If interpretations of data compiled during a five-year demonstration of alternative educational programing in the elementary schools of California's Alum Rock School District are correct, the nature of, the size of, and the parental choice exercized over alternative programs have no significant effect on students' reading achievement, social skills, or perceptions of themselves and others. Originally an experimental educational voucher system within the public school system, the Alum Rock project never became a full-fledged voucher program but did enable researchers to examine the relative cognitive and noncognitive impacts of several variables in educational programing. Neither the experimental group of students (those in alternative programs) nor the control group (those remaining in regular programs) was a random sample, information was lacking concerning the motivations behind parental choices, and test data were imperfect, so researchers had to modify their information to accommodate sample biases before applying analysis of covariance. Over 9,000 students were involved in the study, 60 percent of whom enrolled in alternative programs. The results, presented with caution, indicate little difference in the benefits to students of alternative and regular programs, but show a marked relationship between student outcomes and teachers' perceptions of themselves and their programs. (Author/PGD)