Panel data consists of repeated, systematic observations of the same person, household, and village over time
Methods for collecting panel data
Panel studies have proliferated across a wide range of disciplines in the behavioral sciences, public health, and medicine, but are rare in developing nations and in cultural anthropology Tsimane’ Amazonian Panel Study (TAPS): The first five years (2002-2006) of data available to the public in 2008. Panel data has one large advantage over cross-sectional data: it allows one to use people as their own control and, so, obtain more reliable estimates of change for the person across time. Panel data allows one to remove the confounding role of fixed attributes of the person that remain constant over time and that might cloud results in a standard cross-sectional study, including an ethnographic study.
In TAPS we measure outcomes every year during June-August through structured surveys and direct observations. During September-December we track attriters and enter and clean the data, which is ready for use by January-March. During January-April, field staff pilot-test new modules and time the new survey. We keep the survey constant across survey rounds to enhance comparability across time and villages and to reduce measurement errors. We introduce changes to the survey after pilot testing new questions or modules. Many modules remain constant, but others are used only during some surveys. Experienced translators whom we have trained over the years help to collect the data. After a survey during some years, we have tracked hard and soft attriters— people who have moved nearby (soft attriters) and people who have left the research area (hard attriters)— to reduce the likelihood of introducing attrition bias when analyzing the information.
