Back to Annual Meeting

Back to Annual Meeting

APHA Scientific Session and Event Listing

Erika S. Trapl, MS¹, Elaine A. Borawski, PhD¹, H. Gerry Taylor, PhD, ABPP/CN², LaTylia Clemmons¹, and Danyel Savarda¹. (1) Epidemiology and Biostatistics, Case Western Reserve University, Center for Health Promotion Research, 11430 Euclid Ave., Cleveland, OH 44106, 216-368-0098, est2@case.edu, (2) Pediatrics, Case Western Reserve University, 10900 Euclid Avenue, Cleveland, OH 44106

While the most common method for collecting school-based surveillance or evaluation data is the paper-pencil self-administered questionnaire, there are many limitations to this method. An alternative option is the handheld personal digital assistant (PDA), which has many advantages of computer-assisted self-interviewing (CASI), but is cheaper and more portable than desktop or laptop based systems. However, recent studies (Trapl et al, 2005) have found that these methods could reduce data quality (i.e., reliability) due to the possible masking of reading or attention issues.

This presentation summarizes the results of a NICHD-funded study examining the impact of data collection mode, reading comprehension, and behavioral issues (e.g. inattentiveness) on data quality. Students attending seven urban middle schools were randomly assigned to complete a health survey by one of three modes: paper-pencil; PDA; or audio-enhanced PDA. The health survey asked students about both sensitive (e.g. sexuality) and non-sensitive (e.g. physical activity) topics. Students completed a paper-based debriefing survey and a reading comprehension assessment. Teachers completed behavioral and academic assessments for each student. Survey administrators completed a testing environment checklist and recorded students' start and end time.

Outcomes included inconsistent responses and skip patterns (paper-pencil only), extent of missing data, as well as construct validity across groups and differences in reporting of sensitive and non-sensitive behaviors. Direct relationships of mode, reading comprehension, and behavioral issues to data quality will be discussed. Further, we will discuss how the relationships of reading comprehension and outcomes and behavioral issues and outcomes are moderated by mode.

Learning Objectives:

• Identify characteristics of school-based data collection methods used with adolescents in the collection of sensitive information
• Recognize intrinsic and extrinsic factors that can impact an adolescent’s ability to complete a survey
• Evaluate the implications of survey computerization and audio enhancement on data quality outcomes

Keywords: Adolescents, Survey

Presenting author's disclosure statement:

Not Answered