The purpose of the present study is to demonstrate the utility of data analysis methodology in evaluative research relating pupil and curriculum variables to pupil achievement. Regression models which account for achievement will result from the application of the methodology to two evaluative problems--one of curriculum comparison and another exploring the relationships between achievement and instructional processes in different schools implementing the same curriculum. Evaluative studies focusing on such questions should yield more information about pupil achievement than evaluations following other models when the following practices are reflected in the design and execution of the study: (1) several dimensions of the curriculum, including material and instructional process aspects, are represented in the set of independent variables; (2) curricular and pupil variables are chosen whenever possible from those conceptualized by educational researchers and known to have a likely relationship to achievement; (3) direct measurements on all variables are incorporated in the data set rather than categorical representations of variables, as in a factorial design; (4) the shape as well as the location of the distributions of pupil achievement before and after instruction is represented in the analyses; (5) the techniques of the data analyst guide the model development process. These recommendations result from a review of substantive and methodological literature. (Author/BJG)