Generally, the study found that repeating questions had a negative effect on the accuracy and consistency of children's responses. Results indicated that when a question was asked a second time, the number of accurate responses declined; however, the use of multiple repetitions beyond the first repetition did not lead to cumulative inaccuracy and inconsistency. This suggests that the response to the initial question is most likely to be accurate. The first repeating of the question may imply to the child that his/her response to the initial question was unacceptable, leading to a revision of his/her most immediate and most accurate recollection. This alteration is most likely to be maintained even if the question is repeated again and again. This consistency, however, was not associated with accuracy. The study found that the differing intervals between a repeated question had almost no influence on the accuracy of children's responses. All repeated questions in the study were based on the meaning of the initial question, but using different words. This is the form of question repetition used most often in police interviews. Study participants were 156 children ages 4-9, who watched a staged event followed by individual interviews 1 week later. They were asked eight open-ended questions that were each repeated an additional four times. Half of these open-ended questions could be answered from information in the event; and half were unanswerable, meaning the children should have said "don't know" in response to these questions. The questions were then repeated in a reworded form with varied frequency and intervals. 1 table and 50 references