The major goal of the University of Southern Maine Reading Academy Project was to demonstrate the effectiveness of process-oriented, diagnostic/prescriptive instruction by trained tutors in raising the literacy levels of adults reading at less than a sixth grade equivalency level. Preservice and inservice training using an empirical model of reading development was provided to students who served as tutors to approximately 240 adults. It was hypothesized that the extent of students' gains in literacy and self concept would correlate highly with the extent of tutors' adherence to the specified program. The procedure for measuring program implementation involved translating the process-oriented approach to reading instruction into specific, observable teacher behaviors and measuring their effectiveness. This was accomplished through specification of essential components of the model program, development and use of tutor interviews and rating scales to collect data on program implementation, and analysis of data by means of a multiple regression equation. Student achievement in reading and changes in students' self concept were measured by the Adult Basic Learning Examination and the Self Esteem Inventory. Extent of program implementation was shown to be the factor that weighted most heavily on prediction of residual posttest scores. (YLB)