"SAT Scores Decrease As a Higher Percent of Students Take the Test"

There are too many notable exceptions to the rule (that SAT scores increase strictly as a result of the decrease in the percent of high school students taking the test) to rely too heavily on this rule. A simple correlation of the scores with the percent of test takers indicates that this is about 10, to a maximum of 20, percent of the reason for the wide variation in scores (240 SAT points, from 867 in DC to 1107 in North Dakota) from state to state.

The smallest percent of students taking the test, 4%, is in Utah and Mississippi, but their scores are nowhere close to similar--there's a difference just between these two states of 23 points in SAT math and 17 points in SAT verbal. More than twice as many students, 9%, take the test in Minnesota, but they score 39 points higher than Mississippi in math and 10 points higher in verbal. 12% of the students in Tennessee take the test, yet they score higher in both verbal and math than Mississippi. Clearly there's something more at work here than simply a difference in the number of students taking the test.

The same notable exceptions hold true in the states where the highest percentage of students take the test. Maine's low SAT math score of 469 and verbal score of 427 has been blamed on the fact that 68% of their students take the test. But 80% of Massachusetts' students take the test and they score 8 and 3 points higher, respectively. And 70% of New Hampshire's students take the test and score 22 and 17 points higher respectively. 74% of New York's students take the test and, even though they DO score 8 points lower than Maine in verbal, they score 4 points higher in math. Only 53% of DC's students take the test so they ought to score HIGHER than the above states, but they score LOWER than Maine, 24 in math and 15 in verbal. Similarly, 51% of the students in Oregon take the test, but they score 54 points higher in math and 36 points higher in verbal than students in DC where a similar percentage of students take the test.

This is not to say that there is NO correlation between the percent of test takers in a state and their average SAT scores, just that it's a relatively minor factor relative to others (education spending, environment, race, health, tax rate, types of teachers, amount of central government control, attitude, genetics, and religion, to name a few).

Hit Counter

Modified Thursday, November 12, 2009