Does "Advanced" Predict College? We Checked 1,838 Schools
A school's "Advanced" score tells you a little about where its kids go to college. It tells you almost nothing else. Advanced is the share of kids who hit the top level on California's state test.
We compared 1,838 California high schools. We matched them into 365 pairs that looked almost the same on paper. Same overall test scores, same share of low-income kids, same money in the neighborhood. The only real difference: how many kids scored Advanced.
The schools with more Advanced kids sent more kids to college. About 2 to 3 more per hundred went to a University of California (UC) campus. About 3 more per hundred went to college of any kind, two-year or four-year.
Grad rates didn't move. Community college numbers didn't move either.
And a good chunk of what the Advanced score measures is just how well-off a school's families are. Here's how we checked it, what held up, and what didn't.
How did we test this?
We matched every school to a near-twin that differed mainly in Advanced share. Advanced share and a school's overall score move together almost perfectly, so without matching, you can't tell which one is doing the work. For every school, we found another one with almost the same overall test score, the same share of low-income kids, and about the same money in the neighborhood. That gave us 365 fair matchups to look at.
What did we find?
This is the most solid result in the whole study: those extra four-year students did not come from community college. We checked community college numbers four ways — the full group, no charter schools, bigger graduating classes, and a different year. Every time, the number barely moved, and it moved in different directions. Random chance explained it 570 to 816 times out of 1,000. There's nothing there to explain.
Nothing got taken away from community college. It's a real, strong path, and kids choose it for reasons this study can't see — cost, a job, staying close to family. Those extra four-year students got added on top of what was already there.
Averaged across all 365 pairs, the school with more Advanced kids sent:
- About 2 to 3 more kids per hundred to a UC
- About 3 to 4 more per hundred to a four-year college
- About 3 more per hundred to college of any kind
Those numbers held up under three separate checks: no charter schools, bigger schools only, and a different testing year. Random chance explained the UC gap 0 times out of 1,000. It explained the four-year gap just once. That's about as sure as this kind of study gets.
We can't say the school caused this gap. We can say it isn't random chance.
Here's every outcome we tested, in one picture.
365 pairs of California high schools
We tested seven things. The two biggest gaps are the two that didn't hold up.
Every bar compares two schools that looked the same on paper — same overall test scores, same share of low-income kids, same kind of neighborhood. The only real difference was how many kids scored Advanced.
Gap in kids per 100
See the numbers as a table
| Outcome | Gap | Verdict | Evidence |
|---|---|---|---|
| AP pass rate | +9.0 | Didn't hold up | 330 in 1,000 — it was charters |
| a-g courses | +6.3 | Didn't hold up | 586 in 1,000 in a third year |
| Four-year college | +3.6 | Held up under every check | 1 in 1,000 |
| College of any kind | +3.2 | Held up under every check | 3 in 1,000 |
| UC | +2.3 | Held up under every check | 0 in 1,000 |
| Graduation rate | +0.3 | Didn't hold up | 686 in 1,000 |
| Community college | −0.5 | Didn't hold up | 620 in 1,000 |
So what should you do with this?
Treat Advanced share as one data point among many. A higher share probably means a few more kids per hundred end up at a four-year school, with no real change to community college numbers. It says almost nothing about grad rates. Read it together with the school's overall score — the two tell you different things.
If two schools look the same everywhere else — same overall scores, same neighborhood — the one with the bigger Advanced share is worth a second look. We still don't know why. The college numbers just say it's real.
None of this tells you about your own kid, though. This is 1,838 schools and one graduating class, averaged out. Your child isn't an average.
Check your own school's Advanced share on its profile page, right next to the numbers we compared here — see it for yourself. For more on reading your child's own score report, here's our guide. And here's why Advanced share can look so different at two similar schools.
What didn't hold up
We looked for a reason a bigger Advanced share would lead to more kids in college. Three guesses, three letdowns.
Grad rates. We checked this five ways. None came close to a real gap. Chance alone explained it as often as 7 times out of 10. Whatever Advanced share is picking up, it isn't who finishes high school.
AP pass rates. This looked real at first — a 9-point gap between higher- and lower-Advanced schools. Once we pulled out charter schools, the gap dropped to under 3 points, and chance explained that 330 times out of 1,000. Too weak to trust. A pass rate only counts kids who took the test. Pushing more kids into AP can make a school's own pass rate look worse — for doing something right.
A-g courses — the classes UC and Cal State require for admission. This looked like the answer: a real gap, two years running. Then a third year made it vanish. We're calling it unproven.
We got it wrong twice ourselves before we got here. Advanced share and the overall score move almost in lockstep. A plain comparison can't tell them apart. An early version of this work said Advanced added nothing. A second version said it hurt community college numbers. Neither survived matching schools that actually looked alike. We're telling you that because it's what a wrong answer looks like on the way to a real one.
We still don't know why the college gap is there. We just know it's there.
Method: 1,838 California high schools, class of 2023, matched into 365 near-twin pairs on test scores, share of students on free or reduced-price lunch, and neighborhood income. One class, one year: a single snapshot. Data source: California Department of Education, CAASPP and College-Going Rate data.