I posted a first observations blog about the new 2015 NAEP scores for all Kentucky students earlier. In that post I only compared Kentucky’s scores between the years.

Unfortunately, doing valid comparisons of Kentucky’s NAEP scores to those from other states isn’t so easy. For one thing, statistical sampling errors in NAEP turn many apparent “wins” into nothing more than ties. Other factors such as wildly different student demographics and the ever-present achievement gaps complicate state to state comparison of scores, as well. You can’t just compare overall “All Student” scores from the NAEP and get a valid idea about relative state education system performance. You have to break the data out and do more refined comparisons.

I have done that for the 2015 NAEP math results using the NAEP Data Explorer tool to generate maps that show states where white students scored statistically significantly higher, the same as, or statistically significantly lower than Kentucky on NAEP math.

This first map covers the Grade 8 NAEP math results for 2015.

There isn’t much change here from the results I posted two years ago after the 2013 NAEP was published. We now outscore two states instead of just one in 2013, but the number of ties has gone down so that now one more “jurisdiction” outperforms us than we saw in 2013, as shown in the second map.

Any way you slice this, Kentucky’s middle school math performance does not look so great.

Things are a bit better for the elementary grade tested by NAEP.

This next map ranks our white fourth grade NAEP math results in 2015 against other states. We now outscore four states, up from only one in 2013. Still, we are not exactly doing that much better.

This looks better than our results from 2013, shown in the next figure. But, there is a catch.

Is the better performance in the fourth grade because Kentucky got better, or because other states got worse? This table, which compares Kentucky’s white Grade 4 NAEP Math Scale Scores to the national public school average for white students, provides insight.

As you can see, the white national average scores dropped two points between 2013 and 2015 while Kentucky’s stayed flat. The appearance of better performance for Kentucky isn’t because Kentucky did better; it is because the national situation got worse. That nationwide decline created a lot of statistical ties for our fourth grade students in 2015, but this is because a leveling took place, not an improvement.

In fact, a lot of states only really started to do Common Core State Standards in the past two years. Could it be that this has impacted performance of formerly better-performing states, leveling the playing field by dumbing down the rest? If that is what Common Core produces, I don’t think it will last very long, at least outside of Kentucky.

One more point: In 2013 I did an additional examination of how our poor, school lunch eligible whites did against poor whites elsewhere. I don’t feel comfortable doing that with the 2015 data, unfortunately. A huge problem has arisen with the national school lunch statistics in the NAEP because of a new program where any school that has at least 40 percent of the students eligible for lunch based on real need can now go into the “Community Eligibility Option” (CEO) lunch option and feed ALL of their students, even sons and daughters of wealthy parents. Even worse, this may impact the numbers reported with the NAEP test results in uneven ways across states. That has really messed up the validity of the NAEP school lunch data as a proxy for poverty. Absent a lot more information that so far I have not been able to locate, I won’t do work with the lunch data this year.