The Science Score, the STEM Score, and the Composite

The Science section has 34 scored questions, and the raw number correct gets converted through a scaling table into a score from 1 to 36. That scaled score does two separate jobs. First, it stands alone as a signal of science reasoning skill, compared against a readiness benchmark of 23. Clearing that benchmark is an empirical estimate: students who score at or above it have historically had a strong chance of earning a C or better in a first-year college biology course. It says nothing about chemistry or physics performance specifically, just a general science-reasoning benchmark tied to that one course.

Second, the Science score combines with the math score to build a STEM score. That average is simple: take the math scale score and the Science scale score, add them, divide by two. No English or reading score enters the calculation, and no separate STEM test exists. The STEM score exists to give a quick read on a student's combined quantitative and scientific reasoning, since those two sections draw on overlapping skills like reading graphs, tracking units, and reasoning from data.

Where this gets tricky is superscoring. A superscore builds the best possible composite by taking the highest section score a student has earned across every sitting, even if those sittings happened months apart. If one sitting included Science and a later one left it out, the Science score from that earlier sitting still carries forward. A missing section in a later sitting can't erase or outrank a real score from an earlier one.