Longitudinal Research

Longitudinal means lengthwise; a longitudinal study deals with the growth and change of an individual or group over time. The Handbook of General Psychology differentiates between cross-sectional, longitudinal, and clinical research. In the field of psychology, “the longitudinal method consists of measuring the same individuals at different intervals over a period of time. Some longitudinal studies are relatively short-term, others are impressively long-term: thirty or more years” (Wolman 1973: 864–65). An example of an extremely long-term study is Lewis Terman's “Life Cycle Study of Children with High Ability,” which he started in 1921 and finished in 1991.

Scholars have used other terms besides longitudinal research to describe these studies. The Danish sociologist Kaare Svalastoga preferred “sequential”; the American psychologist John Dollard described life history as the “long-sector view of culture.” Researchers may describe the longitudinal relationships as dynamic or static. A longitudinal study and the data set on which it is based, especially when the time covered is not too protracted, are referred to collectively as a panel.

Scott Menard argues that longitudinal research does not describe a single method but a family of methods. According to R. T. Campbell, longitudinal data contain information that one can order in time. Both Campbell and Menard include repeated crosssectional surveys on new independent samples at each measurement point among longitudinal studies. Such sets of surveys may follow the development and change of categories on an individual level. This, however, seams to be a minority position, as Campbell himself recognizes.

A two-dimensional diagram known as a Lexis diagram, named for the German statistician and economist Wilhelm Lexis (1837–1914), may illustrate the longitudinal design. For a study of a local set of newborn children, the horizontal axis can mean “time” and the vertical axis “age.” The researcher fills in the life courses of people belonging to the local community as sections of lines going diagonally upward and toward the right, rising at an angle of 45 degrees, if she uses the same time scale on both axes. For a person born in the community at time t1, the line will start at the point (t1; 0). For a person who moved to the area at time t1 at age a1, the line will start at point (t1; a1). Each line will go on until it reaches the point of time and age where the person either died or moved out of the area. If the person left but later returned, the line will stop when the person leaves and start again when the person returns. In the diagram, persons whose lines started or ended at about the same time, that is, who were born, died, or moved in or out at the same time, a cohort will be represented by a cluster of parallel life lines running close to one another.

A vertical line at time t will cut the life-course lines at various ages, giving a cross-section of the population at time t. A series of cross-sections will cut the lines at various times, resulting in a set of dependent samples, or panel. In a longitudinal study, one follows the life lines lengthwise, diagonally, and upward. One may make observations at certain intervals when lines cross vertically, as in a panel study, or they may be made continuously.

...