Gavin E' Oxburgh

1
Police Officers Use of Emotional Language During
Child Sexual Abuse Investigations

• Gavin E. Oxburgh
• BSc (Hons), MSc
• Acknowledgements Dr James Ost, Dr Tom Williamson
  Dr Clare Wilson

International Centre for Research in Forensic
Psychology
2
Research Background

• What were the emotional effects of interviewing a
  child victim prior to interviews with offenders?
• Were there gender differences in the use of
  emotional words by interviewers during child
  abuse enquiries?
• Would abuse type have any effect on the offender
  interview regarding emotional words used ?

International Centre for Research in Forensic
Psychology
3
Research Background

• Numerous guidelines Acts on interviewing child
  witnesses (e.g. Memorandum of Good Practice
  Achieving Best Evidence, Vulnerable Witnesses
  (Scotland) Act) but.
• Limited training on interviewing sex offenders
• No known research examining emotional language by
  police officers who interview sex offenders
• Personal-professional dialectic?

International Centre for Research in Forensic
Psychology
4
Research Hypotheses

• (1) Officers who had previously interviewed
  victims would use more emotional words than those
  who had not interviewed the child victim.
• (2) Emotional words used would vary as a function
  of the gender of the interviewer.
• (3) Emotional words used would vary as a function
  of the type of offence

International Centre for Research in Forensic
Psychology
5
Theoretical Model Used
Ekman (2003) 7 basic emotional states

• Sadness
• Anger
• Disgust
• Contempt
• Fear

• Surprise
• Happiness

POS
NEG
International Centre for Research in Forensic
Psychology
6
Research Methodology

• Only Service Police and HD Forces in England
  Wales Utilised (N 45)
• Transcripts of child sex offender interviews
• Officers had to be regularly employed on child
  sexual abuse investigations
• Additional Information

International Centre for Research in Forensic
Psychology
7
Participant Response Rate

• 50 transcripts made available for analysis of
  which 34 were deemed suitable (17 for each
  condition)
• However, 3 transcripts excluded due to small
  number of cases (intra extrafamilial abuse)
• 31 used for analyses
• 20 male and 11 female interviewers
• All interviewees were male

International Centre for Research in Forensic
Psychology
8
Transcript Analyses

• Only words/phrases uttered by main interviewer
  were analysed
• If more than one interview of offender took
  place, with different officers, transcripts were
  split and treated as separate interviews

International Centre for Research in Forensic
Psychology
9
Transcript Analyses
Contempt 26 (2) Anger 23 (1) Disgust 15
Sadness 0 Fear 0 Surprise 10
(10) Happiness 0 (1)
Not used for analyses
Total Utterances 75 (14)
Inter-Rater Reliability 90
International Centre for Research in Forensic
Psychology
10
Transcript Analyses (cont)
Disgust Well, the whole thing is disgusting
isnt it? and you have had sex, not only with
your niece, but a 15 year-old girl. Anger Im
not here to answer your questions Contempt
You get off doing this do you? and Youre not
suggesting it was an accident?
International Centre for Research in Forensic
Psychology
11
Results
(1) Previous interview with child victim Total
NES for officers who interviewed victims prior to
offender interviews (M .86, SD 1.40) compared
to total NES for officers who had not interviewed
victims prior to offender interviews (M 3.18,
SD 3.61). Significant effect in the opposite
direction to prediction (t(29)-2.323, plt.05)
International Centre for Research in Forensic
Psychology
12
Results (cont)
(2) Gender of the interviewer (Male M 1.75
Female M 2.63) No significant effect
(t(29)-.788, pgt.05)
(3) Type of abuse (intra-, extrafamilial) (Intra
M 1.53 Extra M 3.38) No significant effect
(t(29)-.832,pgt.05)
International Centre for Research in Forensic
Psychology
13
Tentative Conclusions

• Overall, low frequency of emotional utterances
  made, nevertheless
• Evidence of an effect of previously interviewing
  child victim on negative emotional utterances
  made during offender interviews (e.g. less
  Disgust Contempt?)
• No significant evidence of gender differences
• No significant effect of abuse type

International Centre for Research in Forensic
Psychology
14
Limitations of Study

• Only interview transcripts used for analysis
• Audio/video tapes would have been useful

• Unable to establish officers previous life
  experiences
• Effects of traumatising events
• Personal attitudes, views, expectations
  pre-conceived ideas about child sex offenders
• Personality type

International Centre for Research in Forensic
Psychology
15
Limitations of Study (cont)

• Small sample size
• Access to additional material required
• Gender of victim unknown in some cases
• Would number of utterances used vary as a
  function of the gender of victim?
• Emotional words uttered may not be representative
  of emotional feelings by officers
• Deliberate strategy to provoke?

International Centre for Research in Forensic
Psychology
16
Future Research

• Compare typed vs. audio recordings of interviews
  with alleged offenders
• Compare interviews with offenders of child sexual
  abuse vs. other serious crimes to establish
  extent of emotional language used
• Study quality of interviews with offender cohort
  to establish if emotional utterances used poor
  quality interview

International Centre for Research in Forensic
Psychology
17
Current Research

• Study quality of interviews with offender
  cohort to establish if higher NES poor
  quality interview?

Effectiveness of an interview should be measured
by the quantity quality of accurate reliable
information that is obtained without prejudice to
the rights of the interviewee (Towards an Agreed
Statement of Principles Concerning Investigative
Interviewing iii2, July 2006)
International Centre for Research in Forensic
Psychology
18
Current Research (cont)
But what is quality effectiveness and how
can they be measured?
Various studies Baldwin (1992) Stockdale
(1993) Williamson (1993) McGurk et al (1993)
Bull Cherryman (1995) Milne Bull (1999)
Cherryman (1999) Clarke Milne (2001) Lamb et
al (2002a, b, c) Griffiths Milne (2006)
Wetherall et al (2006) Most have looked at the
effectiveness of P.E.A.C.E and/or questioning
techniques
International Centre for Research in Forensic
Psychology
19
Current Research (cont)
None have looked at the use/effect of emotional
language, during investigative interviews of
suspected sexual offenders and we know very
little about how the use of negative emotional
utterances affects the overall interview.
International Centre for Research in Forensic
Psychology
20
Current Research (cont)

• Need to study (amongst other things!)
• Context/extent of emotional language used
• Social construction of interview (see Bartlett,
  1932 Baldwin, 1993) to look at
• Behaviour of interviewer
• Behaviour/response of interviewee

International Centre for Research in Forensic
Psychology
21
Thank you!
gavin.oxburgh_at_port.ac.uk
International Centre for Research in Forensic
Psychology