The basics of your ACT score are easy to understand, and that’s really all you need to know to have the test figured out.
You get raw scores for each section – this is just how many questions you get correct (and no marks deducted for wrong answers). This is then scaled using data from previous tests to give you a score out of 36 in each section. The mean average of these is your composite score (the one you report to schools).
Your essay is marked on a 6 point scale by two markers in the categories of Ideas & Analysis, Development & Support, Organization and Language Use. This gives you a score out of 48, which is then scaled on the 1-36 scale. Read more about the new essay here
Here’s what happens:
Another thing that will appear on your score report is ‘sub-section’ scores. English is broken down into ‘Usage/Mechanics’ (grammar, punctuation etc) and ‘Rhetorical Skills’ (making additions/deletions from sentences, reordering sentences and paragraphs). You’ll get a score out of 18 on each of these sections. These are basically meaningless, since these numbers are not involved in the calculation of your composite score. They just show you which questions you’re better or worse at. The only real use of these numbers is for the ACT data crunching and for you to target your preparation.
Here’s what your report looks like (click on sections of the image)
So how does the 1-36 scale come about?
Essentially, the ACT is one large data crunching machine. Most of the information I have in this article comes from a set of 100+ page reports that the ACT produces. Here’s a summary:
The scaling chart/score curve is slightly different for every test, because every test is slightly different in difficulty. Once per year, a random group of test takers sits a collection of new tests and one test that has already been administered. The percentile on the new tests is scaled to the old test so that the same number of students are getting each score. Then, the scaling chart can be calculated for all those new forms. This dispels the urban myth that you should write the test at a certain time of year because the scoring is more generous. This just isn’t true, and you can read the data for yourself from the ACT Technical Manual here .
By how much might my score vary?
This is a really interesting one. Every possible measurement that you can make includes an error (you know this from science class), and the ACT is no different. Here’s a quote from the ACT’s Technical Manual:
“The average standard error of measurement is approximately 2 points for each test score and subscore, and 1 point for the Composite score.”
Standard error of measurement (SEM) is how far you could statistically be expected to deviate from your true score. So, if you get 45/60 on the Math section, you true score could be anywhere between 41 (if your true score is a 43 and you were on the upper bound of your high score) or a 49 (if your true score is a 47 and you were on the lower bound this time). In terms of your composite score, if you got a 27 this time, your true score could be anywhere between 25-29 (SEM on the composite score is only one point). This is proof of why you need to write the test more than once, straight from the mouth of the ACT themselves!