Are reporting errors due to encoding limitations or retrieval failure? Surveys of child vaccination as a case study

Surveys of childhood vaccinations are often highly inaccurate, due to paren tal misreporting. We conducted three experiments to examine the source of t he inaccuracies. In Experiment 1, we provided parents with memory aids; the se aids did little to improve reporting accuracy. Two further experiments a sked whether parents forgot what they know about their children's vaccinati ons, or whether they never knew the information. In Experiment 2 we surveye d parents both immediately and ten weeks after their child's medical visit. Accuracy was only slightly better than chance immediately afterwards; ten weeks later performance had not changed significantly. Experiment 3 compare d reports in both recall and recognition conditions. Although the recogniti on condition lowered the response burden on parents it did not produce more accurate reports. We conclude that low levels of accuracy in parental repo rts on vaccinations appear to reflect poor initial encoding rather than ret rieval failure. Copyright (C) 1999 John Wiley & Sons, Ltd.