During last Thursday’s meeting of the Kentucky Senate’s Education Committee, Dr. Bob Rodosky, the statistical person from Jefferson County Public Schools, asserted his school district outperformed charter schools across the nation on the National Assessment of Educational Progress (NAEP) in 2011.
That assertion surprised me for several reasons. First, charter school performance has been improving on the NAEP. Second, the latest NAEP data is from 2013, not 2011. Why cite out of date statistics?
So, I got curious. I cranked up the Main NAEP Data Explorer Web tool to generate the following tables. These tables compare NAEP reading and math performance for the fourth and eighth grade from Jefferson County and the nation’s charter schools. This 2013 data shows Dr. Rodosky’s comments are no longer accurate.
The most notable example is for eighth grade math, where white, black and Hispanic charter school students all outscored their racial counterparts in Jefferson County Schools by a statistically significant amount. In fact, most of the score differences shown are much larger than just statistically significant – they are just plain significant (I’ll be happy to send the Excel spreadsheets with full information, including the sampling error metrics, to anyone really interested).
It’s also worth noting that the racial demographics in the nation’s charter schools vary considerably from those in Jefferson County. Whites only make up 30 percent of the nation’s charter school enrollment, but they comprise 52 percent of the Jefferson County enrollment. Because whites in both cases significantly outscore the minorities, it essential to break the scores out by race if you really want to understand what is happening. Just looking at overall average scores can be highly misleading, a problem with test comparisons I have discussed many times before.
Continuing on, things don’t look much better for Jefferson County when we look at NAEP eighth grade reading.
Charter school whites and blacks both outscored their Jefferson County racial counterparts by a statistically significant amount. Hispanic scores were tied, once you consider the sampling errors in the NAEP.
The situation is similarly troubled for Jefferson County when we examine fourth grade math results. Here whites and Hispanics in charter schools did better than Jefferson County’s students. Black scores are statistically tied even though the blacks in charter schools did score four points higher than Jefferson County blacks did.
Finally, when we look at the scale scores from the NAEP Grade 4 Reading Assessment, we see no advantage for Jefferson County, once the statistical sampling error in NAEP is considered (Rodosky never mentioned statistical sampling error).
The bottom line here is that one of Rodosky’s main arguments against charter schools falls apart when we examine the latest NAEP data. It is clear that even average charter school performance across the nation is better overall now than what Jefferson County is producing. The clear message is that Jefferson County could most definitely benefit from even average charter school performance.
Of course, you first have to make charter schools legal in Kentucky, an action some adults in the school system seem to want to block at all costs, even if their reasoning is wrong.