Citation

Boekeloo BO, Schamus LA, Simmens SJ, Cheng TL. Am. J. Prev. Med. 1998; 14(3): 209-216.

Affiliation

The George Washington University Medical Center, Institute for Health Policy, Outcomes, and Human Values, Center for Health Promotion and Disease Prevention, Washington, DC 20037, USA.

Copyright

(Copyright © 1998, Elsevier Publishing)

DOI

unavailable

PMID

9569222

Abstract

INTRODUCTION: Difficulty in measuring sensitive behaviors in 12-15-year-old adolescents is a barrier to research. This study determined whether early adolescents reported substance use and sexual activity similarly in assisted paper-and-pencil versus touch-tone telephone responses. METHODS: Adolescents 12-15 years old completed confidential, interviewer-assisted questionnaires first in a physician office by paper-and-pencil and then at home by touch-tone approximately 3 months later. Adolescents were from a high-risk urban area, 71% were minority, and all had parent consent to participate. RESULTS: The follow-up participation rate was 94% (follow-up n = 207). Test-retest stability was generally poor for low-frequency behaviors such as injection drug use, anal intercourse, and sexual behaviors in 12-13-year-olds. Test-retest stability was fair to good for common substance use items. Test-retest stability was generally good among females and 14-15-year-old adolescents, and poor to fair among males and 12-13-year-olds, for common sexual experiences in the last 3 months. Test-retest stability was generally good to excellent for all lifetime sexual experiences except among 12-13-year-olds in which it was generally poor. Internal consistency of the self-esteem scale was high using both response technologies. Both response technologies reproduced correlations between substance use and lifetime sexual experience. CONCLUSION: A high participation rate and reliable data capture were achieved when assessing sensitive behaviors of 14-15-year-olds using touch-tone telephone response. Sexual behaviors were more reliably captured using a "lifetime" versus "last 3-month" reference period. Low prevalence contributed to poor reliability in 12-13-year-olds.

Language: en