The thesis of this paper is that, in evaluating most innovative educational projects that are conducted in schools, tenets of formal experimental design and associated inferential data analysis methods should be given limited emphasis. The basis of this thesis lies in the problems and difficulties that undermine the design and implementation of the typical project. Sections of the paper include discussions on design problems and difficulties, data collection, data analysis and reporting, statistical proportions as an evaluative measure, and graphs. These are followed by a presentation of the considerations that may be made in an evaluation report in the face of the problems and difficulties. (Contains 14 references.) (RR)