Week Two

Hello there again random stranger on the internet, you are once again reading the blog of Charlie Rhodes.

I meant to put this post out earlier this week, but, things and life got in the way. Not the least of those things was the fact that we had to put our cat down earlier this week. She was 20 years old, and her kidneys had started to fail completely.

Suffice it to say, it’s been a long week.

But, back to the topic at hand, I made some interesting discoveries about the links between scores and pass rates this week building on the analysis I performed last week which can be found in Week One Part Two. I won’t go into too much detail on that analysis, as it should still be up from last week, but suffice it to say, I discovered that there was a major difference between English scores and English pass rates.

It turns out that while pass rates and scores do differ and are regressed on different variables for English, the same is not true of math. Both pass rates and scores for math can be explained by two variables, Percentage White (PW) and Median Household Income (MHI). In both cases these variables explain more than 40% of the variation, and both PW and MHI are statistically significant.

One of the interesting parts of this finding that I am only just beginning to get into in my analysis, is that the percentage of a district which is African American (PAA) matters more than PW for English, while the reverse is true for math. It is to the point where Median Household Income, one of the staple explanations for variations in test scores, is not significant for math pass rates when we regress on PAA, but is extremely significant when regressed on PW.

I am currently trying to figure out why this is, as well as why there is so much less variation between math scores and pass rates compared to what we saw in English last week.

So, that’s been my week, hopefully the next one is better.