Given the recent articles on proficiency, I decided to clear the cobwebs out of my stats brain and run a few numbers to add context to the discussion.
I took the last three periods of proficiency and percentages on special needs across the 7 district dashboards, ran Pearsons, converted it to a T-Stat, and determined a probability value (I welcome anyone to correct my math here- seriously; it’s been a while since I ran numbers in this manner and kind of did this in between meetings)
***Corrected- a friend pointed out that one of my formulas was referencing the wrong cell-***
I got a p-value of 0.019 for math proficiency, which indicates statistical significance
A p-value of 0.030 for Science which indicates statistical significance
A p-value of 0.000018 for ELA which indicates statistical significance for language arts proficiency; however, Language Arts is the strongest out of the three areas for Berkeley Heights, with a 77% proficiency rate.
We can’t really draw any strong conclusions especially given the sample size is a few short of the minimum for Pearson’s but it lends more credibility to exploring this avenue.
The percentage of Special Needs MAY impact our proficiency and this small snapshot may be a clue that its an area we need to dig deeper on. It does point out that we can’t just throw out a percentage and claim it’s the reason with the conversation stopping there.
Just because we may have more students with special needs should not automatically cause us to accept lower proficiency scores—especially when you consider that we spend more per pupil than the six other Districts on the Dashboard in this area.
So if Special Needs students are in fact a significant factor in our proficiency concern (which MAY be the case) parents of special needs students should really be asking why.
Here is the spreadsheet with the numbers and math.
More Context:
Just a note that Special Needs MAY be a factor in connection to proficiency, the information above does not mean that it is a factor specific to our District in connection to our declines – the information above only shows a correlation across all schools between two variables.
Millburn spends significantly less per pupil on Special Needs, has the fifth largest 3 year average percentage of Special Needs Students while ranking highest in every proficiency category out of the 7 Districts.
Even More Context (955pm)
So I got lazy and ran the spreadsheet through Gemini to determine the extent differences in proficiency can be explained by special needs:
For Math:
25.4% of the variation in math proficiency is explained by the special needs percentage. The strength of this relationship is weak, suggesting that there are other important factors that influence math proficiency scores.
For Science
22.3% of the variation in science proficiency is explained by the special needs percentage. The strength of this relationship is weak, suggesting that there are other important factors that influence science proficiency scores.
For ELA
62.8% of the variation in English proficiency is explained by the special needs percentage. Special needs percentage explains a relatively substantial proportion of the variation in English proficiency scores between districts.