History of Longitudinal Statistical Analyses

Abstract

Synonyms Longitudinal data; Longitudinal data analysis; Longitudinal research; Longitudinal studies; Repeated measures (Re)Constructing the History of Longitudinal Statistical Analysis Longitudinal statistical analysis has a long past but a short history. In fact, until very recently, longitudinal statistical analysis did not exist as a subject, but was inextricably tied to substantive research in different disciplines. Even today, most publications on longitudinal statistical analysis are written from the perspective of a certain discipline and focus on a specific research design and data structure. Examples include pertinent work on large sample panel data (Hsiao 2014) and single-subject time series data (L€ utkepohl 2005) in economics, crossover experimental designs in medical and pharmaceutical research (Jones and Kenward 2014), and the typical applications in the social sciences, often involving multiple individuals and a moderate number of repeated measurement occasions (Singer and Willett 2003). Depending on the field of research and the focus on a specific design or data structure, the history of longitudinal statistical analysis may thus be construed quite differently. This encyclopedia entry is primarily devoted to the history of longitudinal statistical analysis in relation to geropsychology and developmental psychology. As a field, however, modern geropsychology is highly interdisciplinary and thus subject to various influences. As the influences of neighboring disciplines wax and wane, so does the importance of their historical backgrounds. For this reason a description of the history of longitudinal data analysis remains always preliminary, reflecting our present-day view on the most important trends and influences that have shaped the field up to now. As the field changes, the perception of its history may change as well. Taking on today's notion of the main rationales of longitudinal research and the best way to address them, this encyclopedia entry will sketch the development of the most prominent approaches to longitudinal statistical analysis in geropsychology. Thereby, the focus lies on identifying relevant historical trends rather than presenting a chronology of discrete events. Specific dates mentioned in the following are thus best understood as exemplary milestones of more general developments.