California 1992 and 1996 SAT Scores
for IMP Schools

May 1997


Quantitative Scholastic Aptitude Test scores of California public high schools from 1992-1996 are available on "The Web" and the Interactive Mathematics Program (IMP) reported 64 California high schools associated with their curriculum in 1995-96. This report looks at those two facts only. No claims are made of cause and effect nor replicability in other schools. This data verifies that CA schools that use IMP dropped over 7 points on a per student basis from 1992 to 1996 while California as a whole dropped by 5. The data does not indicate a dramatic increase in the number of students taking the SAT. That number has gone up 6.9% at these schools, roughly half that of California overall which has gone up 13.6%. It is often reported by IMP supporters that students stay in mathematics longer, i.e., elect more mathematics classes than students in traditional programs. Assuming that is true, this data indicates that one should wonder about the effectiveness of that endeavor. Perhaps doing a better "selling job" of traditional courses or more stringently mandating them for college entry to state supported schools would be more effective at improving the percent of seniors taking the SAT and their subsequent performance.

There is a wealth of valid reasons why using the simple fact that an individual school appears on the IMP list might be misleading information. Perhaps the most telling is a quote from Diane Resek, one of the IMP Co-Directors, in a letter to me of September 25, 1996:

"To our knowledge, there are no schools that offer IMP as the only college intending mathematics program. We encourage everyone to offer students, parents, and teachers a choice of programs, including both IMP and a traditional mathematics program."

In some schools, mathematically stronger students have been encouraged to enroll in IMP. In others, stronger students have been encouraged to avoid IMP. Some students have been encouraged to use outside resources to prepare for the SAT, some are "savvy" enough to do that even if not encouraged to do so, etc. Suffice it to say that much more careful research would need to be done to make any genuinely definitive statements about the wisdom of IMP in California schools in regard to eventual performance on the SAT. There may be schools or subpopulations within schools for which it is inferior, neutral, or superior and such studies should be underway.

These overall numbers are obviously too close to be meaningful but there are indications of concern and even a little hope in the individual school data. The IMP list divided the schools into nine geographical regions and there is some indication that the further away from the bay area home of IMP, the worse the performance. Perhaps better training and support have been available close to home? Only one of the regions showed an actual gain in SAT score performance and only two exceeded the California 5 point decline. They are the two San Francisco regions, North and South San Francisco Bay. A couple of others showed declines of about 7.5 and the rest were in the 13-18 range. Live Oak High School in Santa Clara County and Santa Cruz High School each had SAT jumps in the 30 point range but both declined in absolute numbers of students electing to take the SAT. Determining the actual level of IMP involvement of top end students at these schools might be informative.

There are dramatic declines in a few schools as well. Roosevelt High School in LAUSD dropped 25 points and 39 fewer seniors elected to take the SAT. Marshall High, also in LAUSD, dropped 41 points and 64 fewer students took the SAT. Those numbers are bad enough that concern should have already been raised by district supervising personnel. The usual disclaimers obviously apply here as well; perhaps the IMP involvement is minimal and has nothing to do with the drop. They are both schools that signed up in 1992, however, so should now be reaping gains if they are to be had. Given the socioeconomic nature of the populations at these schools, concern must also be raised that perhaps the program is less beneficial for some populations than others.

Three comments about details should be made. The first is that 13 of the schools, as listed by IMP, were not recognizable by the SAT website so not in the data. Another is that several of the schools were reported as starting IMP in Fall 1995. Since this testing was done less than one year later, and since the program would probably be being phased in slowly, those schools were omitted as well. The third is that the SAT scoring was "recentered" with the 1996 scores so the apparent jump in most scores is misinformation. On the College Board web page is the concordance table for translating old scores into the recentered ones and that was used to convert the 1992 results into a form comparable with those of 1996.

If anyone would like further details, please contact me. I would be pleased to answer questions about the methodology I used and provide a copy of my chicken scratchings.

Respectfully submitted,

Wayne Bishop
Mathematics & Comp Sci
Cal State LA
Los Angeles, CA 90032