Thomas, A., Chess, S., Birch, H. G., Hertzig, M. E., & Korn, S. (1963). Issues in longitudinal studies. In A. Thomas, S. Chess, H. G. Birch, M. E. Hertzig, & S. Korn, Behavioral individuality in early childhood (pp. 8-23). New York, NY, US: New York University Press.
http://dx.doi.org/10.1037/14328-002
Although many aspects of growth and development may be adequately investigated by means of successive matched age stage samples, longitudinal study is uniquely effective in the exploration of: 1] the relation between earlier characteristics of the individual and features of later development (developmental process or prognosis); and 2] the delineation of individual behavioral sequences. For most other questions serial sampling of appropriate populations at different ages represents an entirely adequate and more efficient study design. In view of the logic of the longitudinal method and the kinds of problems to which it is pertinent, it is clear that, when psychological characteristics are considered, the method is principally applicable to the study of age stage characteristics of an individual's functioning. Further, if the method is to be effectively employed, one must, as in the case of the longitudinal studies of physical growth, identify objectively measurable characteristics of individual functioning which may either themselves or through their sequelae be followed in time. These considerations provide the fundamental prerequisities for pursuing any longitudinal study of psychological development. (PsycINFO Database Record (c) 2016 APA, all rights reserved)