Does use of the calendar in surveys reduce heaping?

Abstract

This study examines heaping of reported durations of three postpartum variables in five countries based on two methods of data collection in retrospective surveys. Two Demographic and Health Surveys were conducted for each of the countries, one that used a five-year month-by-month calendar to record the durations and a subsequent survey in which duration information was collected from simple questions in the body of the questionnaire. Heaping indexes were calculated for breastfeeding, amenorrhea, and abstinence at six, 12, 18, and 24 months postpartum. Use of the calendar was associated with a significant reduction in the proportion of heaped responses for both breastfeeding and amenorrhea, and the effect increased for longer durations. To obtain more accurate information about these variables, the authors recommend the use of the calendar.