Regular readers know that the 2015 scores from the National Assessment of Educational Progress (NAEP) were released yesterday, and I have already posted some initial blogs about the results. Some key points:
• Kentucky’s NAEP performance in both math and reading in both the fourth and eighth grades is flat in 2015 compared to the state’s results back in 2009, a year before the state adopted the Common Core State Standards. If Common Core had impacts in Kentucky, they are too small for the NAEP to confidently detect them.
• A fair comparison of the NAEP math performance of Kentucky’s white students to counterparts in other states shows our eighth grade students continue to perform very poorly. Kentucky’s eighth grade whites only outscored whites in just two other states in the entire country and were outscored by whites in 42 other states and Washington, DC. While the state made some progress in fourth grade white math performance, that appears to be mostly due to scores dropping in other states rather than any improvement here. In 2015 Kentucky’s white fourth grade students only statistically significantly outscored whites in just three other states.
Now, it is time to examine another question. Are the 2015 NAEP results just a fluke? Was this just a “bad test day” thing for Kentucky?
An answer to this question comes from another set of tests given to all Kentucky eighth grade students, the ACT, Inc.’s EXPLORE tests. EXPLORE also tests reading and math, and it reports results against carefully developed “Benchmark Scores” that are tied to a real likelihood that students as of the eighth grade are on track to be ready for college.
When we compare the percentages of Kentucky students who meet or exceed the EXPLORE Benchmark Scores to the percentages of the same student cohorts that score proficient or more on the NAEP, we discover rather astonishing agreement has existed over time in the scores and in the messages both tests have been sending.
This first graph shows the EXPLORE and NAEP math results. NAEP math results for each school term are shown by the blue bars and the EXPLORE Benchmark results by the red bars.
Let’s first discuss the trend in the NAEP. To begin, just as we discussed with the NAEP Scale Scores in an earlier blog, there is statistical plus and minus sampling error in all the NAEP data in the graph. In fact, according to the 2015 NAEP Report Card supporting Excel spreadsheet for math scores, there is no statistically significant difference between the 2015 NAEP score and any earlier NAEP score shown.
Still, the movement in the most probable NAEP scores, as shown by the bars, was a very small improvement in scores between 2008-09 (which covers the NAEP test in 2009) and 2010-11 (Which covers the NAEP test for 2011) followed by a very slow decay in performance over the next two NAEP test administrations.
However, to reiterate, the changes in the NAEP are below the “noise level” of the statistical sampling errors in the scores. So, let’s look at more refined data from EXPLORE, which has no sampling error because every Kentucky eighth grader takes this assessment.
First, note that the EXPLORE results for each year are quite close to the results from the NAEP. Within the NAEP’s sampling error, the two tests are generally showing the same thing.
Next, note that while EXPLORE rose slowly from 2006-07 to 2012-13, it also shows a decline for Kentucky in the latest 2014-15 school term. In fact, the 2014-15 EXPLORE result is no different from the score back in 2010-11, one year before Kentucky began testing with the Common Core aligned Kentucky Performance Rating for Educational Progress (KPREP) tests.
In fact, the picture from EXPLORE is even more unsettling than the graph shows. While NAEP recently has only been given in late winter in odd-numbered years (2007, 2009, 2011, 2013, 2015), EXPLORE has been given every year since the 2006-07 school term. In the 2009-10 term – an off year for NAEP, so not shown above – Kentucky’s EXPLORE Benchmark Score peaked at 35.6 percent. EXPLORE performance has been lower every year since. The 2009-10 school term just happens to be the same year that the state adopted, but had not yet implemented, Common Core.
So, if we examine the more extensive and accurate information (no sampling error) in EXPLORE math, it refines what the NAEP also tells us about this key subject. Kentucky has not fared well, so far, in the Common Core era.
Let’s see what happened reading. This next graph shows that story.
First, the NAEP scores are statistically flat from 2008-09 to the 2014-15 year. NAEP can detect no real change in performance for Kentucky from the year prior to Common Core adoption to the present.
The more refined data from EXPLORE shows some reading performance did occur between the early days of use of that test and 2012-13. However, just as with NAEP, the reading performance of Kentucky’s eighth grade students reversed in the new numbers for the 2014-15 school term. Between 2010-11 and the present the EXPLORE shows only the barest detectible increase in performance.
So, the message from the new NAEP results and the somewhat more precise EXPLORE tests is largely the same. To the present, Kentucky’s performance in the Common Core era has so far been flat, at best. Furthermore, both tests indicate a decay in performance occurred between 2012-13 and 2014-15.
NAEP scores come from the Excel spreadsheets that were released in the NAEP Report Card online resources on October 28, 2015. These are available from pulldown menus found in the “Data Quick View” section at the bottom of this web page.
The EXPLORE Benchmark Scores can be accessed from the Supplemental Data, “Assessment” section of the Kentucky Department of Education’s Open House portal by clicking here.