statistics question about high school rigor

I mentioned that Temple and I are writing an op-ed on U.S. high schools.......and I'm stumped by a statistical issue.

What does it mean to say that multivariate analysis shows that a certain factor is highly predictive of a particular outcome, while another factor is less predictive?

What does this form of analysis imply about causality, if anything?

I ask because of an apparently highly influential government report published in 1999: Adelman, C. (1999). Answers in the Tool Box: Academic intensity, attendance patterns, and bachelor’s degree attainment. Washington, D.C.: U.S. Department of Education.

This study finds an extremely high correlation between rigor of high school curriculum and students' eventual completion of a college degree--far higher than the correlations with the factors we're used to hearing about, such as parents' level of educational attainment, socioeconomic status, and race (and substantially higher than high school GPA and SAT scores).

The report itself, which I've barely skimmed, as well as other accounts of it, seem to imply that the relationship is causal. It's not that being smart and motivated in the first place causes a student to take a more rigorous high school curriculum and attend and complete college, but that the more rigorous high school curriculum sets him or her up to succeed in college.


from the American Educator:

Academically well-prepared students are likely to graduate from college regardless of their social background. Unprepared students of all backgrounds are not likely to do so.

The graph below breaks students into quintiles based on their level of academic preparation and their socioeconomic status (SES). As you can see, among the lowest SES students, a bachelor’s degree was earned by 62 percent of those who were well prepared, but only 21 percent of those who were not. Similarly, among the highest SES students, 86 percent of those who were well prepared--but only 13 percent of those who were not--earned a bachelor’s degree.

Percentage of students who graduated from a four-year college by socioeconomic status (SES) and academic preparation.


key words: rigorous high school curriculum predicts graduation from college
how can you tell whether A caused B?
low birth weight paradox
how good are our best students?
statistics and law

Back to main page.