30 April 2012

A Debate Over Academic Standards: Response Staples

To John Staples:

Thank you for your interesting and thoughtful response. Your argument is that even if Fredonia is losing ground on SATs relative to competitors this need not bother us because the SAT is a flawed measure of academic potential.

1. SAT as Measure: I disagree. Once the restricted range is corrected for, the correlation between SAT scores and first-year GPA is reasonable. On one large 2001 meta-analysis it was somewhere in the 0.44 to 0.62 range. The likely reason for this is that SATs correlate strongly with general intelligence (g) as measured by IQ (.82 correlation on one study) and IQ scores predict a wide range of socially important behaviors, often with greater accuracy than parental socio-economic status.

2. Plausibility: The notion that SATs do not track any useful information is implausible. In About.com, Old Westbury’s mid 50% SATs are 955-1110, Yale’s are 1410-1590. Is it plausible to think that these scores do not tell us anything about the relative academic ability of incoming freshmen at these schools from these scores? I suspect we would agree on this issue.

3. GPA as Measure: Even if you think the SATs are unreliable, I would be surprised if you thought the same about a measure that combined high school GPA and SATs. Again, high school GPA when combined with the SAT score is a reasonable predictor of a student’s first-year college GPA, especially relative to its other measures. Among SUNY colleges in 2011, the high school GPA of Fredonia is ranked 7th behind the usual suspects: Geneseo, New Paltz, Cortland, Oneonta/Brockport (tied), and Oswego. I do not know if this was true in 1990.

4. SAT and Graduation: One former Provost of SUNY (Peter Salins, 1997-2006), argued that for a four-year period in SUNY campuses with highly differentiated missions, whose high school GPAs improved by the same modest amount (2-3%), only those whose SAT scores improved substantially saw gains in graduation rate. Those that didn’t change their SAT scores did not see an increase. I have doubts about how strong evidence this is toward the relation between SATs and graduation, but it is worth considering.

Thank you again for the note,
Steve K

No comments: