Application of computer-assisted interviews to sexual behavior research.

Area of Science:

• Public Health
• Epidemiology
• Behavioral Science

Background:

• Collecting sensitive data is crucial for understanding health behaviors.
• Traditional face-to-face interviews (FTFI) may be subject to social desirability bias.
• Computer-assisted interview techniques offer potential for improved data accuracy.

Purpose of the Study:

• To compare the validity of sensitive behavioral data collected via video-enhanced, computer-assisted, self-administered interviews (V-CASI) versus FTFI.
• To assess the impact of interview method on reporting of socially desirable and undesirable behaviors.
• To evaluate differences in statistical modeling outcomes based on data collection method.

Main Methods:

• A randomized crossover design was used, comparing V-CASI and FTFI.
• Participants were women diagnosed with Chlamydia trachomatis infection.
• Eight close-ended behavioral questions (socially undesirable, desirable, neutral) were administered.

Main Results:

• Kappa scores indicated good-to-excellent agreement between V-CASI and FTFI.
• Women reported socially undesirable behaviors more frequently using V-CASI.
• Logistic regression models for condom use yielded different results depending on the interview method.

Conclusions:

• V-CASI can reduce social desirability bias in sensitive sexual behavior research.
• V-CASI demonstrates potential for improving the validity of collected data.
• The choice of data collection method impacts research findings and requires careful consideration.