ABSTRACT
Their many advantages in behavioral and educational research have encouraged the use of longitudinal designs in large-scale sample surveys. In the context of surveys they are referred to as “panel” designs, implying that each subject has made a commitment to participate in a number of future data collections. Such designs not only make possible true prospective studies and measurement of change, but they may actually be less expensive than a comparable series of cross-sectional studies. The initial cost of locating the subjects is so large in a national probability sample survey that it is more economical to call again on the same subjects than to interview new ones. For this reason, most commercial market research is based on panels of consumers, even when the data are used only cross-sectionally.
