Age Heaping in CATI Surveys in Kenya and Nigeria

Age heaping is the tendency for respondents to round age to multiples of 5. This study explores the prevalence and correlates of age heaping in eight large national computer-assisted telephone interviewing (CATI) surveys in Kenya and Nigeria. Using Whipple’s Index, a standard measure of age heaping, we found significant age heaping across all surveys, particularly in Kenya. Age heaping is more prevalent among less educated respondents (Kenya only) and Swahili speakers (Kenya only), with no differences by respondent gender (both countries). Notably, age heaping was highly sensitive to different aspects of the data collection process. There were significant differences across call centers (“house effects”) and interviewers. Heaping was more pronounced among female interviewers (Nigeria only), inexperienced interviewers, breakoff interviews, and weeks 4+ of the data collection period (Kenya only). Collectively, our findings suggest that age heaping is not solely a function of respondents’ characteristics but is also influenced by the data collection process. While rates of age heaping were high, our research suggests that age heaping can be mitigated through careful oversight, training, and survey design.