Academic Performance Index (API)

Archived Q&A and Reviews


Parsing the API Scores

Jan 2015

My local school does not have high API scores. The usual explanation for this is that there are a high # of English Language Learners, socioeconomically challenged families, and special needs kids, and these bring the scores down. School boosters say to look at the demographic subgroups and you will get a better idea of how well the kids are served. That is, kids from certain ethnic groups that might have higher parental education will score higher. This makes sense, and these kids in fact do score higher. However, when I go to the schools with higher API scores it turns out that ALL the kids do better including the ESL, school lunch, and LD kids. They don't do as well as the highest achieving subgroups, but they do markedly better than at my school. The high achievers do better too. So this means it's my local school failing to perform as well as other schools, right? And we should take the scores seriously after all? Trying to Keep an Open Mind

Causality is never a simple question. Higher API scores for a school indicate a whole bunch of things going on, which are difficult (if not impossible) to sort out. Certainly, higher API schools have higher levels of the particular ethnic/socioeconomic categories you mention. However, also consider the following possibilities:

  • More experienced teachers are given precedence when a school position comes open. They often want to obtain positions at higher API scores because the work is easier and (perhaps) more enjoyable. New teachers, with little experience, often end up at the lowest API schools, where the work is hardest.
  • Wealthier, better educated parents often contribute substantial money through a PTA or similar group, to finance all kinds of activities, extracurriculars, after-school programs, teacher classroom materials and assistance, playground repairs, etc. etc., which districts often cannot afford these days.
  • Higher API schools also tend to have higher levels of parent volunteers -- parents do everything from leading weekly reading groups in their child's classroom, to helping set up science fairs, to presenting special presentations on interesting jobs they may have.
  • Research shows that when the majority of kids in a school are high-achieving, that becomes the ''cool'' thing to be, and lower-achieving kids strive for it. When high-achieving kids are in a minority this doesn't always happen.
  • District officials often watch the lower-API schools very carefully to make sure they are adhering to all the curricular requirements of that district. The higher API schools are not watched so carefully (because they're doing fine) -- which sometimes means they can do activities and lessons that are more interesting and engaging for the kids.

It's a shame that it is this way; If there were more resources available to make all schools better, then all kids would do better, and the discrepancy would not be so great. And it's not the case that all low-API schools are bad schools, with poor teachers and nobody around who cares -- far from it! My mother worked at a low-income, low-performing school for nearly all her life, and she loved those kids. But many low-API schools don't have enough money and don't have many parents who can do all this donating and volunteering. Their teachers are often forced to spend nearly all their teaching time on mandated curriculum materials (which may not meet the kids where they are) while dealing with a severe lack of everything from books and new desks to paper and pencils. Some don't even have working libraries. All of this is going to have an effect on the kids there. anonymous

The truth is, nobody knows for sure. They might say they do, but they really don't. There are so many factors, so difficult to isolate and analyze. The prudent parent might err on the side of caution with this one--unless there is some wildly compelling reason to send your child to the lower-performing school. Anon

A school that is 10% English language learners/socioeconomically challenged is in a different position than one that is made up 70% of these groups. The more challenged school is putting out more fires and has teachers' and administrators' attentions divided between more students in need. The more advantaged school may be in a better position to support their 10%. So does that mean that the disadvantaged school is failing relative to the advantaged school? Probably not. If the demographics were reversed, I suspect that the APIs would reverse as well. When I look at APIs by subgroup, what I look at is the difference between the different subgroups, with me being more impressed with a narrow gap than a wide one. There is a lot of variation among schools in the size of the gap between Asian, white, black, hispanic, socially disadvantaged, and ELL. When one subgroup comes out so far behind, it raises red flags for me about that school (regardless of whether my kid is in the subgroup or not). I might have also parsed the data

I'm just guessing, but you might want to compare how much $$ the PTA raises each year. It would seem to me that more funding=more support for kids who might otherwise struggle. For instance, at our well-funded Oakland elementary, the PTA pays for aides in kindy and 1st grade, so the teacher isn't having to teach 30 little kids all by herself. She can focus more attention on the struggling kids if needed. Our PTA also pays for a ''pull-out'' teacher--again, for kids who are struggling. So that could be a partial explanation. Then there's the question of how big the difference in scores is between high-achieving kids at your school is versus the other school. If we are talking 50 points I wouldn't worry at all. OUSD mom

How did Montclair Elementary get the highest scores?

May 2010


The new API rankings are out: ( and Montclair Elementary has overtaken Hillcrest and Thornhill to become the high ranking elementary school in Oakland. I'm wondering if anyone knows how this was achieved? What has the school improved to get a higher score this year? -oaklandparent

You're talking about a difference of ONE point between Montclair and Hillcrest--that's hardly statistically significant. I doubt Montclair did anything very differently last year; this handful of schools always has the highest test scores (big correlation between student socio-economics and scores, of course). Anyway, rather than focusing on a few points here and there, why not look at the big picture? I'd be happy to have my kids at any one of the top 15 (in terms of test scores) Oakland elementary schools (and lucky for us, they are--though, horrors, their school hasn't topped 900 yet!). I don't mean to be facetious, but I find test scores to be a small percentage of what makes a great school (and ours is). If you look at the percentage of free/reduced lunch and/or English language learners, you'll see that almost none of the top 15 schools (in terms test scores) have more than 20% free/reduced lunch or 5% ELL students. Notable exceptions: Peralta, Grass Valley and Lincoln. Now, THAT's good teaching: high test scores, low-income kids and ELLs. Loving our Oakland school

Was the poster surprised? These three schools have been close in scores for many years. Oddly many people forget to group Montlcair with Thornill and Hillcrest (in terms of performance). All three have (sadly) less diversity and (unfairly) more parent resources than other Oakland schools.

Why did Montclair top out this year? ... An excellent principal? A down-turned economy sending more potential private school kids (with parents with even more resources) to the local school? Luck of the draw?

We have a kindergartner there and are very happy. But the API doesn't add to the experience. We're looking at teacher quality, kid happiness, enrichment, etc. Montclair is a winner, top score or not. We are fortunate. -Kim

A bit of clarification. These are not new scores. Scores come out at the end of Aug/beginning of Sept based on the previous spring testing. What came out this week was the ''Similar Schools Rank'' indicating how a school is doing first compared to the state and then compared to 100 similar schools. (Similar in size, demographics, economic levels, blah blah blah).

What you need to know is that scores can change from year to year based on so much (and so little)- Five kids were sick when testing, a tired kid, etc. If a school scores over 900, do not worry-it is a great ''scoring school''. A school with 940 vs. 960 is not going to be any different in quality. I say great ''scoring schools'' because you also need to know that there are schools scoring in the 700s that are perfectly fine and have quality education. Testing only tells us so much. It is an important measure but is only one measure. Don't allow the over emphasis on testing to stress you out. High School Administrator/Testing coordinator

Please be careful in assuming that the school has ''improved'' something in order to have the highest ranking API score. I'm not implying that they are doing anything wrong either -- but many things outside the school's control can affect the API.

First, like any test score, it is not so accurate that a small difference is meaningful. There's an ''error band'' around these API scores, and for the Oakland hills schools, I'm pretty sure their error bands all overlap. This means that they are really indistinguishable, even though one comes out a couple of points higher than the other -- i.e. Montclair is one point higher than Hillcrest and 15 points higher than Thornhill on a scale of that goes up to 1000 -- differences like this don't mean much, and schools can change rank from year to year without actually changing anything. Second, several of the hills schools are overenrolled (more students are in that school's area than the school can actually hold). This is true for Hillcrest, and I believe for Thornhill too. Last summer, Montclair added a bunch of new portables to its campus, and I believe this was so they could take in the extra students. Changes like this can account for small differences, as family educational and economic background can have a significant influence on API scores.

Finally, the API is only one part of what a school is. School climate, parent involvement, teacher philosophy, principal leadership style, and so on all are important in a child's education, and none show up in the API. Karen

I don't have children in Oakland Unified but speaking as a parent who has seen testing in the public schools for nearly a decade my first reaction to your question was: Who cares which school ''beat out'' another school on a standardized test? Education is not a baseball game. These are children--each one unique in their own right.

Children may or may not be attentive or focused on the day of the test. The questions may be worded in a way that they don't understand, even though they actually know the information. In addition, testing only on reading and math is a crude way of measuring a child's overall intelligence and problem-solving ability.

And does anyone think testing has done much to solve the achievement gap? Seems like more effective parenting and better preschool options have done more to address this issue. Sorry, but I think standardized testing is a failed program and can't wait until we move on to a new education ''reform'' fad in public schools. I would rather see testing done away with and more money going into the performing arts, science, and physical education. anon


Information about 1999 API Test Results

Feb 2000


The Berkeley Unified School District has all its schools' scores online together on one page here:

The State of California website will take some digging around and it looks like the documents are not in a very nice readable form:

The Oakland Unified School District has put all the state schools online in a very nice searchable interface - search for Oakland, Alameda County, or any district in the state:

Piedmont USD scores:

Albany: I could not find any info about API scores on the AUSD web site

Contra Costa: West Contra Costa USD just links to the state web page