The study involved an attempt to promote use of evaluation data in compensatory education through a technical assistance effort at the state education agencies (SEA) level. It included a secondary analysis of Hawaiian Title I longitudinal data and dissemination of findings through a series of consultations with evaluation and program personnel at the SEA and local education agencies (LEA) levels. Impetus for the study came from increased awareness of information needs on the part of project staff for program design and improvement purposes. The study suggests several potent variables for predicting Title I student achievement. These variables include school type, school/project enrollment, pretest achievement status, per pupil cost, absenteeism, project setting and instructional approach. Outcomes of the study suggest that face-to-face technical assistance is not only a viable but perhaps one of the most effective ways of promoting use of evaluations. (Author/PN)