Mapping repeated interviews

The present study introduces an adaptation of the Griffiths Question Map (GQM; Griffiths & Milne, 2006) which extends the chronological, visual map of question types used in an investigative interview to include child interviewee’s responses (through the addition of the Interview Answer Grid, IAG). Furthermore, it provides a rare evaluation of repeated interviews with children. From a sample of transcripts of Scottish repeated interviews with child victims, two ‘good’ and two ‘poor’ first interviews were chosen based on interviewer question types. First and second investigative interviews of these four children were mapped using the GQM and IAG in order to examine across the two interviews the similarity of interviewer and interviewee behaviours and the consistency and investigative-relevance of information provided. Both ‘good’ and ‘poor’ interviews were found to include practices discouraged by interviewing guidelines, which would not have been identified by examining question proportions alone. Furthermore, ‘good’ first interviews were followed by second interviews which began with poor question types, suggesting a possible impact of confirmation bias. Social support was also assessed and found to be used infrequently, mainly in response to the child being informative rather than pre-emptively by interviewers in an attempt to encourage this. Children were also found to disclose throughout their second interviews, suggesting that rapport-maintenance is vital for single and multiple interviews. The use of the GQM and IAG are encouraged as techniques for determining interview quality.

The present study introduces an adaptation of the Griffiths Question Map (GQM; Griffiths & Milne, 2006) which extends the chronological, visual map of question types used in an investigative interview to include child interviewee’s responses (through the addition of the Interview Answer Grid, IAG). Furthermore, it provides a rare evaluation of repeated interviews with children. From a sample of transcripts of Scottish repeated interviews with child victims, two ‘good’ and two ‘poor’ first interviews were chosen based on interviewer question types. First and second investigative interviews of these four children were mapped using the GQM and IAG in order to examine across the two interviews the similarity of interviewer and interviewee behaviours and the consistency and investigative-relevance of information provided. Both ‘good’ and ‘poor’ interviews were found to include practices discouraged by interviewing guidelines, which would not have been identified by examining question proportions alone. Furthermore, ‘good’ first interviews were followed by second interviews which began with poor question types, suggesting a possible impact of confirmation bias. Social support was also assessed and found to be used infrequently, mainly in response to the child being informative rather than pre-emptively by interviewers in an attempt to encourage this. Children were also found to disclose throughout their second interviews, suggesting that rapport-maintenance is vital for single and multiple interviews. The use of the GQM and IAG are encouraged as techniques for determining interview quality.