Handheld computers for self-administered sensitive data collection: a comparative study in Peru.

Area of Science:

• Public Health
• Medical Informatics
• Survey Methodology

Background:

• Low-cost handheld computers (Personal Digital Assistants - PDAs) show potential for efficient sensitive data collection in surveys.
• Evaluating the data quality from PDAs against traditional paper questionnaires is crucial for validating new technologies.

Purpose of the Study:

• To assess the quality of sexual behavior data collected using PDAs compared to paper-based questionnaires.
• To determine the feasibility and reliability of PDA-based data collection in a developing country context.

Main Methods:

• Developed a PDA data collection program using open-source tools.
• Conducted two cross-sectional studies in Ancon (Lima), comparing PDA-collected sexual behavior data with paper questionnaire data.
• Analyzed agreement and correlation for categorical and numerical variables, and assessed response rates, inconsistencies, and missing values.

Main Results:

• General agreement between paper and PDA data was 86%, with higher agreement for categorical variables (Kappa: 0.43-0.86) than numerical variables (Spearman: 0.76-0.95).
• Agreement was higher among participants with at least a high school education.
• PDA use resulted in significantly fewer inconsistencies and missing values compared to paper questionnaires.

Conclusions:

• PDA-based systems are valuable for collecting sensitive data, achieving high agreement with paper methods.
• PDAs offer a feasible, low-cost alternative for field data collection in developing nations, improving data accuracy.
• The study demonstrates the successful implementation of a low-cost PDA application for reliable data gathering.