ABSTRACT
This chapter analyzes how evaluation in its early days could have been so "off-base" and summarizes the lessons from that period. It presents a five-tiered approach to program evaluation as an attempt to build those lessons into a graduated model of evaluation. Since the key to obtaining "truth" in the natural world was seen to be rigorous logic and empirical study, science held the promise of unlocking the mysteries of every other sphere of life, including human development. Program evaluation has evolved not only through internal process, but also in interaction with a new set of programs developing in a markedly different political climate. Program evaluations serve many masters; the audience for a particular evaluation will not only influence decisions concerning data collection and analyses but also study dissemination. The preimplementation tier, creates the conditions for all subsequent evaluation efforts. Many respected researchers have persisted in applying "context-stripping methods of investigation" to their studies.
