A framework for understanding methodological practices from the perspectives of internal validity, external validity, statistical control validity, and construct validity is presented. One hundred doctoral dissertations completed between 1980 and 1988 at a single urban public university were analyzed for various methodological practices and types of statistical techniques used. Techniques were further coded into categories of univariate or multivariate designs and subcategories within these categories. In all, 201 techniques were discovered across papers, which were generated by the Departments of Educational Leadership and Foundation, Curriculum and Instruction, and Special Education of the university. Demographics, size, and other information about sampling techniques were also assessed. Randomization and convenience were the most typical sample selection techniques. Means of establishing reliability of the dependent variables and whether surveying was involved for each study were also determined. Additional features concerning the discussion of results in the dissertations were coded as well. The analyses indicate that the instance of use of specific statistical techniques has changed little during the last 9 years. Three data tables are included. (TJH)