Comparison of Data Quality between PAPI and CAPI Modes of Data Collection in Longitudinal Household Surveys: RLMS–HSE Experiment Results

Abstract

This article presents the results of an experimental study on how the transition from PAPI to CAPI modes affected data quality in longitudinal household surveys. The study was conducted in 2018–2019 within the Russian Longitudinal Monitoring Survey (RLMS–HSE). In the previous paper, which was based on data from the 26th wave of the RLMS HSE, it was shown that the use of CAPI leads to a significant decrease in the rate of non-substantive responses (“Don’t know”), as well as significant differences in sensitive questions. This paper was aimed at verifying these findings using new data collected during the 27th wave of the RLMS–HSE. The results show that the use of CAPI leads to a decrease in the rate of non-substantive responses, which helps to improve data quality. However, it was shown that the use of CAPI could lead to an increase in social desirability bias.

Keywords

• papi
• capi
• data quality
• panel household survey
• rlms–hse
• rate of nonsubstantive responses
• sensitive questions
• social desirability effect