A study was conducted to compare the psychometric qualities of two identical forms of a survey administered in paper-and-pencil format and in Web format to 197 teachers randomly split into two paper and Web groups. While the rates of response were quite discrepant, the respective measures of reliability were extremely similar for the two versions of the survey. A followup study was conducted with the nonrespondents in the Web group to investigate reasons behind their decision not to complete the survey. Several methodological issues are raised as a result of the followup. (Contains 3 tables and 16 references.) (SLD)