Summary of Self-Reported Methods of Test Preparation by LSAT Takers for Testing Years 1991-1992 Through 1996-1997 (TR-97-02)
Andrea E. Thornton

Executive Summary

This investigation of Law School Admission Test (LSAT) preparation patterns for the 1991-1992 through 1996-1997 testing years represents a replication of earlier studies by Wightman (1990) and McKinley (1991). As with the earlier studies, all analyses reported here were descriptive in nature, and no attempt was made to evaluate the effectiveness of different test preparation methods.

In this study, five different types of analyses were performed. First, analyses were conducted to compare the response rates for each testing year to determine whether there were appreciable differences in response rates across these years, and to assess the extent to which response rates in this study differed from those reported in earlier studies. Second, analyses designed to compare respondents and nonrespondents in terms of mean age and mean LSAT score to determine the extent to which the respondents were typical of the entire testing population were carried out. Third, an evaluation of the utilization rates for the different methods of test preparation was carried out to assess the frequency of use of the different methods. Fourth, the extent to which test takers used multiple test preparation methods was evaluated. Finally, users and nonusers were compared for each method in terms of mean LSAT score and mean age to evaluate the extent to which users of a particular method are different from nonusers.

Overall, the patterns of results for respondents and nonrespondents were consistent across testing years. The mean LSAT score was higher for respondents than for nonrespondents, and the mean age was higher for nonrespondents than for respondents. This was similar to the patterns reported for the earlier studies. These results indicate that the respondents differ systematically from the nonrespondents, and caution should therefore be exercised in generalizing any of the findings of this study to the nonrespondents. However, the response rates for all of the testing years was so high that this represents only a very minor limitation in the interpretation of the results.

The patterns of usage for the various methods of test preparation appeared to remain relatively stable across testing years. The most popular method of preparation was use of the sample questions in the LSAT/LSDAS Registration and Information Book (referred to here as the Information Book).

On the average, respondents used between two and three methods of preparing for the LSAT. The most common number of methods used was one (selected by approximately one-third of the respondents for each testing year), but typically from 14 to 20% of the respondents reported using two, three, or four methods. The lowest average number of methods used, and the highest percentage who report using only one method, was for the 1993-1994 testing year. Those respondents who are not fluent in English and who are Puerto Rican tended to report relatively low numbers of methods used, as did respondents indicating English was not their dominant language.

The most significant finding in the analysis of users and nonusers of each method is that LSAT scores were higher for respondents indicating use of any of the first five methods—sample questions in the Information Book, sample test in the Information Book, official Law School Admission Council (LSAC) test preparation materials, book not published by LSAC, commercial test preparation school—and lower for respondents selecting the last four options—undergraduate institution courses, self-study, other methods, and no methods. 

Among the conclusions reached for this study were the following:

  • Test takers in the earlier testing years (i.e., 1991-1992) were more likely to respond than were test takers in the later testing years.

  • Females were slightly more likely to respond than were males.

  • Caucasians were more likely to respond than were members of other ethnic subgroups; members of the Canadian Aboriginal subgroup were least likely to respond.

  • Those for whom English is not the dominant language, or who are not fluent in English, were less likely to respond than were those for whom English is the dominant language or who are fluent in English.

  • Respondents tended to be, on the average, one year younger than were the nonrespondents, and tended to score one to more than two points higher on the LSAT than did the nonrespondents. 

  • The largest differences between respondent- and nonrespondent-LSAT scores tended to be for Canadian Aboriginal and Puerto Rican subgroups, with Canadian Aboriginal nonrespondents outperforming Canadian Aboriginal respondents and Puerto Rican respondents outperforming Puerto Rican nonrespondents.
  • The Information Book and test preparation materials published by LSAC were clearly heavily utilized, as were books from other publishers.

  • Courses provided by undergraduate institutions were used by relatively few test takers.

  • Females tended to use more methods of preparation than did males.

  • Puerto Rican test takers and test takers not fluent in English tended to use fewer methods of test preparation than did other subgroups.

  • Caucasian test takers tended to use more methods of preparation than did other subgroups. 

  • Users of undergraduate institution courses, other methods, or no methods tended to have lower scores than nonusers of these methods.

  • Users of LSAC-published test preparation materials, commercial schools, and non-LSAC test preparation books tended to have higher scores than did nonusers of these methods.

  • Users of the Information Book materials tended to be older than nonusers, while users of commercial schools, self-study, and other materials tended to be younger than nonusers.


Summary of Self-Reported Methods of Test Preparation by LSAT Takers for Testing Years 1991-1992 Through 1996-1997 (TR-97-02)


Research Report Index