Differential recall in a case-control study of child pedestrian injuries

I conducted a validation study to examine data collected in a case-control study of child pedestrian injuries for evidence of differential recall. I compared questionnaire responses for variables related to the safety of the home environment with direct observations of the same parameters. A comparison of all child pedestrian injury cases with all controls provided no evidence of differential recall. There was, however, differential recall for a subset of the cases, specifically those injured in residential driveways. For these cases, differential recall had a substantial effect on the odds ratio.