This report is concerned with the extent of potential bias introduced into the 2003 U.S. study through nonresponse on the part of schools and students. Three bias analyses were conducted to address the issues of school nonresponse, student nonresponse, as well as possible achievement differences between students in the spring and fall testing sessions. The analysis compares selected characteristics likely to reflect bias in participation from participating and nonparticipating schools. The results suggest that there is little potential for nonresponse bias in the PISA original participating sample based on the characteristics studied. The results also suggest that, based on the characteristics studied, there is some potential for nonresponse bias in the PISA student sample, using base weights.
Online Availability
