You are greatly mistaken, although it does depend on the quality of the preparation.

I have seen scores improve 100-200 points per skill on average for middle range scores, and 50-100 points per skill for higher range scores with quality preparation.

Reasons:

- familiarity with test format

- improved test taking strategies

- familiarity with test content

- developing the mental stamina to perform at a high level throughout the relatively long test

- lowered anxiety and (often) increased confidence

The SAT goes from 200 - 800. If test prep raises average score by 100 points, there still is some room for the test to show actual knowledge.

The college board says that an SAT is better predictor of success in college than GPA, which based on my experience of American high schools makes sense. I get that wealth probably has a positive correlation to both. But it seems that should matter a little.

Why do you exclude learning the material as a reason? My test prep was largely a matter of learning material that I missed or wasn't taught effectively during school.

I would group that under familiarity with test content. Maybe an opaque label on my part.

although? Or the most important factor is, quality of prep. A man/women who have continuously, kept practicing to improve academically, becomes insignificant. However, keep throwing ball in hoop would help math and science innovations. This is absolutely insane.

Is this just personal anecdata?

An acquaintance of mine runs a high end SAT training program. He gets these results reliably for folks who have had little or no preparation before coming to him. Trained test takes is in the hundreds (maybe more).

Is this data or anecdata? You be the judge. He finds it easy to bank on professionally.

I encourage you to talk to someone who has worked at a high-end SAT training program (not Kaplan or that type). They can tell you the real deal, both successes and failures. Don’t rely on talking to the businesses themselves, because they will be selling their services rather than telling the complete truth (unless you know them personally).

Fwiw, failures are usually due to one of a few factors:

1. anxiety

2. learning blocks

3. physical issues (e.g., lack of sleep, sick, etc.).

Cramming for the test right at the end with SAT prep will add a few points but won’t do anything dramatic.

But if you get put in the “honors” track of a lot of high schools where the name of the game is to teach material informed by the SAT, ACT, AP, and IB teats then you almost can’t help but do well since you’ve been cramming for 4 years.

It turns out actually learning the material on a test over the span of a few years helps you do well on the test. That used to be called getting an education.

Right! But the problem is that you don’t actually get much of an education along the way. You get a bunch of not super useful knowledge that you’ve almost surely forgotten and some test taking skills. This is the downfall of so many programs that teach to the test.

The problem with this reasoning is that the SAT isn't an arbitrary test. Presumably the course material came before the SAT was developed to test mastery of the material. Sure, the course material then changes to better capture what the SAT measures, and the SAT adjusts in kind. But this adversarial process doesn't optimize arbitrarily. Learning SAT material (rather than say learning tricks) is learning substantial math and reading skills, the kind that serves you well in college. So characterizing this as anything other than getting an education is disingenuous.

Teaching the test and getting an education are not the same.

> cramming for 4 years.

That's not called cramming. It's called "learning". It means the test is working as intended.

> since you’ve been cramming for 4 years.

You could say the same for pretty much every educated person who has done an undergraduate or postgraduate degree. Seems like a lot of human progress has been made as the result of people that have been "cramming for 4 years"

Any evidence to support this?

That’s right, the average effect of test prep is around 0 increase.