13 September 2021

SAT Test Preparation Isn't Very Effective

The main problem with moving away from standardized tests is that almost all of the alternatives are less meritocratic. Contrary to popular belief, test preparation classes to which the wealthy has disproportionate access don't make much of a difference for the vast majority of students taking them.

One pillar of the case against standardized testing is the widespread belief that wealthy students carry an advantage because they can afford expensive test prep courses and tutors. That’s what critics mostly mean when they say the SAT is a test of family wealth, not of academic ability.

Is this true?

Let’s start with some findings that pretty much everyone who studies this stuff seems to agree on. 
First: It’s true that test prep, which I’ll define as outside help that costs money and requires an investment of time, is generally used by wealthier and better-connected students. 
But second: The effects of test prep have been studied pretty extensively, and while there’s far from any consensus on why some students do better than others, the published studies agree that the range of improvement, once controlled for a variety of factors like the fact that students who enroll in and complete test prep courses will likely be a self-selected group, is about 10 to 35 points.

Does test prep really help everyone who has the money to sign up for a course, even if it raises their scores just a little? Not quite. 
Two studies found that when you disaggregate for ethnicity, Americans of East Asian descent benefit far more from test prep than any other group, including white and other Asian American students. (There’s an interesting if somewhat unrelated distinction to make here: One-on-one tutoring seems to help nobody. Commercial test prep, which ranges from cram schools in East Asian enclaves to the Princeton Review, has some effects.) This might explain why Asian Americans’ SAT scores have steadily been rising over the past decade.

According to a study conducted by Julie Park and Ann Becks in The Review of Higher Education, “East Asian Americans were the only group where a form of test prep predicted a higher SAT score (about 50 points).” For everyone else, SAT prep has no significant effect or even, in some cases, a negative one. A previous study found that the majority of this improvement took place in East Asian immigrant enclaves like Flushing, in Queens, which has dozens of cram schools that serve ethnic communities.

From the New York Times. 

2 comments:

Tom Bri said...

So, perhaps a study of their training techniques is in order? Are the students different, the teachers, the materials?

andrew said...

My suspicion is that the distinctive factors are (1) teaching substantive math competence as opposed to mere test prep (since the subject matter for math is much more compact than verbal), and (2) resolving some common English language defects common to children whose parents aren't native speakers of a language. I also suspect that the number of hours involved in those test prep classes is at least 10x the programs that they are being compared to.