Presents a model for conducting research on the career development of special populations. Reviews available research paradigms and offers an alternative framework designed to make maximum use of available data sources while minimizing potential theoretical bias. Presents an illustration of research conducted within the recommended framework. (Author/RC)