August, 1969
Law School Admission Test
VALIDITY STUDY SERVICE
In 1965 the LSAT Council determined that validity studies for individual schools were an important service which should be provided free to interested member schools. During the following four years 128 studies have been conducted for 97 schools.
In June 1968 the Council voted to make the Service a permanent feature of the program. It is hoped that new schools will continue to join and that schools already having had studies will have new studies every three or four years.
It is important in using test scores for academic selection to know the extent to which they sort students in roughly the same fashion as does a measure of academic success: that is, the extent to which the scores are valid as predictors. One accepted way of determining the validity of test scores is to study the statistical correlation between the scores and a criterion of later success, usually grades. While the statistical prediction of grades is not the only, or perhaps even the most important, function of law school admissions procedures, it is useful in suggesting efficient ways of combining scores and undergraduate grades for predictive purposes.
One outcome of the validity study will be a prediction chart on which a predicted first-year law school average can be located for any combination of undergraduate average and LSAT morning score. The report, which will be partially computer-produced, will include both statistical results and a general explanation of basic concepts relevant to validity studies.
Beginning this year, the service will include a study of the predictive value of the data on undergraduate colleges listed in the LSAT Statistical Summary by Undergraduate College Attended. For this part of the study, we will need the college code designated on the LSAT score report slip. You are also asked to supply the student’s registration number shown on the score report. This information will make your data more useful for future research studies designed to improve the test. Moreover, if you wish us to study an additional predictor for which you can supply data, we can readily do so.