A comparison of paper-based questionnaires with PDA for behavioral surveys in Africa: Findings from a behavioral monitoring survey in Kenya

Simon Pierre Tegang, Gideon Emukule, Samuel Wambugu, Inoussa Kabore, Peter Mwarogo


Background: The concept of using personal data assistant (PDA) software for survey data collection is not new. However, usage is not widespread, especially in sub-Saharan Africa where paper-based questionnaires remain the main data collection tool. In 2007, FHI1 conducted a behavioural survey among high-risk groups in three provinces of Kenya. To assess the strengths and weaknesses of different data collection methodologies, paper questionnaires were used in one province and PDAs were used in two others.

Methodology: The PDA-based questionnaire was designed using Visual CE, with controls to guide data entry and reduce errors. Additional resources required included PDA equipment and accessories, necessary software and design applications to suit the survey, and PCs for downloading and backing-up data.

Results: Use of touch-screen PDAs revealed several challenges, including power outages, slow downloading in database format, accidental loss of data or damage of equipment, and security of PDAs. However, quality of data from PDAs was better and more quickly available for analysis than from paper questionnaires. Using PDAs was more expensive, but startup costs could be reduced over time as rental or purchase fees are spread among many surveys or other uses. Significant differences in managing PDA and paper questionnaire by data collection teams were identified. Main lessons were the need to design relational databases ahead of time, download data in text format, develop or adapt manuals and standard operating procedures for data management, procure sufficient supplies of back-up batteries and accessories, and strengthen technical skills of the data collectors.

Conclusion: This experiment showed that PDAs may be successfully used in implementing surveys in the African context. Use of PDAs speeds data collection, improves completeness and, most importantly, increases data quality. If all PDA challenges were addressed ahead of time, their use, as opposed to paper-