White minus black achievement gaps continue problematic in new NAEP results – Grade 8 Math

The 2017 National Assessment of Educational Progress (NAEP) results for Grade 4 and Grade 8 mathematics and reading have finally been released, and I already discussed Kentucky’s Grade 4 NAEP white minus black students’ reading gap and mathematics gap, and the Grade 8 Reading gap. So, let’s look at Kentucky’s NAEP Grade 8 white minus black proficiency rate gap for math.

Here is how the NAEP Grade 8 Mathematics Assessment proficiency rates have trended for Kentucky’s white and black students since NAEP first reported state results for this school grade and subject in 1998.As is the situation for the earlier cases I’ve already discussed, this isn’t a happy graph. Click the “Read more” link to learn why.

G8 Math White Minus Black Gaps on NAEP by Year

First, a few comments about the graph’s design.

All the proficiency rate numbers come from the NAEP Data Explorer web tool and are shown in yellow shaded boxes. All of these NAEP proficiency rates after 1996 include results for learning-disabled students who received NAEP-accepted accommodations. The superscript number 1 for earlier years indicate only non-accommodated students were tested.

White proficiency rates are found along the blue line and black scores are found along the red line.

The gap between the white and black proficiency rates for each year is shown by a number in the middle of the doubled-headed arrow for each test year. For example, in 1990 Kentucky’s whites scored 11 percent proficient or above on NAEP Grade 8 reading while only 2 percent of the black students met this level of performance. That difference produces a proficiency rate achievement gap of 9 points.

Now, let’s explore what else the graph shows.

First off, the Kentucky NAEP Grade 8 white minus black achievement gap in math has been growing for some time. In fact, using the NAEP Data Explorer’s statistical significance testing tools, all the gaps identified with an asterisk are statistically significantly smaller than the latest gap for 2017. So, even using the not-very-sensitive-to-small-changes NAEP, the eighth grade math gap in Kentucky has definitely grown since the early days of KERA. Nominally, the gap of 23 percentage points in 2017 is the worst white minus black gap Kentucky has posted to date on the NAEP, though once the statistical errors in NAEP are considered, the best we can say is that the gap has been flat since 2007.Even more disturbing, there has been very little improvement in black math proficiency since 1990, over a quarter of a century ago. No rational person would cheer about Kentucky’s black eight graders only posting a single-digit math proficiency rate in 2017.Clearly, unless the NAEP is very badly off target somehow, math performance for Kentucky’s blacks is simply inexcusable. For the state’s whites to improve proficiency rates from 11 to 32 percent while blacks only went from two percent to nine percent proficient is just horrible performance.

As we discussed before, some have raised concerns the NAEP’s transition to digital testing in 2017 might have created some problems for comparing the new results to earlier testing, and the impacts might be more severe for less advantaged students. To examine that possibility, I looked at the state’s KPREP eighth grade math scores for both races for 2015 and 2017. The results are in this table.

Keep in mind that, unlike NAEP, Kentucky tests virtually all public school eighth grade students, so there are essentially no sampling errors in these KPREP scores. That doesn’t mean the scores necessarily present an accurate picture of proficiency, however. Without question KPREP scoring is less demanding than the NAEP. For example, in 2017 KPREP declared 52.1 percent of Kentucky’s white eighth grade students were proficient in math while the NAEP declared only 32 percent met the proficiency muster. For blacks, KPREP declared in 2017 that 25.8 percent did math proficiently but NAEP says Kentucky’s black eighth grade math proficiency is far less than half of that at just nine percent.

These are notable differences, much larger than anything that can be excused by NAEP’s sampling errors. Basically, for the subject of eighth grade math KPREP shows proficiency rates significantly higher than those shown by the NAEP.

KPREP G8 Math for 2015 and 2017

The trends in KPREP are interesting to compare to NAEP. Note that between 2015 and 2017, KPREP reported that white math proficiency in Kentucky increased by 4.7 percentage points to a rate of 52.1 percent. Referring back to the graph of the NAEP Grade 8 Reading scores, the federal test says white proficiency increased by two percentage points between those two years, though that change was not statistically significant.

However, NAEP also says Kentucky’s black math proficiency in Grade 8 decreased by a statistically insignificant three points while Kentucky’s own test showed blacks gained four points in proficiency. That is a pretty sizable difference in the trends. Could it be that there was some sort of differential functioning problem for NAEP Grade 8 Math in 2017 that put blacks at some sort of disadvantage compared to whites?

To examine that a bit more, I looked at the trend in the national black students’ NAEP Grade 8 Math Scale Scores from 2015 to 2017 (scale scores are more precise than the proficiency rates). The nationwide black score stayed perfectly flat at 260. In notable contrast, Kentucky’s black eight grade math scale score dropped from 257 to 252. That notable drop still wasn’t big enough to be statistically significant, however. So, once again, the statistical measurement errors in NAEP might be covering up a small problem with testing changes in 2017, but the data isn’t sensitive enough to tell us that.

In any event, when NAEP says only a single-digit percentage of Kentucky’s black students can do math proficiently and that blacks notably lag behind whites after more than a quarter of a century of KERA, we have some very serious problems in our public school system.