. The potential links between CF and empathy remain unclear and raises

questions relating to the theoretical background of CF as a construct.

Awareness of the moderate levels of distress, which are greater than community

norms, should encourage further studies and support, particularly with literature

revealing comparable levels of distress (Greinacher et al., 2019; Turgoose et al., 2017).

Addressing police wellbeing in the UK has recently been enhanced through Oscar Kilo

and the Bluelight frameworks (Coleman et al., 2018). To date, the impact of trauma has

prioritised the needs of those officers more obviously exposed to trauma (e.g. through

direct violence to themselves, road traffic fatalities or breaking bad news to families of

victims). The data, revealing elevated distress and its correlates, suggests the emotional

needs of those occupying less obviously distressing roles in interviewing should be

further assessed and supportive interventions developed.

Future research

62

Although it is reasonable to assume that being indirectly exposed to trauma might have

negative impacts or contribute to psychological distress, the cross-sectional nature of

this explorative study cannot provide a definite answer regarding causality. Studies

utilising larger samples and longitudinal designs can contribute to more robust

understanding of how psychological and occupational distress and resilience develops

in investigative interviewers. Whilst both UK and Norwegian interviewers had been in

post for some time, it would also be useful to examine those who have been in post for

significant periods to examine the potential cumulative impact of their work

The CF-CS model could potentially benefit from some restructuring and a

clearer examination of how occupational distress or resilience forms. Although empathy

has been suggested to be the lead process through which CF and CS is developed, these

findings do not support that. It would be beneficial to understand better how these

processes occur, to provide a more robust theoretical standing to the model.

ConclusionThis is the first exploratory study to examine levels and predictors of work-related

distress and resilience in investigative interviewers. It revealed elevated psychological

morbidity and work-related distress, as well as lower CS in national police services in

the UK and Norway. The suggested role of empathy as a mediating factor was not

evidenced and more research is required both to further understand factors influencing

wellbeing, and strategies to counter distress. Nevertheless, findings suggest that further

organisational strategies should be offered to support police interviewers, which might

help to maintain resilience and sustain them in key roles to aid the vulnerable and secure

prosecutions.

63

ReferencesAndersen, J. P., & Papazoglou, K. (2015). Compassion fatigue and compassion

satisfaction among police officers: An understudied topic. International Journal

of Emergency Mental Health and Human Resilience, 17(3), 661-663

Arnold, D., Calhoun, L.G., Tedeschi, R & Cann, A. (2005). Vicarious posttraumatic

growth in psychotherapy. Journal of Humanistic Psychology. 45(2), 239-263

Baron-Cohen, S., & Wheelwright, S. (2004). The empathy quotient: An investigation of

adults with Asperger syndrome or high functioning autism, and normal sex

differences. Journal of Autism and Developmental Disorders, 34(2), 163–175

Bjelland, I., Dahl, A. A., Haug, T. T., & Neckelmann, D. (2002). The validity of the

Hospital Anxiety and Depression Scale: an updated literature review. Journal of

Psychosomatic Research, 52(2), 69-77.

Bruusgard, R., Bjerknes, O.T., Buseth, S. et al. (2013). Riksadvokatens arbeidsgruppe:

Avhørsmetodikk i Politiet. Available via https://www.riksadvokaten.no/wp-

content/uploads/2017/10/Avh%C3%B8rsmetodikk-i-Politiet-med-vedlegg.pdf

retrieved 24 April 2020

Bull, R. (2014). Investigative interviewing. New York: Springer Science & Business

Media

Canfield, J. (2005). Secondary traumatization, burnout, and vicarious traumatization: A

review of the literature as it relates to therapists who treat trauma. Smith College

Studies in Social Work, 75(2), 81-101.

Coleman, R., Kirby, S., Birdsall, N., Cox, C., & Boulton, L. (2018). Wellbeing in

Policing: Blue Light Wellbeing Framework. University of Central Lancashire

and Lancashire Constabulary

College of Policing (2019). Investigation: Investigative interviewing. Available via

https://www.app.college.police.uk/app-content/investigations/investigative-

interviewing/ Retrieved 23 April 2020

Collins, S., & Long, A. (2003). Working with the psychological effects of trauma:

consequences for mental health‐care workers–a literature review. Journal of

Psychiatric and Mental Health Nursing, 10(4), 417-424.

Crawford, R., Disney, R., & Simpson, P. (2017). Police officer retention in England and

Wales. Institute For Fiscal Studies for Economic and Social Research Council.

Available via https://www.ifs.org.uk/publications/8824 Retrieved 25 Mai 2020

64

Crawford, J. R., Henry, J. D., Crombie, C., & Taylor, E. P. (2001). Normative data for

the HADS from a large non‐clinical sample. British Journal of Clinical

Psychology, 40(4), 429-434.

Cross, L.C. & Ashley, L. (2004). Police trauma and addiction: Coping with the dangers

of the job. FBI Law Enforcement Bulletin, 73(10), 24-32.

De La Rosa, G. M., Webb-Murphy, J. A., Fesperman, S. F., & Johnston, S. L. (2018).

Professional quality of life normative benchmarks. Psychological Trauma:

Theory, Research, Practice, and Policy, 10(2), 225.

Duarte, J., Pinto-Gouveia, J., & Cruz, B. (2016). Relationships between nurses’

empathy, self- compassion and dimensions of professional quality of life: A

cross-sectional study. International Journal of Nursing Studies, 60, 1-11.

Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived

organizational support. Journal of Applied Psychology, 71(3), 500-507.

Eisenberger, R. (2019). Perceived Organizational Support.

http://classweb.uh.edu/eisenberger/perceived-organizational-support/ Retrieved

on 12th February 2019

Elwood, L. S., Mott, J., Lohr, J. M., & Galovski, T. E. (2011). Secondary trauma

symptoms in clinicians: A critical review of the construct, specificity, and

implications for trauma focused treatment. Clinical Psychology Review, 31(1),

25-36.

Evans, S., Huxley, P., Gately, C. et al. (2006). Mental health, burnout and job

satisfaction among mental health social workers in England and Wales. The

British Journal of Psychiatry, 188(1), 75-80

Evces, M. R. (2015). What is vicarious trauma? In G. Quintangon & M. R. Evces (eds),

Vicarious trauma and disaster mental health (pp. 9-23). Routledge.

Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G* Power 3: A flexible

statistical power analysis program for the social, behavioral, and biomedical

sciences. Behavior Research Methods, 39(2), 175-191.

Figley (1995). Compassion Fatigue: Coping With Secondary Traumatic Stress Disorder

In Those Who Treat The Traumatised. Routledge, New York NY.

Goubert, L., Craig, K. D., Vervoort, T., et al. (2005). Facing others in pain: the effects

of empathy. Pain, 118(3), 285-288.

65

Haldorsson, O. L. (2017). Barnahus Quality Standards Summary: Guidance for

Multidisciplinary and Interagency Response to Child Victims and Witnesses of

Violence. Council of the Baltic Sea States Secretaries and Child Circle.

Available via http://www.childrenatrisk.eu/promise/

wp-content/uploads/sites/4/2015/11/Guidance_short_reformatted.pdf

Hanway, P. & Akehurst, L. (2018). The voice from the front line: Police officers’

perceptions of real-world interviewing with vulnerable witnesses. Investigative

Interviewing: Research and Practise. 9(1) 14-33

Her Majesty’s Inspectorate of Constabulary (HMIC) (2012). Policing in Austerity: One

Year On. Available via https://www.justiceinspectorates.gov.uk/

hmicfrs/media/policing-in-austerity-one-year-on.pdf Retrieved on 31 May 2020.

Hunt, P.A., Denieffe, S. & Gooney, M. (2017). Burnout and its relationship to empathy

in nursing: A review of the literature. Journal of Research in Nursing. 22(1)

Karlsson, I. & Christianson, S-A. (2003). The phenomenology of traumatic experiences

in police work. Policing: An International Journal of Police Strategies &

Management, 26(3), 419-438.

Kronenberg, M., Osofsky, H. J., Osofsky, J. D. et al. (2008). First responder culture:

Implications for mental health professionals providing services following a

natural disaster. Psychiatric Annals, 38(2), 114–118.

Lawrence, E. J., Shaw, P., Baker, D., et al. (2004). Measuring empathy: reliability and

validity of the Empathy Quotient. Psychological Medicine, 34(5), 911-920.

Lassiter, G. D. (Ed.) (2004). Interrogations, Confessions, and Entrapment. New York:

Springer.

Leiknes, K. A., Dalsbø, T. K. & Siqveland, J. (2016). Psychometric assessment of the

Norwegian version of the Hospital Anxiety and Depression Scale (HADS). Oslo:

Norwegian Institute of Public Health.

MacEachern, A. D., Jindal-Snape, D., & Jackson, S. (2011). Child abuse investigation:

police officers and secondary traumatic stress. International Journal of

Occupational Safety and Ergonomics, 17(4), 329-339.

Maslach, C., & Jackson, S. E. (1981). The measurement of experienced

burnout. Journal of Organizational Behavior, 2(2), 99-113.

Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach Burnout Inventory. 3rd ed.

Palo Alto, CA: Consulting Psychologists Press.

66

Mathieu, F. (2007). Running on empty: Compassion fatigue in health professionals.

Retrieved from http://www.compassionfatigue.org/pages/RunningOnEmpty.pdf

Retrieved 4 May 2020

McFadden, P., Campbell, A., & Taylor, B. (2014). Resilience and burnout in child

protection social work: Individual and organisational themes from a

systematic literature review. British Journal of Social Work, 45(5), 1546-1563.

Milne, R. & Bull, R. (1999). Investigative Interviewing: Psychology and Practice.

Chichester: Wiley.

Najjar, N., Davis, L W., Beck-Coon, K. & Doebbeling, C.C. (2009). Compassion

fatigue: A Review of the Research to Date and Relevance to Cancer-Care

Providers. Journal of Health Psychology. 14(2) 267-277

Papazoglou, K., Bonanno, G., Blumberg, D., & Keesee, T. (2019). Moral injury in

police work. FBI Law Enforcement Bulletin.

Papazoglou, K., Koskelainen, M. & Steuwe, N. (2017). Exploring the Role of

Compassion Satisfaction and Compassion Fatigue in Predicting Burnout among

Police Officers. Open Journal of Psychiatry & Allied Sciences. 9(2), 107-112

Perneger, T. V. (1998). What's wrong with Bonferroni adjustments. Bmj, 316(7139),

1236-1238.

Risan, P., Milne, R., & Binder P. (2016). Regulating and coping with distress during

police interviews of traumatised victims. Psychological Trauma: Theory,

Research, Practise and Policy. 8(6), 736-744

Shore, L.M. & Tetrick, L.E. (1991). A Construct Validity Study of the Survey of

Perceived Organizational Support. Journal of Applied Psychology, 76(5), 637

643

Silverstone, P. H. (1994). Poor efficacy of the Hospital Anxiety and Depression Scale in

the diagnosis of major depressive disorder in both medical and psychiatric

patients. Journal of Psychosomatic Research, 38(5), 441-450.

Stamm, B. H. (2002). Measuring compassion satisfaction as well as fatigue:

Developmental history of the Compassion Satisfaction and Fatigue Test. In C. R.

Figley Ed.), Psychosocial stress series, no. 24. Treating compassion fatigue (p.

107–119). Brunner-Routledge.


67

Thomas, J. T., & Otis, M. D. (2010). Intrapsychic correlates of professional quality of

life: Mindfulness, empathy, and emotional separation. Journal of the Society for

Social Work and Research, 1(2), 83-98.

Turgoose, D., Glover, N., Barker, C., & Maddox, L. (2017). Empathy, compassion

fatigue, and burnout in police officers working with rape

victims. Traumatology, 23(2), 205.

Zigmond, A.S. & Snaith, R.P. (1983). The hospital anxiety and depression scale. Acta

Psychiatrica Scandinavica, 67(6)

Watts, J., & Robertson, N. (2015). Selecting a measure for assessing secondary trauma

in nurses. Nurse Researcher, 23(2).

Worley, J. A., Fuqua, D. R., & Hellman, C. M. (2009). The survey of perceived

organizational support: Which measure should we use? SA Journal of Industrial

Psychology, 35(1), 112-116.

Yang, Y. H. & Kim, J. K. (2012). A literature review of compassion fatigue in

nursing. Korean Journal of Adult Nursing, 24(1), 38-51.

68

Appendices:Mandatory appendices are marked with an asterisk *

69

Appendix A*: Guidelines to authors for journal targeted for literature review

About the Journal: Psychiatry, Psychology and Law is an international, peer-reviewed

journal publishing high-quality, original research. Please see the journal's Aims &

Scope for information about its focus and peer-review policy. Please note that this

journal only publishes manuscripts in English.

Psychiatry, Psychology and Law accepts the following types of article: original articles

and empirical studies, analyses of professional issues, controversies and developments

in these areas, case studies and case commentaries, book reviews.

Peer Review and Ethics: Taylor & Francis is committed to peer-review integrity and

upholding the highest standards of review. Once your paper has been assessed for

suitability by the editor, it will then be double blind peer reviewed by independent,

anonymous expert referees. Find out more about what to expect during peer review and

read our guidance on publishing ethics.

Preparing Your Paper: All authors submitting to medicine, biomedicine, health

sciences, allied and public health journals should conform to the Uniform Requirements

for Manuscripts Submitted to Biomedical Journals, prepared by the International

Committee of Medical Journal Editors (ICMJE).

Structure: Your paper should be compiled in the following order: title page; abstract;

keywords; main text introduction, materials and methods, results, discussion;

acknowledgments; declaration of interest statement; references; appendices (as

appropriate); table(s) with caption(s) (on individual pages); figures; figure captions (as a

list).

Word Limits: Please include a word count for your paper. A typical paper for this

journal should be no more than 12000 words, inclusive of tables, references, figure

captions.

Style Guidelines: Please refer to these quick style guidelines when preparing your

paper, rather than any published articles or a sample copy. Please use British (-ise)

spelling style consistently throughout your manuscript. Any form of consistent

quotation style is acceptable. Please note that long quotations should be indented

without quotation marks.

Psychological manuscripts should be prepared in accordance with the format and style

specified in the ‘Publication Manual of the American Psychological Association’, fifth

edition. Pages should be numbered consecutively. References should be cited in the text

70

as specified in the Publication Manual of the American Psychological Association, fifth

edition. A concise description of APA referencing style can be found her.

Personal communications should be cited as such in the text and should not be included

in the reference list. For an overview of APA style (including referencing)

visit http://www.lib.monash.edu.au/tutorials/citing/apa.html

Formatting and Templates: Papers may be submitted in Word format. Figures should

be saved separately from the text. To assist you in preparing your paper, we provide

formatting template(s). Word templates are available for this journal. Please save the

template to your hard drive, ready for use. If you are not able to use the template via the

links (or if you have any other template queries) please contact us here.

References: Please use this reference guide when preparing your paper.

Checklist: What to Include

1. Author details. Please ensure everyone meeting the International Committee of

Medical Journal Editors (ICMJE) requirements for authorship is included as an

author of your paper. All authors of a manuscript should include their full name

and affiliation on the cover page of the manuscript. Where available, please also

include ORCiDs and social media handles (Facebook, Twitter or LinkedIn). One

author will need to be identified as the corresponding author, with their email

address normally displayed in the article PDF (depending on the journal) and the

online article. Authors’ affiliations are the affiliations where the research was

conducted. If any of the named co-authors moves affiliation during the peer-

review process, the new affiliation can be given as a footnote. Please note that

no changes to affiliation can be made after your paper is accepted. Read more on

authorship.

2. Should contain an unstructured abstract of 150 words.

3. Graphical abstract (optional). This is an image to give readers a clear idea of

the content of your article. It should be a maximum width of 525 pixels. If your

image is narrower than 525 pixels, please place it on a white background 525

pixels wide to ensure the dimensions are maintained. Save the graphical abstract

as a .jpg, .png, or .tiff. Please do not embed it in the manuscript file but save it as

a separate file, labelled GraphicalAbstract1.

71

4. You can opt to include a video abstract with your article. Find out how these

can help your work reach a wider audience, and what to think about when

filming.

5. At least 10 keywords. Read making your article more discoverable, including

information on choosing a title and search engine optimization.

6. Funding details. Please supply all details required by your funding and grant-

awarding bodies as follows: For single agency grants. This work was supported

by the [Funding Agency] under Grant [number xxxx]. For multiple agency

grants: This work was supported by the [Funding Agency #1] under Grant

[number xxxx]; [Funding Agency #2] under Grant [number xxxx]; and [Funding

Agency #3] under Grant [number xxxx].

7. Disclosure statement. This is to acknowledge any financial interest or benefit

that has arisen from the direct applications of your research. Further guidance on

what is a conflict of interest and how to disclose it.

8. Supplemental online material. Supplemental material can be a video, dataset,

fileset, sound file or anything which supports (and is pertinent to) your paper.

We publish supplemental material online via Figshare. Find out more

about supplemental material and how to submit it with your article.

9. Figures. Figures should be high quality (1200 dpi for line art, 600 dpi for

grayscale and 300 dpi for colour, at the correct size). Figures should be supplied

in one of our preferred file formats: EPS, PS, JPEG, TIFF, or Microsoft Word

(DOC or DOCX) files are acceptable for figures that have been drawn in Word.

For information relating to other file types, please consult our Submission of

electronic artwork document.

10. Tables. Tables should present new information rather than duplicating what is in

the text. Readers should be able to interpret the table without reference to the

text. Please supply editable files.

11. Equations. If you are submitting your manuscript as a Word document, please

ensure that equations are editable. More information about mathematical

symbols and equations.

12. Units. Please use SI units (non-italicized).

Using Third-Party Material in your Paper: You must obtain the necessary

permission to reuse third-party material in your article. The use of short extracts of text

72

and some other types of material is usually permitted, on a limited basis, for the

purposes of criticism and review without securing formal permission. If you wish to

include any material in your paper for which you do not hold copyright, and which is

not covered by this informal agreement, you will need to obtain written permission from

the copyright owner prior to submission. More information on requesting permission to

reproduce work(s) under copyright.

Disclosure Statement: Please include a disclosure statement, using the subheading

“Disclosure of interest.” If you have no interests to declare, please state this (suggested

wording: The authors report no conflict of interest). For all NIH/Wellcome-funded

papers, the grant number(s) must be included in the declaration of interest

statement. Read more on declaring conflicts of interest.

Clinical Trials Registry: In order to be published in a Taylor & Francis journal, all

clinical trials must have been registered in a public repository at the beginning of the

research process (prior to patient enrolment). Trial registration numbers should be

included in the abstract, with full details in the methods section. The registry should be

publicly accessible (at no charge), open to all prospective registrants, and managed by a

not-for-profit organization. For a list of registries that meet these requirements, please

visit the WHO International Clinical Trials Registry Platform (ICTRP). The registration

of all clinical trials facilitates the sharing of information among clinicians, researchers,

and patients, enhances public confidence in research, and is in accordance with

the ICMJE guidelines.

Complying With Ethics of Experimentation: Please ensure that all research reported

in submitted papers has been conducted in an ethical and responsible manner, and is in

full compliance with all relevant codes of experimentation and legislation. All papers

which report in vivo experiments or clinical trials on humans or animals must include a

written statement in the Methods section. This should explain that all work was

conducted with the formal approval of the local human subject or animal care

committees (institutional and national), and that clinical trials have been registered as

legislation requires. Authors who do not have formal ethics review committees should

include a statement that their study follows the principles of the Declaration of Helsinki.

Consent: All authors are required to follow the ICMJE requirements on privacy and

informed consent from patients and study participants. Please confirm that any patient,

service user, or participant (or that person’s parent or legal guardian) in any research,

73

experiment, or clinical trial described in your paper has given written consent to the

inclusion of material pertaining to themselves, that they acknowledge that they cannot

be identified via the paper; and that you have fully anonymized them. Where someone

is deceased, please ensure you have written consent from the family or estate. Authors

may use this Patient Consent Form, which should be completed, saved, and sent to the

journal if requested.

Health and Safety: Please confirm that all mandatory laboratory health and safety

procedures have been complied with in the course of conducting any experimental work

reported in your paper. Please ensure your paper contains all appropriate warnings on

any hazards that may be involved in carrying out the experiments or procedures you

have described, or that may be involved in instructions, materials, or formulae.

Please include all relevant safety precautions; and cite any accepted standard or code of

practice. Authors working in animal science may find it useful to consult

the International Association of Veterinary Editors’ Consensus Author Guidelines on

Animal Ethics and Welfare and Guidelines for the Treatment of Animals in Behavioural

Research and Teaching. When a product has not yet been approved by an appropriate

regulatory body for the use described in your paper, please specify this, or that the

product is still investigational.

Submitting Your Paper: This journal uses ScholarOne Manuscripts to manage the

peer-review process. If you haven't submitted a paper to this journal before, you will

need to create an account in ScholarOne. Please read the guidelines above and then

submit your paper in the relevant Author Centre, where you will find user guides and a

helpdesk.

Please note that Psychiatry, Psychology and Law uses Crossref™ to screen papers for

unoriginal material. By submitting your paper to Psychiatry, Psychology and Law you

are agreeing to originality checks during the peer-review and production processes. On

acceptance, we recommend that you keep a copy of your Accepted Manuscript. Find out

more about sharing your work.

Publication Charges: There are no submission fees, publication fees or page charges

for this journal. Colour figures will be reproduced in colour in your online article free of

charge. If it is necessary for the figures to be reproduced in colour in the print version, a

charge will apply. Charges for colour figures in print are £300 per figure ($400 US

Dollars; $500 Australian Dollars; €350). For more than 4 colour figures, figures 5 and

74

above will be charged at £50 per figure ($75 US Dollars; $100 Australian Dollars; €65).

Depending on your location, these charges may be subject to local taxes.

Copyright Options: Copyright allows you to protect your original material, and stop

others from using your work without your permission. Taylor & Francis offers a

number of different license and reuse options, including Creative Commons licenses

when publishing open access. Read more on publishing agreements.

Complying with Funding Agencies: We will deposit all National Institutes of Health

or Wellcome Trust-funded papers into PubMedCentral on behalf of authors, meeting the

requirements of their respective open access policies. If this applies to you, please tell

our production team when you receive your article proofs, so we can do this for you.

Check funders’ open access policy mandates here. Find out more about sharing your

work.

My Authored Works: On publication, you will be able to view, download and check

your article’s metrics (downloads, citations and Altmetric data) via My Authored

Works on Taylor & Francis Online. This is where you can access every article you have

published with us, as well as your free eprints link, so you can quickly and easily share

your work with friends and colleagues. We are committed to promoting and increasing

the visibility of your article. Here are some tips and ideas on how you can work with us

to promote your research.

Article Reprints: You will be sent a link to order article reprints via your account in

our production system. For enquiries about reprints, please contact the Taylor & Francis

Author Services team at [email protected]. You can also order print copies of the

journal issue in which your article appears.

Queries: Should you have any queries, please visit our Author Services website or

contact us here.

Updated 6-05-2020

75

Appendix B: Expanded details on search methodology

Database Applied filters Secondary trauma

key terms

Legal profession

key terms

PsychINFO English

Peer reviewed

Peer-reviewed

journal

Exclude

dissertations

burnout OR

“compassion

fatigue” OR

“secondary

trauma*” OR

vicarious trauma*

OR PTSD OR

“Post-traumatic

stress*” OR “Post

traumatic stress*”

solicitor* OR

barrister* OR

lawyer* OR

attorney* OR

judge* OR legal

PILOTS Peer reviewed

Journal article

English

PubMed English

Journal articles

Web of science Article

English

Westlaw UK Journal

Business source

premier

Academic journal

Article

Criminal justice

abstracts

Scholarly peer-

reviewed journals

English

Article

Academic journal

76

Appendix C: Data extraction form

Reference

Year

Country of study:

Quality appraisal score:

Secondary trauma

definition:

Aims/hypotheses:

Sample (size,

demographic)

Demographics

Age:

Gender:

Ethnicity:

Other

Sampling method

Study design (procedure)

Measures (validity,

amendments)

Analysis

(Reliability/validity

statistic, parametric/non-

parametric/effect size)

Reliability/Validity Statistic:

Cronbach’s Alpha:

Sensitivity:

Specificity:

Other:

Results (effect size):

Conclusions:

Limitations/biases:

Clinical Implications/

Implications for further

research:

Additional comments:

Data extraction form

77

Appendix D: AXIS Quality appraisal tool items

Introduction

1. Were the aims/objectives of the study clear?

Methods

2. Was the study design appropriate for the stated aim(s)?

3. Was the sample size justified?

4. Was the target/reference population clearly defined? (Is it clear who the research

was about?)

5. Was the sample frame taken from an appropriate population base so that it

closely represented the target/reference population under investigation?

6. Was the selection process likely to select subjects/participants that were

representative

7. Were measures undertaken to address and categorise non-responders?

8. Were the risk factors and outcome variables measured appropriate to the aims of

the study?

9. Were the risk factors and outcome variables measured correctly using

instruments/measurements that had been trialled, piloted or published

previously?

10. Is it clear what was used to determined statistical significance and/or precision

estimates? (eg. P values, Cis)

11. Were the methods (including statistical methods) sufficiently described to enable

them to be repeated?

Results

12. Were the basic data adequately described?

13. Does the response rate raise concerns about non-response bias?

14. If appropriate, was information about non-responders described?

15. Were the results internally consistent?

16. Were the results for the analysis described in the methods, presented?

78

Discussion

17. Were the authors’ discussion and conclusions justified by the results?

18. Were the limitations of the study discussed?

Other

19. Were there any funding sources or conflicts of interests that may affect the

authors’ interpretation of the results?

20. Was ethical approval or consent of participants attained?

Downes, M. J., Brennan, M. L., Williams, H. C. & Dean, R. S. (2016). Development of

a critical appraisal tool to assess the quality of cross-sectional studies (axis).

BMJ Open, 6(12). 10.1136/bmjopen-2016-011458

79

Appendix E – Table of findings for literature review

Table 12.Study characteristics and results relating to research questions.Author(year)country

Title Aims Design Sample Measures Analysis Summary of findings

Jaffe et al.(2003)

USA

Vicarioustrauma injudges: Thepersonalchallenge ofdispensingjustice

1. Rates andtypes of VTsymptomsexperiencedby judges.2.Relationshipbetween VTexperiencesand judges’characteristics (age, workexperience,gender).

QuantitativeCross-sectional.Survey

Conveniencesampling

105 judgesMean age 51(SD=8.1)Gender54.3%Male45.7%Female

Surveydeveloped bythe authors.Open-endedquestionsbased ontraumasymptoms.

Paper testsadministered

ANOVA andchi-squareanalysis used.

Grouping ofVTsymptoms.

Prevalence63% of the participantsreported one or more VTsymptom.

Associated factorsFemales reported moreVT symptoms on averagecompared to males (73%vs. 54% (1, N=104) =3.83, p< .05)More work experiencepredicted more VTsymptoms (F (1, 103) =6.56, p<.05.)Age, sex and experiencereported as possible riskfactors of VT in judges.

Leclerc etal. (2019)

Canada

The unseencost of justice:Post-traumaticstresssymptoms in

1. AssessaveragePTSDsymptomseverity.

QuantitativeCross-sectionalBetween-groupdesign.

Conveniencesampling.476 lawyers.Age:

LEC-5PCL-5HSCL-25WHOQOL-BREF

Two-sidedtests, alphalevel of .05.

Prevalence9% of participants metcriteria for PTSD DSM-5.

Associated factors

80

Canadianlawyers

2. Identifytheproportion oflawyersmeeting thediagnosticcriteria ofPTSD.3. Levels ofpsychologicaldistress andquality oflife,understanding lawyers’wellbeing.

Comparinglawyers withno, moderateand highwork-relatedtraumaexposure.

20-29 (13.7%)30-39 (34.9%)40-49 (24.8%)50-59 (18.3%)60+ (8.4%)Gender :42.4%Male57.6%FemaleEthnicity:Canadian(84.5%)Caucasian(88.9%) livingin Quebec(59%)

Onlinesurvey.

Descriptives(chi-squares,t-tests andANOVA).Post-hocBonferronicontrastanalysis wasconductedbetweengroups.ANCOVAused to testmoderatinghypotheses.

Higher trauma exposurepredicted higher levels ofPTSD symptoms (F (1,471) = 4.79, p=.029,partial η2=.01).Prior trauma historypredicted higher levels ofPTSD symptoms.Females more likely todevelop PTSD symptoms,and had more severesymptoms relative to men.Higher number of weeklywork hours increased therisk of PTSD by 1.89times

Levin &Greisberg(2003)

USA

Vicarioustrauma inattorneys

Exploringlevels of VTin attorneyscompared tomental healthprofessionals.

Quantitative.Cross-sectionalBetween-groupdesign.

Social workers,mental healthprofessionalsand Attorneys

STQ,Itemsassessingburnout.Caseload,personal

Method ofanalysis notreported.

PrevalenceAttorneys showed higherlevels of STS and burnout,compared to socialworkers and mental healthprofessionals.

81

Age: reports“similar age.”Gender: Reports“mostlyfemale.”

trauma andtreatmenthistory.

(no statistical datareported)

Associated factorsLarger caseload predictedhigher scores on STS andburnout (no statistical datareported).

Levin etal. (2011)

USA

Secondarytraumaticstress inattorneys andtheiradministrativesupport staffworking withtrauma-exposedclients.

Examine howcaseload oftrauma-exposedclientspredictshighersymptomload andmediatingvariables. Doattorneyshave moresymptomsthan adminstaff.

Quantitative.Cross-sectionalBetween-groupdesign.

Conveniencesampling

238 AttorneysAge45.72 (SD=11)GenderFemale (55.5%)Male (44.5%)

109 Adminsupport staffAge45.07(SD=11.32)GenderFemale (86.2%)Male (12.6%)

DemographicsTraumaexposureassessmentsIESCES-DSDSProQOL

T-tests

CorrelationsusingBonferronicorrection.

SEMmodelling

Prevalence11% of the lawyers metcriteria for PTSD DSM-5.PTSD symptoms,depression, functionalimpairment, burnout, andSTS were consistentlyhigher among attorneyscompared with adminsupport staff: PTSD (11%vs. 1%), depression(39.5% vs. 19.3%),functional impairment(74.8% vs. 27.5%), BO(37.4% vs. 8.3%), andSTS (34% vs. 10.1%)

Associated factorsBivariate correlationanalysis showed that theIES-r and PROQOL(excluding CS) correlated

82

significantly (p<.001) withhours worked per week(IESr: r=.25-.29,ProQOL5: r=.26-.40) andthe number of trauma-exposed clients (IESr:r=.24-.28, ProQOL5:r=.30-.37).

Participant groupsignificantly explained 7%of the variance in PTSDsymptoms. Work-relatedexposure significantlyexplained 14% of thevariance in PTSDsymptoms. Work-relatedexposure variablesmediated (albeit notexclusively) the attorneys’vulnerability to PTSDsymptoms.

Participant groupsignificantly explained 20% of the variance in STSand burnout scores. Work-related exposure variablesmediated albeit notexclusively attorneys’

83

vulnerability to STS andburnout.

Lustig etal. (2008)

USA

Burnout andstress amongUnited Statesimmigrationjudges

InvestigatedwhetherImmigrationJudges sufferfrom STSand burnout.

QuantitativeCross-sectionalExplorative

Conveniencesampling

96 immigrationjudgesAge: Mean 53(SD=6.55)Gender:Female: 44.8%Male: 55.2%

DemographicsSTSSCBI

Onlinesurvey

Percentageand meanscores.

PrevalenceComparisons with otherprofessional groups usingthe same psychometricmeasures showed thatimmigration judgesexperienced more burnoutthan hospital or prisonstaff (no statisticalsignificance reported).

Associated factorsFemale judges reportedmore STS than malejudges (M=2.50 vsM=1.84, p>.0005).

Maguire &Byrne(2017)

Australia

The law is notas blind as itseems:Relative ratesof vicarioustrauma amonglawyers andmental healthprofessionals

Compares theextent towhichexposure totraumaticinformationaffectshelpingprofessionalsand legalpractitioners.

Quantitative.Cross-sectional.Between-groupdesign.

Conveniencesampling

36 lawyers30 MHP (21psychologist, 9social workers)

Age:18-24 (7%), 25-34(32%) 35-44(17%) 45-54

DemographicsIES-rVTSDASS-21TIPI

T-testsCorrelations

PrevalenceThe VTS score wassignificantly higher forlawyers, (M=39.86,SD=7.81), than for MHPs,(M=33.13, SD=6.96, 95%CI (3.06 to 10.40),t(64)=3.66, p=.001. Largeeffect size (d=.90).IES-R scores for lawyers(mean rank =40.00)significantly higher than

84

Exploreswhetherpersonalitytype caninfluence thelevel of riskto traumareaction.

(23%) 55+(21%)Gender:20% Male80% Female

for MHPs (meanrank=24.83), U=280.00,z= -3.235, p=.001.Medium effectsize(r=.40).Lawyers weresignificantly higher on allmeasures of the DASS-21scales.Depression: Lawyers(M=11.88, SD=4.49) vs.MHPs (M=8.67, SD=2.86,p<.001, d=0.84)Anxiety: Lawyers(M=11.42, SD=4.39) vs.MHPs (M=7.87, SD=1.33,p<.001, d=1.01)Stress: Lawyers (M=14.58, SD=4.80) vs.MHPs (M=11.28,SD=3.18, p<.002, d=0.80)

Associated factorsNo effect of gender.Profession (lawyers vsMHP) accounted for 15%of the variance. This effectwas mediated by enteringthe personality traitEmotional Stability

85

(neuroticism), whichincreased explainedvariance by 20%. Thescores from the remainingfour personality traitsincreased explainedvariance again by 18% intotal. Personal traumahistory added 6% to theexplained variance.

Miller etal. (2018)

USA

Judicial stress:The roles ofgender andsocial support.

Researchquestions:What isjudges’ stresslevel? Dosocialsupport,gender, andotherdemographics relate tostress, health-relatedoutcomes,and job-relatedoutcomevariables?Does gendermoderate the

QuantitativeCross-sectional.Explorative

Conveniencesample

221 judges

Age:30-39 (0.01%)40-49 (14.02%)50-59 (41.62%)60+ (40.72%)

Gender:61.3% Males28.7% Female

Surveydeveloped bythe authorsrelating toperceivedgeneralstress,mental/physical health,jobsatisfaction,socialsupport, anda 13-itemCompassionfatigue scale.

“Clicker”responder tosurvey,

Chi-square

Correlations

Linearregressions

Associated factorsThere were no significantmain effects for gender orsocial support and nointeraction effect on thesecondary traumameasure.There was no main effectfor gender, but there wasan interaction effectindicating a negativerelationship betweensocial support andburnout: increased socialsupport led to lessperceived general stress,but only for male judges(R2=.07, F(3, 136)=3.45,p<.05; interaction term,B=–0.480, p< .05)

86

relationshipbetweensocial supportand stress,health-relatedand job-relatedoutcomes?

presented at aseminar.

Social support was relatedto less perceived stress,less burnout and more jobsatisfaction, but only formales.

Piwowarczyk et al.(2009)

USA

Secondarytrauma inasylumlawyers

Examinationof asylumlawyers andthe impact oftheir workwith traumasurvivors ontheiremotionalwellbeing.

Quantitative.Cross-sectional.Explorative

Conveniencesampling

57 asylumlawyers

Gender:60% maleEthnicity:93% born in theUSA

DemographicsJobsatisfactionand jobstress.LEC-5STSSPerceivedstress scale

Method ofanalysis notreported.

Prevalence87% of the participantshad two or moresymptoms of STS present.9% of participants scoredhigher than 30 on theSTSS scale.

Associated factorsWork hours per week wasassociated with higherlevels of STS (p<.007).

Schreveret al.(2019)

Australia

Thepsychologicalimpact ofjudicial work:Australia’sfirst empiricalresearchmeasuringjudicial stressand wellbeing

To explorethe nature,prevalence,severity andsources ofjudicial stressin Australia.

Mixed-methods study.Explorative

Conveniencesampling125 judges (part1 and 2)Age: <50 (12%),50-59 (38.4%),60-69 (47.2%),>70 (2.4%)Gender:47.2%Female

Part 1.Perceivedstress andsatisfactionBPNSWscaleK-10DASS-21Part 2.IDLS

Percentages,means andnormativecomparison.

PrevalenceThe majority (83.6%) ofjudicial officers endorsedat least one STS symptom,with approximately a thirdscoring in “moderate” to“severe” ranges. Almostone in three judicialofficers reported“problematic” alcohol use.

87

52.8%Male MBI-GSSTSSAUDIT

Interpretation of STSSscores indicated that30.4% of the judicialofficers met DSM-5criteria for PTSD.

Judicial officers’ rates ofpsychological distresswere markedly highercompared to the generalpopulation and barristers,but lower than lawstudents and solicitors (nostatistical significancereported).

Vrelevski& Franklin(2008)

Australia

Vicarioustrauma: Theimpact onsolicitors ofexposure totraumaticmaterial

Toinvestigatethe impact ofworking withtraumatisedclients andtheirtraumaticmaterial onmembers ofthe legalprofession.

Quantitative.Cross-sectional.Between-groups design.Explorative

Conveniencesampling

50 criminallawyers50 non-criminallawyers

Age:Mean 39.7yrs(SD=11.08)Gender:64% Female36% maleEthnicity

DemographicsVTSDASSIES-rTABS

T-testsMann-Whitney test

Authorsdecided touse asignificancelevel of .025.

PrevalenceCriminal solicitorsreported higher scores onVTS (M=41.50, SD=6.36)compared to non-criminalsolicitors (M=26.32,SD=11.18, p<.001).No significant differencebetween groups on IES-rscores.Criminal solicitorsreported higher scores onVTS (M=41.50, SD=6.36)compared to non-criminal

88

73% Anglo-Saxon, 8%European, 4%Middle Eastern,3% Asian (14%not reporting)

solicitors (M=26.32,SD=11.18, p<.001).Criminal solicitorsreported higher scores onelements of the TABS:self-safety (p<.014) othersafety (p<.021) andintimacy (p<.022)compared to non-criminalsolicitors.

Associated factorsDirect trauma affectedVTS scores irrespective ofthe level of traumaexposure at work.

Key: VTS: Vicarious Trauma Scale; DASS-21: The Depression, Anxiety and Stress Scale; IES-r: Impact of Events Scale-revised; TABS:Trauma and Attachment Belief Scale; BPNSW: Basic Psychological Needs Satisfaction at Work; K-10: Kessler Psychological Distress Scale;IDLS: International Depression Literacy Survey; MBI-GS: Maslach’s Burnout Inventory – General survey; STSS: Secondary Traumatic StressScale; AUDIT: World Health Organisation’s Alcohol Use Disorders Identification Test; LEC-5: Life Events Checklist for DSM-5; TIPI: TenItem Personality Measure; CBI: Copenhagen Burnout Inventory: PTSD Checklist for DSM-5; HSCL-25: The Hopkins Symptoms Checklist-25;WHOQOL-BREF: World Health Organisation’s Quality Of Life –BREF; CES-D: Center for Epidemiologic Studies Depression Scale; SDS:Sheehan Disability Scale; ProQOL; Professional Quality of Life Scale.

89

Appendix F*: Guidelines to authors for journal targeted for research project

Background to the Journal: Since its inception in 2007, the iIIRG’s research outlet

was the Bulletin, however, given its success, it was decided to re-name the Bulletin to

make it more focussed on the group’s main activities (e.g., research and practice). As

such, from April 2012, the Bulletin was renamed to: Investigative Interviewing:

Research and Practice (II-RP), and is the official journal of the iIIRG. From March

2013, II-RP will be included in EBSCOhost™ databases.

Submission Guidelines: Investigative Interviewing: Research and Practice (II-RP) is

the official journal of the iIIRG and is a peer-reviewed, professional outlet for members

and non-members of the iIIRG to publish their articles. Given the multi-disciplinary

nature of the Group, the worldwide circulation of this Journal and strong practitioner

focus, a wide range of articles will be considered for inclusion. These may include

individual research papers or practitioner case studies in relation to the following

specialist areas: investigative interviewing of suspects, witnesses or victims; expert

advice to interviewers; interview training and policy; interview decision-making

processes; false confessions; detecting deception; forensic linguistics. The list of topic

areas is purely indicative and should not be seen as exhaustive.

Submission of Articles: All submissions made to Investigative Interviewing: Research

and Practice (II-RP) should be in English and submitted online via www.iiirg.org.

Authors should prepare and upload one version of their article. The article will be dealt

with by the Editor or the Deputy Editor. Each article will be independently read by at

least two referees (blind review process).

Submission of an article to II-RP will be taken to imply that it represents original work

not previously published and is not being considered elsewhere for publication and, if

accepted for publication, it will not be published elsewhere in the same form, in any

language, without the consent of the II-RP editor. It is a condition of the acceptance by

the editor of an article for publication that the iIIRG acquires the copyright of the article

throughout the world. Prior to publication, the author/s will be required to sign and

return a Copyright Agreement before the article can be published in II-RP. The Editorial

team retains the discretion to accept or decline any submitted article and to make minor

amendments to all work submitted prior to publication. Any major changes will be

made in consultation with the author/s.

90

Format of Articles: Articles can be in these formats: (i) practitioner case studies; (ii)

reviews of previous bodies of literature; (iii) reviews of conference or other specialist

events; (iv) opinion papers; (v) topical commentaries and; (vi) book reviews.

Manuscripts should be typed in double-space with wide margins (3 cm). As a general

guide, articles should not exceed 5,000 words, although the Editor retains discretion to

accept longer articles where it is considered appropriate. Please make sure that all

acronyms are clearly defined in brackets the first time they are used.

Title page: This should contain the title of the paper, the name and full

institutional/organisational postal address of each author and an indication of which

author will be responsible for correspondence and proofs (this should also include e-

mail address). Abbreviations in the title should be avoided.

Abstract: This should not exceed 200 words and should be presented on a separate

page (together with the title of the article) summarising the significant coverage and/or

findings.

Keywords: Abstract should be accompanied by five key words or phrases that between

them characterise the contents of the paper. These will be used for indexing and data

retrieval purposes.

Style guidelines: If you are an academic, it is expected that prior to submission your

article will be formatted to the latest standards of the Publication Manual of the

American Psychological Association (APA). If you are not an academic, there is no

requirement for your work to conform to the format standards of the APA, however,

you must reference your article (where appropriate) and the Editor/s will format it prior

to publication (should it be required). Any consistent spelling style is acceptable. Use

single quotation marks with double within if needed. If you have any questions about

references or formatting your article, please contact the Editor, Kirk Luther,

at [email protected].

Figures: All figures should be numbered with consecutive Arabic numerals, have

descriptive captions and be mentioned in the text. Figures should be kept within the

text. It is the author’s responsibility to obtain permission for any reproduction from

other sources.

Preparation: Figures must be of a high enough standard for direct reproduction with all

the lettering and symbols included. Axes of graphs should be properly labelled and

91

appropriate units given. Photographs intended for half-tone reproduction must be high

resolution.

Size: Figures should be planned so that they reproduce to 10.5 cm column width. The

preferred width of submitted drawings is 16 – 21 cm with capital lettering 4 mm high,

for reduction by one-half. Photographs for halftone reproduction should be

approximately twice the desired size.

Tables: Tables should be clearly typed with double spacing and should be kept within

the text. Tables should have consecutive Arabic numerals and give each a clear

descriptive heading. Avoid the use of vertical rules in tables. Table footnotes should be

typed below the table, designated by superior lower-case letters.

Copyright and Author Rights: It is a condition of publication that authors assign

copyright or license the publication rights in their articles, including abstracts, to the

iIIRG BEFORE the article or case study can be published in II-RP. This enables the

iIIRG to ensure full copyright protection and to disseminate the article, and of course

the Journal, to the widest possible readership in electronic format (and occasionally

print) as appropriate. Authors are, themselves, responsible for obtaining permission to

reproduce copyright material from other sources. Download the iIIRG Copyright

Agreement, complete and sign it, then return to the Editor, Kirk Luther, electronically

at [email protected] (please mention the article/case study title in your email so

that it can be easily identified).

Exceptions are made for certain Governments’ employees whose policies require that

copyright cannot be transferred to other parties. We ask that a signed statement to this

effect is submitted when returning proofs for accepted papers.

92

Appendix G*: Ethics letter of approval from UOL, UK

93

Appendix H*: Norwegian ethics approval from Regional committee for medical

and health related research ethics (written in Norwegian)

From: [email protected] <[email protected]>

Sent: Thursday, September 12, 2019 2:26 PM

To:

Subject: Svarbrev

Alle skriftlige henvendelser om saken må sendes via REK-portalen

Du finner informasjon om REK på våre hjemmesider rekportalen.no

Region:

REK midt

Saksbehandler: Vår dato:

12.09.2019

Vår referanse:

7168

Deres referanse:

7168 Empati, resiliens og omsorgstrettet hos etterforskere av overgrep mot barn

Forskningsansvarlig: Norges teknisk-naturvitenskapelige universitet

Søker:

Søkers beskrivelse av formål:

Vold og overgrep mot barn er et prioritert område. For å kunne prioritere dette, må

fokuset rettes mot de som jobber med å etterforske disse sakene. Tidligere forskning har

vist at denne type arbeid kan være helseskadelig, også for ansatte i politiet, som anses

som en gruppe som er spesielt motstandsdyktig når det gjelder stress. Det er begrenset

med forskning på hvilke faktorer som er avgjørende for den totale belastningen i slike

yrker, men studier fra USA, samt foreløpige funn fra en kvalitativ studie fra Trøndelag

politidistrikt (STRESOB), indikerer at arbeidets organisering og støtte fra kollega og

leder er avgjørende. I denne studien vil det fokuseres både på personlige og

organisatoriske ressurser, hos etterforskere med og uten spesialisering i tilrettelagte

avhør av barn. Studien vil inkludere ansatte i politiet i Norge og Storbritannia. Funnene

fra dette forskningsprosjektet vil kunne bidra til optimal organisering av arbeidet med å

etterforske vold og overgrep mot barn.

REKs vurdering

95

Vi viser til søknad om forhåndsgodkjenning av ovennevnte forskningsprosjekt.

Søknaden ble behandlet av Regional komité for medisinsk og helsefaglig

forskningsetikk (REK midt) i møtet 21.08.2019. Vurderingen er gjort med hjemmel i

helseforskningsloven § 10.

Forsvarlighet: Vi har vurdert søknad, forskningsprotokoll, målsetting og plan for

gjennomføring. Under forutsetning av at du tar vilkårene nedenfor til følge vurderer vi

at prosjektet er forsvarlig, og at hensynet til deltakernes velferd og integritet er ivaretatt.

Forbedring av informasjonsskriv

- Opplys hvor lang tid det vil ta å besvare spørreskjemaet.

- Opplys om at studien innebærer et samarbeid med Storbritannia.

Prosjektets varighet

Prosjektstart og prosjektslutt er i søknaden registrert med samme dato. Vi ber deg om å

sende oss en prosjektendring der du spesifiserer hva sluttdatoen skal være.

Vilkår for godkjenning

1. Revidert informasjonsskriv skal sendes komiteen til vurdering. Vi vil opprette en

oppgave til deg for dette i portalen. Du vil motta en epost fra oss når oppgaven er

opprettet.

2. Komiteen forutsetter at ingen personidentifiserbare opplysninger kan framkomme

ved publisering eller annen offentliggjøring.

3. Komiteen forutsetter at du og alle prosjektmedarbeiderne følger institusjonens

bestemmelser for å ivareta informasjonssikkerhet og personvern ved innsamling, bruk,

oppbevaring, deling og utlevering av personopplysninger.

4. Av dokumentasjonshensyn må du oppbevare opplysningene i 5 år etter prosjektslutt.

Opplysningene skal oppbevares avidentifisert, dvs. atskilt i en nøkkel- og en datafil.

Opplysningene skal deretter slettes eller anonymiseres.

Vedtak

Godkjent med vilkår

96

Appendix I*: Participant information sheet

97

Appendix J*: Online consent form

99

Appendix K*: Complete online questionnaires, with public measures.

100

References

Baron-Cohen, S., & Wheelwright, S. (2004). The empathy quotient: An investigation of

adults with Asperger syndrome or high functioning autism, and normal sex

differences. Journal of Autism and Developmental Disorders, 34(2), 163–175

Eisenberger, R., Huntington, R., Hutchison, S., & Sowa, D. (1986). Perceived

organizational support. Journal of Applied Psychology, 71(3), 500-507.

Eisenberger, R. (2019) Perceived Organizational Support.

http://classweb.uh.edu/eisenberger/perceived-organizational-support/ Retrieved

on 12th February 2019


Zigmond, A.S. & Snaith, R.P. (1983). The Hospital Anxiety and Depression Scale. Acta

Psychiatrica Scandinavica. 67(6)

110

Appendix L: Exploration of missing dataThe missing gender data was explored using a one-way between subjects ANOVA,

coding data as male, female and missing. The effect of gender was explored to compare

means of CS, STS, burnout, anxiety, depression, empathy and perceived organisational

support. There was no significant effects of gender with an alpha value set at p=.01; CS

(F(2,89)=.094, p=.911), STS (F(2,89)=.012, p=.988), burnout (F(2,89)=2.499, p=.088),

anxiety (F(2,89)=.632, p=.534), depression (F(2,89)=.156, p=.856), empathy

(F(2,89)=.462, p=.632) and perceived organisational support (F(2,89)=.193, p=.825).

These data suggest that the missing gender data did not significantly differ from the data

reported as male or female.

111

Appendix M: Evaluation of assumptions.

After missing data were addressed, data were then explored to determine whether

statistical assumptions for parametric analyses were met. Normality was assessed by

checking histograms, Q-Q plots, box plots, skewness and kurtosis values. No significant

findings were found in skewness, but there were significant findings of negative and

positive values for kurtosis. Several outliers were identified in box plots and histograms.

The Kolmogorov-Smirnov test, which is preferred to Shapiro-Wilk because N exceeded

50 (Clark-Carter, 2010), further indicated non-normality. For t-tests, Levene’s statistics

were also referred to for each of the comparative variables.

Assumptions were explored for bivariate correlations, to examine linearity,

homoscedasticity and type of data scales (interval and ratio). Gender data was included

as nominal data, as the data was coded in two levels. Scatterplots indicated that

homoscedasticity was present. Given violations of normality and linearity, a

Spearman’s rho correlation was chosen.

To undergo linear regressions, assumptions of multivariate normality,

multicollinearity and homoscedasticity were explored. Two outcome variables were

found to be normally distributed. A multiple regression was completed for the CS

model with three predictor variables (anxiety, EQ and SPOS) and a Simple linear

regression for the STS model with one predictor (anxiety). Both models met all

assumptions when examining multicollinearity, P-P plots, scatterplots, std. residuals and

Cooks distance. The outcome variable burnout with two predictor variables (anxiety and

depression), failed the assumption of normality. Other assumptions were met when

examining standard residuals, and ANOVA indicated significant linear relationship.

112

Appendix N: Q-Q plots and Normality statistics.Table 13Test of Normality and assumptions statistics

Kolmogorov-Smirnova Skewness KurtosisStatistic df Sig. Statistic Statistic

PROQOL_CS ,086 92 ,087 -.304 .118PROQOL_BO ,114 92 ,005 -.337 1.497PROQOL_STS ,070 92 ,200* .189 -.490HADS_Depression ,196 92 ,000 -.762 .294HADS_Anxiety ,114 92 ,005 -.384 .385CBS_EQ ,047 92 ,200* .080 -.135SPOS_total ,056 92 ,200* -.277 -.207*. This is a lower bound of the true significance.a. Lilliefors Significance Correction

Figure 2. SPSS output of Normal Q-Q Plot for ProQOL Compassion Satisfaction

113

Figure 3. SPSS output of Normal Q-Q Plot for ProQOL Burnout

Figure 4. SPSS output of Normal Q-Q Plot for ProQOL Secondary Traumatic Stress

114

Figure 5. SPSS output of Normal Q-Q Plot for HADS depression

Figure 6. SPSS output of Normal Q-Q Plot for HADS Anxiety

115

Figure 7. SPSS output of Normal Q-Q Plot for ProQOL Cambridge Behaviour Scale –

Empathy Quotient

Figure 8. SPSS output of Normal Q-Q Plot for Survey of Perceived Organisational

Support.

116

Appendix O*: Checklist assuring confidentiality and anonymity of participants.

Checked in

Executive

Summary/

Abstract/ Overview

(If included in

assignment)

Checked

in main

text

Checked in

appendices

Pseudonym or false initials used

N/A N/A N/A

Reference to pseudonym/false

initials as a footnote N/A N/A N/A

Removed any reference to names of

Trusts/ hospitals/ clinics/ services

(including letterhead if including

letters in appendices)

Removed any reference to names/

specific dates of birth/ specific date

of clinical appointments/

addresses/location of client(s),

participant(s), relatives, caregivers,

and supervisor(s). For research thesis

– supervisors can be named in the

research thesis “acknowledgements”

section.

Removed/altered references to

client(s) jobs/professions /nationality

where this may potentially identify

them. [For research thesis – removed

potential for an individual research

participant to be identifiable (e.g., by

a colleague of the participant who

might read the thesis on the internet

and be able to identify a participant

117

using a combination of the

participants specific job title, role,

age, and gender)]

Removed any information that may

identify the trainee (consult with

course staff if this will detract from

the points the trainee is making)

N/A N/A N/A

No Tippex or other method has been

used to obliterate the original text –

unless the paper is subsequently

photocopied and the trainee has

ensured that the obliterated text

cannot be read

The "find and replace" function in

word processing has been used to

check the assignment for use of

client(s) names/other confidential

information

118

Appendix P*: Chronology of research process

Dates Research activity

January 2018 Research proposal form submitted to University.

Research supervisor allocated.

May 2018 Initial research proposal submitted.

June 2018 Panel review meeting of initial research proposal.

July –

November 2018

Development of research proposal

December 2018 University peer review process completed.

February 2019 Service user review completed

May 2019 Ethical approval and sponsorship from University of Leicester for

English data.

October 2019 –

March 2020

Recruitment for empirical study.

September 2019 Ethical approval from REK in Norway.

October 2019 1st draft of systematic literature review.

February 2020 Final draft of systematic literature review

May 2020 Received data from Norway

June 2020 Thesis submitted to University of Leicester (04.06.2020)

July 2020 Research viva

July –

September 2020

Preparation and submission of work for publication

119

Appendix Q*: Statement of epistemological position taken

The researcher adopted a positivist epistemological position for both the literature

review and the research report. Positivism considers research as an objective empirical

truth that can be scientifically verified, preferably by making use of logical and

mathematical proof (Cohen et al., 2013). Variables are considered as objective

constructs, which are measured using validated measurable methods. The proposed

research questions and chosen methodological measures and procedure considered

quantifiable data, where statistical analyses were used to inform interpretation and

drawing conclusions.

Prior to carrying out this project, the researcher had limited knowledge and

experience in the field of occupational distress and secondary trauma, both regarding

the legal professions and law enforcement. As such, the researcher was perceived to

have an objective viewpoint to the research findings and holding limited prior

conceptions. In addition, the research was kept independent from any data-collection or

possible participants who completed self-reporting questionnaires via an online survey.

Although a positivistic stance can be helpful to identify and quantify reported or

observable trends in a population, some limitations of the model are still acknowledged.

The empirical nature of the positivistic model can limit potential understanding of

personal experience and meaning-making which is not specifically captured by the

measures.

Cohen, L., Manion, L., & Morrison, K. (2013). Research methods in education. New

York: Routledge.

120

empathy, resilience and distress in norwegian and english

Documents