This study is primarily an analytical investigation into certain aspects of the theory and practice of educational evaluation. The philosophical position taken is outlined in chapter 1. Briefly, it is that educational evaluation, except in special cases, necessarily contains an irreducible residue of indeterminacy that logically precludes the setting up of mechanistic evaluation routines for the assessment of educational phenomena. In the literature, reservations about the advisability of highly deterministic evaluation systems have been expressed primarily because of ethical concerns, the potentially damaging effects evaluations often have, and the limited basis provided by many evaluations for understanding how programs work. In this thesis, similar reservations are expressed, but for fundamental philosophical reasons that have to do with the constitution and use of criteria and standards.
Of the numerous definitions of evaluation that exist in the literature, the one adopted here is that evaluation is the determination of the merit or worth of something, whether that something be an educational program, a curriculum, or a piece of student work. A judgmental approach to evaluation, especially one which tests basic value assumptions, currently has only a small number of advocates in the field, the most notable being Michael Scriven. Several reasons for the relative unpopularity of this position are discussed in chapter 2, together with some speculations as to how a judgmental stance can be reconciled with a desire to advance and serve rather than hinder the practice of education. Apart from being in accord with the etymological roots and ordinary interpretation, evaluation viewed as a judgment about quality facilitates the identification of a number of issues common to a wide variety of evaluations. In particular, it makes it possible to focus on the emergence and functions of criteria, the definition and use of standards, and the conceptualization of composites, as principal elements in evaluative activity- These aspects are discussed theoretically in some detail in chapters 3-5, which with chapter 2 make up Part A of the thesis.
In Part B (chapters 6-10), a number of issues in evaluation design are explored. In particular, some claims and proposals made in writings about methods for organizing program evaluations are critiqued, two partial designs for data collection in different circumstances are developed, and the proposition that evaluations should ordinarily be accompanied by recommendations for change is analyzed. The final section of the study, Part C (chapters 11-12), contains accounts of two evaluations, one a field study and the other an evaluation of a curriculum principle. The investigation reported here thus contains both description and theoretical analysis, from which a number of normative principles are derived.