-There are fundamentally two data sets in play here, a national one, NCES, that provides basic school information: name, # of students, demographics, etc. And a State data set.
- Each state administers their own state assessment tests for most, not all, grades. So for example, Massachusetts does tests for grades 3-8 and grade 10. Other states do a different mix. To assign a "score" to an Elementary, Middle, or High School you combine the grades, but there is judgement around, for example, what constitutes a middle school? Grade 6? Grade 8? It varies by school district, let alone state.
-The subjects covered are typically Math and English. Some sites, especially for high school, also test Science or additional subjects. In our case, we include ALL available tests in the score calculation. GreatSchools provides no insight into whether they do this or not. SchoolDigger used to just use Math/English but recently added all tests to their rankings.
-The "results" for, say, Grade 3 English are typically broken out by Advanced, Proficient, Not Proficient, Failing. But this, too, varies by state, some include 5 buckets.
- Special needs students. This could be a really active community for you.
- Arts/Music/Theater programs and participation rates.
- Innovative educational programs. One of the HS I attended had an on-campus fish hatchery that was almost entirely student run, it also had a professional choral program--as in, students in the "varsity chorus" got paid professional rates for a summer show series. Another that I attended had a pretty unique combined history/literature and later math/science program. I could imagine parents being interested in finding those kinds of things.
- University affiliation. Many colleges/universities are affiliated with K-12 schools where they are involved in applying the latest in educational practices.
- Early college programs. My wife managed to get her first 3 semesters of her Mech-E degree paid for through an innovative relationship between her HS and the local Community College. The local CC also has a great direct-to-university relationship (all credits transfer) with two local Universities. Allowed her to graduate undergrad a year early and with half the student loans she otherwise would have.
- Proximity to cultural institutions. Friends Select in Philadelphia is located on the Ben Franklin Parkway... walking distance to the Franklin Institute, Academy of Natural Sciences, Phila Art Museum, Barnes Foundation, Rodin Museum... It's almost embarrassing.
Standardized test scores are important, but they aren't necessarily the only thing parents are looking for.
If you can showcase that, your website could be a really powerful resource!
-Special needs: Hadn't considered this, but the data is there to break out numbers and performance of special needs students. Are there other factors we should consider?
- Arts/Music/Sports etc: Yes, would LOVE to include data on this. I'm not aware of a dataset that contains it. Manually it'd be tough to do for 100,000+ schools.
- Innovative/college programs: Yup, great idea. As we expand the site we want to find ways to highlight features like this. A simple example we can do now is Language Immersion programs. But we want to extend it in the future.
- Another "problem" with the education data in general is that there's no way to compare schools across states. Common Core might have addressed this but currently you can only say, This is a top high school in Massachusetts. What if I live across the border in New Hampshire? How do 2 schools compare?
You could try to crowdsource this information, or contact schools in an effort to get them to self report.
The two main issues when we've thought of this are:
1) how do you ensure the data is accurate/up-to-date?
2) the graveyard issue where you don't want to show what's NOT on your site. As an example (and I know I'm probably picking on them unfairly) check out GreatSchool's crowd-sourced photos here: http://www.greatschools.org/search/search.page?q=boston%20ma)
Scraping sites is probably more of the way to go or just using Wikipedia as Google does, for example, in their school info boxes.
It's definitely something that, longer term, would be good to seriously look into.
It might not be the full-scope of what you are trying to do, but it might be a way to help bootstrap part of your development.