Reducing sensitive survey response bias in research on adolescents: a comparison of web-based and paper-and-pencil...

Using the Problem Oriented Screening Instrument for Teenagers (POSIT), the effect of mode of administration on (1) students' willingness to disclose sensitive information and (2) response rates was investigated.

A 2 × 2 unequal N factorial analysis of variance (ANOVA) design was employed. Mode of administration (paper-and-pencil vs. Web-based) was crossed with grade level (middle vs. high school).

The study was conducted in two middle and two high schools.

A total of 628 middle and high school students completed the survey.

The POSIT is a self-report measure with 139 yes/no items that identifies stressors in 10 functional areas (e.g., Substance Use).

An unequal N 2 (mode) × 2 (grade level) factorial ANOVA was employed.

No statistically significant differences were found for self-reported risk across modes of administration. Students completing the Web-based version of the survey were four times more likely to skip an item.

Effect of Mode on Reporting of Sensitive Information-Students completed the Web-based surveys in computer labs with other students. The intent of the Web-based survey was to increase perceived privacy but the environment likely negated any effect. Effect of Mode on Response Rates-The higher response rate for the paper-and-pencil survey was the opposite of what was expected and revealed that students were more likely to skip sensitive items on the Web survey.