1. Recorded conversations between parents and children when suspecting child sexual abuse Iiirg 2011, Dundee Julia Korkman, PhD Helsinki University Hospital & Åbo Akademi University Co-authors: Aino Juusola, Jason J. Dickinson, Katarina Finnilä-Tuohimaa and Pekka Santtila

2. Introduction Often, allegations of child sexual abuse (CSA) and other forms of child maltreatment come about through the child telling a close person about the events However, how parents discuss with children when suspecting CSA is scarsely researched

3. Introduction Bruck, Ceci & Francoeur, 1998: Mothers’ memories of conversations with their 4-year- old children: – The mothers remembered the meaning of the discussions better than the wording – They did not recall whether information had come about spontaneously or not – Warnings to pay attention to the kind of questions used did not improve their performance

4. Aim of the present study To analyse a sample of recorded conversations between parents and their children – The conversations had been delivered to the police by the parents, most likely in order to give further evidence for the alleged abuse  Analyse the strategies used by parents to retrieve information from their children when suspecting child abuse  Expectations: the interviews were expected to be leading, in view of previous research on formal interviews conducted by untrained interviewers (Lamb et al., 2000; in Finland Korkman, 2006; Korkman et al., 2008)

5. Method Collecting recorded parent-child discussions from the police Transcribing the discussions and coding them in terms of: – Question types (Lamb et al., 2000) – Information provided and whether it was introduced into the interview by the interviewer or the child – Positive / negative feedback by the parent, and – Type of suspicion

6. Sample 20 cases were coded, of which one was excluded as it was not clear if it was related to an abuse allegation Evident by qualitative analyses that the majority of the cases were part of custody disputes The ”interviewers” were parents, step parents and foster parents – 32% one male parent, 47% one female parent and in 21% one male and one female The children were 9 boys and 10 girls, aged 2 – 8 years (mean age 5,4 yrs)

7. Distribution of question types

8. New information provided within the interview When looking at the new, allegation-related information, the parent provided the large majority of the information: – In 13/19 (68%) of the interviews, all new information was provided (first) by the parent – In 4/19 (21%) all new information was provided by the child – In 2 interviews (11%) both the child and the parent provided new information

9. The association between question types and new information The type of question used was significantly associated with the child providing new information X² (5) = 33.76, p <.001

10. The information provided by the children were given in response to the following question types Question typenResulted in % of all the info provided by the children Open-ended61 Facilitators13812 Focused20722 Option-posing23830 Suggestive34935  65% of the information provided came in response to leading and suggestive questions

11. Positive/negative feedback The child being unresponsive was associated with the interviewer using more positive/ negative feedback (X²(1)=34.20, p <.001) In 20% of the cases where the child had not provided any abuserelated information, the parent put pressure on the child to do so, whereas they did so only in 8% of the cases (mostly thanking)

12. Variation between the discussions? Dynamics highly similar in most cases Parents in almost all recorded conversations relied heavily on leading prompts Open-ended questions present in only four cases One exceptional case: only open-ended, facilitative and focused questions were used in this very short interaction – This was the only case where a conviction was made

13. Discussion In roughly 70% of the interviews, the child did not provide any new information whatsoever, yet the parents seemed to think the recorded conversation lended support to the abuse allegation The information that the children did provide came mostly in response to leading and suggestive questions While Bruck et al (1998) found that mothers did not remember how the information was elicited from the children, the results of the present study might indicate that the parents do not even recognise the difference btw spontaneous and coerced utterances in the first place

14. Practical relevance? Very often child abuse is recognised through children telling their parents: all parental reports should not be discredited – but the present results call for (yet more) caution – Possible motives or strong preconceptions by the caretakers should be carefully adressed (confirmation bias  suggestion) – When collecting information from caretakers, warning signs could be ”I had long suspected” or ”I clearly felt there was something more he wanted to tell me”

15. Practical relevance? When planning an interview on the basis of what the child allegedly has told an adult, the (formal) interviewer should plan the interview so that the possibility of misunderstandings / manipulation is taken into consideration: ”I heard you told mum/dad that...”  ”I heard you and your mum/dad talked about... – Tell the child adults may misunderstand or that it is unclear what really has happened, and that the interviewer would like the child to tell what has happened according to him/her

16. Limitations The sample is very small The vast majority of the cases were related to custody disputes (i.e., are perhaps not representative of all parents suspecting abuse) There may have (and in some cases clearly had) been discussions prior to recording

17. Future research Experimental study with a ”typical” conversation, based on the present sample  let a nr of randomly assigned (i.e., ”normal”) parents listen and assess what has been told and how (spontaenous vs. coerced) the childrens’ accounts are  Similarily for judges Comparative real-case samples from other countries!

18. Thank you! Contact: julia.korkman@hus.fi