Urban Elementary Schools and MCAS (Cambridge, Somerville: low income, neighborhood, best schools)

[semiurbanite]

As a parent of young kids in an urban neighborhood, I have been researching schools both here and in the suburbs, looking at MCAS scores, touring a few schools and talking to other parents. One parent that is very happy with the urban schools urged me to look at the data from a demographic perspective, so I went for it. Here is an xls showing MCAS scores for "Top" schools like Brookline and Weston compared the "below-average" urban schools like Somerville and Cambridge. I added Arlington because if we every felt forced to more for the school we would probably move there.

We have no desire to leave for the suburbs, and based on my research so far I see absolutely no reason to move to another city for the elementary schools. Most may find this shocking, but if you look at "Non-Low Income" student only across these schools, Somerville and Cambridge schools actually do quite a bit better in terms of percentage of student scoring "Advanced" on the MCAS tests than Weston and Brookline!

https://docs.google.com/file/d/0B1OS...it?usp=sharing

[missionhill]

Thanks, semiurbanite. This seems like very useful information and a welcome contrast to the reflexive response that city schools are inferior.

[claiz]

Semiurbanite - interesting data. My only concern is since you are looking at per-year, per school, sub-group data, the data set is sliced really thin. And I think primary school test score is hard to judge upon since the school does not have enough time to really change/educate the student. Middle and High school data is more reliable in my opinion.

If I look at non-low-income group 2012 data, but at the whole district to broaden the data set to enough data points, here is what I found (I throw in Lexington since it's where I live so purely my own curiosity):

At Grade 3 Math:
Cambridge - 50A+32P
Arlington - 38A+34P
Somerville - 35A+45P

Weston - 48A+34P
Brookline - 49A+34P
Lexington - 60A+28P

At Grade 3 Reading:
Cambridge - 32A+44P
Arlington - 33A+51P
Somerville - 30A+45P

Weston - 35A+50P
Brookline - 29A+53P
Lexington - 39A+49P

Grade 10 Math
Cambridge - 66A+23P
Arlington - 70A+22P
Somerville - 36A+28P

Weston - 75A+21P
Brookline - 79A+16P
Lexington - 87A+9P

Grade 10 Reading
Cambridge - 49A+42P
Arlington - 69A+29P
Somerville - 25A+54P

Weston - 70A+28P
Brookline - 68A+29P
Lexington - 77A+22P

You can almost make an (overly simplistic) argument that this non-low-income group enters the school systems in elementary school with similar, sometimes even higher, academic performance, but after they got educated in their respective school systems, the performance parity disappears when it comes to grade 10.

[semiurbanite]

Claiz,

The numbers do change after elementary school. However I think there is another factors at play here. First, there are quite a few very highly educated parents in the area, many are faculty and researchers at MIT, Harvard and Tufts, or Biotech workers that work in Kendall. These kids likely perform very well, coming from involved, educated parents, and skew the averages upward. After 4th or 5th grade, some of them do tend to move out of the city for more space, or because they see the scores at the middle school level. Some also go to private school. But as more and more stay, as I am seeing, I think we may see things change quickly. Either way, after doing my research I have little concern until my kids are approaching middle school.

As for using the whole district, that is interesting. I cherry picked the top schools for all of the districts. In Somerville that will make a difference, as the top 2 schools probably have 80-90% of the highest income earners. In the other schools, while they may have just as many Non-Low Income students, they are probably much closer to the income threshold.

[claiz]

Semiurbanite - I understand what you are saying and I agree that may very well be what's happening. I still think using a broader data set can give you more reliable insight. Your data shows % of Advanced Reading student in Graham and Parks from 32% in grade 3 to only 4% in grade 4 then 46% in grade 5. For Brown in Somerville, % of Advanced Math student goes from 37% to 10 then to 55%. Which means your dataset is quite "noisy" and hard to derive conclusion from.

Having said all this, regardless of test scores and school systems and all that, I think parental involvement is the most important predictor in children academic success. So the fact that you spent so much time researching school performance means you care deeply about your kids' academics and likely to stay involved, which is the most positive factor you can have.

[semiurbanite]

I see your point, the sample size is small. However the degree to which these "sub par" schools outperformed their peers in the best schools does say something in my opinion. The discrepancy from year to year could also be the teacher, as they have the same teacher all day at this age.