Inconsistencies in children's recall of witnessed events: The role of age, question format and perceived reason for question repetition


Correspondence should be addressed to Dr Pauline Howie, School of Psychology A18, University of Sydney, Sydney, New South Wales 2006, Australia (e-mail:


Purpose. Children's inconsistencies when answering repeated questions about past events are a source of concern in forensic, educational, and other contexts. Theoretical accounts of these inconsistencies have predominantly assumed that children shift because they infer adult dissatisfaction with their initial answer. This study aimed to test this account via examination of the effects of question format on shifting, as well as via direct questioning of children.

Method. Four-, five-, and seven-year-olds (N = 226) were asked 17 recall questions about a recent classroom activity, with eight target questions repeated in one of four formats: no-correct (mildly misleading questions to which the correct answer was ‘no’), yes-correct (mildly misleading questions to which the correct answer was ‘yes’), specific open wh- questions, and forced-choice questions. They were then asked about the adult's reasons for repetition. Accuracy, shifting, and interpretations of question repetition were examined.

Results. Shifting from accuracy decreased with age, and was affected by question format in 4-year-olds only, who shifted more to no-correct than to forced-choice questions. Shifting towards accuracy was more common in forced-choice questions than either no-correct or open questions, but there were no significant age differences. First-answer-unsatisfactory interpretations of question repetition were rare, especially in the two younger groups. The oldest group offered a wider range of interpretations and also showed some evidence of an association between first-answer-unsatisfactory interpretations and shifting.

Conclusions. Overall, our findings present a challenge to first-answer-unsatisfactory explanations of young children's shifting in recall settings, at least where overt pressure to shift is low. Forensic implications are considered.