A few days ago I saw in my feed the headline “High-Poverty D.C. Charter School Students Outscore Wealthy Neighbors in Math.” When I started blogging about 15 years ago, stories like this were very common and I would dig into the data to find if the claim was true. When a school’s results were proved to be “too good to be true” they became known, sarcastically, as ‘miracle schools.’

Being a miracle school debunker is kind of a sad specialty, like being an autopsy doctor or one of those people who sifts through airplane crash wreckage to find the cause of the crash. If I’m successful at my investigation I show that the underdog did not actually win this time.

But in this era of social media, it is important to fact-check things like this otherwise what can happen is that the false claims start to influence politicians and eventually it leads to ineffective policies getting expanded. Like in Tennessee where the idea that charter schools were performing miracles in New Orleans led to about a billion dollars being wasted in a state takeover district.

So the claim in this article is that 70% of the 8th graders in Center City Congress Heights charter school in D.C. 8th Ward while only 64% of the 8th graders in the affluent 3rd Ward passed the same test.

So the first thing that I did was find if the claim is accurate at all. I found the database and verified that this school had 20 8th graders of which 14 passed the test so, yes, 14 is 70% of 20 so the claim at least is based on something is true. But 20 students is a very low number of students. A school usually has at least 75 students per grade. And even in this school most I checked and the other grades in this school have around 25 students. So attrition might play a part in this statistic since 14 out of 25 would be just 56%, for example.

The comparison to the affluent 3rd Ward, I think, might be that 288 out of 450 8th graders in a 3rd Ward school called Deal Middle School passed the same test, which is 64%.

Center City is a charter school network of which the Congress Heights school is just one of six schools in the network. So I looked up the other five schools in the network and found that they didn’t do quite as well.

The Brightwood campus had 10 out of 25 passing (40%)

The Capitol Hill campus had 5 out of 24 passing (20.8%)

The Congress Heights campus (the one in the article) had 14 out of 20 passing (70%!)

The Pentworth campus had 6 out of 24 passing (25%)

The Shaw campus had 3 out of 21 passing (14.3%)

The Trinidad campus had 4 out of 25 passing (16%)

So taken as a whole, in the network they had 42 students pass out of 139 which is 30%. This is slightly better than the D.C. average of 25% but definitely the 70% school was an outlier which is an excellent example of how data can be ‘cherry picked.’ If this charter network had some secret that worked so well on one campus, why not just use the same methods in their other five schools?

I also checked to see how charter schools, in general, are doing on this test so I looked up the gold star KIPP D.C. network. They only had 13% passing, with one of their schools with 3 out of 156 which is less than 2%. Now I know the article wasn’t about KIPP, but still

It is such a shame that propaganda sites like The74 are still running articles like this. The idea that there is a quick-fix in education has been proved wrong over and over. And though I don’t enjoy having to be that guy who has to give the full story and spoil the party, I still think it is important to call out these stories to give a full context.

To the 14 students (and to their teachers) who did pass the Algebra 1 test at Center City Congress Hall, you should be proud. Only 25% of students in D.C. passed it so each student who passes should be celebrated. But you don’t want your individual success to become something that gets used in propaganda and ultimately used as a weapon against other hard working kids and teachers.

Here is the spreadsheet with the public data I used if you want to check my numbers

2024-25 Public File School Level DCCAPE and MSAA Data 1Download