Opting In: The Power of Testing – PSSA Math

In my most recent Opt In post, I argued that PSSA data provides powerful insight into student achievement and school performance.

Today I take a break from the theoretical case against “Opt Out” and get very practical.   What data is available from standardized tests?  How is that information be used?  What insight does that information provide to parents, administrators, teachers, and school directors?

In my next three posts, I will publish data taken from PDE’s public test score database, which is found here.  First up — how are our students and schools performing on PSSA math?

The scatterplots below are drawn using PSSA’s analysis tools, which are available to the public. For help understanding the meaning of the data, you can watch video explanations here.

The first scatterplot shows UCFSD’s position (red square) relative to all PA school districts (yellow ovals).  In addition, the blue rectangles show Chester County school districts.   School that are performing well on achievement and growth are those in the top right quadrant.

Chart 1:  PSSA Math – UCFSD vs. All PA School Districts


UCFSD students are at the very top on the achievement dimension (left axis), with only two districts in the state achieving at a higher level.

On the growth dimension (horizontal axis) UCF is at approximately the 90th percentile.  Although our schools are demonstrating an ability to impart more math learning in a year than the average PA school district, about 10% of districts show better growth than UCF.   So we are very good, but not yet elite on growth.

Let’s zoom in on Chester County.  Within Chester County, how do our individual schools perform on PSSA math?

Chart 2:  UCF Building Results vs. Chester County


At the building level, the results are not universally positive.  Although achievement results are quite strong in all five buildings, growth scores are mixed.  Pocopson (PES) and Hillendale (HES) growth scores are not materially different from average (0 growth index).  Even Chadds Ford  (CFES) and Patton Middle School (CFPMS) are short of the top quartile.  Only Unionville Elementary (UES) has a strong growth score on math.

Note that PES is the highest performing of the five schools on the achievement dimension, but the lowest on the growth dimension.

What does this mean is going on at PES? PDE’s guide ( PDE Growth and Achievement Reporting) gives us some clues.   First, PDE tells us that achievement results are highly correlated with socio-economic background.  And they are also influenced by factors outside of school.   On the other hand, growth scores show little effect from socio-economic factors, and reflect largely what happens inside school.

So one simple interpretation is that PES students come from affluent families who emphasize academics (achievement).  But PES is doing only an average job of delivering math education.  With 2014 being a transition year to a new math curriculum (Math in Focus / Singapore math) perhaps there were implementation challenges.  (Of course, these same challenges were faced in the other four district schools that had higher growth scores).NO Mediocrity

There are certainly other possible explanations.  Maybe it was just a bad year and growth scores will be better in 2015.  Maybe the school has high achieving students but doesn’t challenge them enough.   (The administration has access to reams of reports and data to better answer these questions).

But my main point is this:  don’t we want to have this kind of information available to parents, administrators, teachers, and board members?  And if we want the data, we need to have annual standardized testing of all students.

Without this PSSA-derived information we might look at PES’ high achievement results and assume all is well and that the school is firing on all cylinders.  Instead, the growth score tells us that there are dozens of schools in the county doing a better job teaching math, and that (at least in 2014) PES was not among the best on this dimension.

Next up:  PSSA reading scores.