<p>^ neomom, I agree with most of what you wrote, but we can't disregard the reality that one's SAT Writing score correlates almost exactly as well with college grades as high school GPA does.</p>
<p>I suggest those interested actually read the study in question. A few points worth noting. This study was released in 2008. I tmerely concludes that the changes made to the SAT did not substantially change how well it predicts FIRST YEAR college performance. Some basic analysis: the general rule of thumb is that a large correlation starts at about .5. The raw R for SAT-w is .33, the adjusted R is .51, so this is hardly an impressive performance. SAT-W does have the highest correlation with FYGPA of the SAT sections, but this is hardly impressive. SAT-CR has Rs of .29 and .48, and SAT-M has .26 and .47. (The adjusted figures deal with the restriction of range issue.) Calling this "a very strong predictor" , as the CB does, is a bit extreme.
Additionally, the 2009 study which breaks down by gender/race/language shows considerable underprediction - as high as -0.26 .
To put this all in context, other studies have shown that SAT to be quite weak at predicting grades. Vars and Bowen, in 1998, assert that the SAT offers about the same predictive value as looking at an applicant's parental education level. THe UC study , in 2001, asserted that SAT 1 scores added so little to the predictions made based on other data as to be completely valueless. Bates, which dropped the SAT requirement in 1990 also found that SAT scores were of very little value.</p>
<p>^ If you consider GPA and SAT to be weak predictors of grades, what do you instead recommend?</p>
<p>Tis not I who considers both GPA and SAT weak predictors of college grades, tis the studies by CB, UC and Bates.
There are no very good predictors of college grades. SAT scores and high school GPAs are among the factors most commonly used and, of the less incendiary data used, are of some value. But a correlation of , at best, .51, is not a strong correlation.</p>
<p>Does anyone have any validity statistics on the ACT? The stuff on their Web site that I have been able to read through hasn't had what I am looking for.</p>
<p>According to [url=<a href="http://www.act.org/aap/pdf/ACT_Technical_Manual.pdf%5Dthis%5B/url">http://www.act.org/aap/pdf/ACT_Technical_Manual.pdf]this[/url</a>], the correlation between ACT scores and college GPA is .42 (page 93 I believe). Is this comparable to the SAT's .53?</p>
<p>The SAT essay is very predictable. If you look at any of the essays that get 12's they're all the same format. You need examples from your knowledge, critical analysis, and enough writing to fill the two pages. For example, I had an introduction, two examples from literary works, a short discussion, and a conclusion. I just about filled the two pages, maybe with one line blank, and I received a 12. And believe me, I'm not the best writer out there.</p>
<p>Yes it is in a way somewhat biased and cruel, but if you know what they're looking for, why not just do it?</p>
<p>I went from 7-11 on my essay...</p>
<p>I think the main change was length and organization.</p>
<p>@ silverturtle: it is college board published material</p>
<p>I'd like to see an update of the UC study now that the "new" SAT has had a comparable amount of data. But given the CB's own study, it would surprising if the result were much different from the original UC conclusion, which was the following.</p>
<p>The UC study found the following average ranking of predictor strength for first-year college GPA (from lowest to highest):</p>
<p>1) SAT I
3) SAT II
4) SAT I + SAT II
5) HSGPA + SAT I
6) HSGPA + SAT II
7) HSGPA + SAT I + SAT II</p>
<p>(#1 was much less predictive than #2; also, #6 and #7 were basically indistinguishable)</p>
silverturtle: it is college board published material
<p>What are you referring to?</p>
<p>I got a 10 on my essay in June and was fully expecting an 11 or 12. Not that a 10 is bad...it's still an improvement from my 8 in March, but it was soooo much better than my March one..I don't know where I could have improved? It took up the full 2 pages given, and I had strong examples. :/ I wonder if it's because I took what was probably the unpopular side on the topic? I dunno...</p>
<p>^Your side on the topic doesn't matter.
You need a nice, big helping of luck in order to jump from a 10 to a 12. Essay's subjective. Always has been, always will be.</p>
I got a 10 on my essay in June and was fully expecting an 11 or 12. Not that a 10 is bad...it's still an improvement from my 8 in March, but it was soooo much better than my March one..I don't know where I could have improved? It took up the full 2 pages given, and I had strong examples. :/ I wonder if it's because I took what was probably the unpopular side on the topic? I dunno...
<p>It is impossible to ensure that you will receive 12, and it is very difficult to make 12 likely. Graders are not consistent: there are three times as many 11's as 12's. That means that the graders gave different scores.</p>
<p>question: does anyone know how the ACT essay scoring is compared to the SAT essay? Is it less subjective, not about length, and overall better?</p>
question: does anyone know how the ACT essay scoring is compared to the SAT essay? Is it less subjective, not about length, and overall better?
<p>Probably about the same. There's nothing about the SAT itself that makes the essay grading bad; it's human nature.</p>
<p>What is the point difference between a 10 essay and a 12 essay? Like how many points am I losing out on?</p>
<p>It really depends on your MC grade. It could be anywhere between nothing and 50.</p>