Beware New Quality Counts State Rankings

Education Week just released the final segment on its 2019 ranking of state education systems, and it is unfortunate that this generally pretty good news outlet continues to mess this up.

Point of Order: I just ranked Kentucky’ s All Student 2017 NAEP Grade 4 Reading Scale Score using the NAEP Data Explorer’ s tools. The state ranks 17th, not 22nd. But, that sort of simplistic ranking is misleading because NAEP is a sampled assessment and there are sampling errors in all the scores. Once you allow for the statistical sampling errors in NAEP (you can also do this with the NAEP Data Explorer), Kentucky’ s fourth graders were statistically significantly outscored by 19 states, tied 16 states, and scored statistically significantly higher than 15 states and the DC school system. So, Kentucky placed somewhere in a rather large and vague middle of the pack, but claiming it outscored a number of states with somewhat similar Grade 4 reading scores is not valid. Here’s another point: Kentucky still has a very high enrollment of white students in its public schools, much higher than many other states. By only ranking overall scores for all students, Education Week actually ranks a lot of Kentucky white students against minority students in other states. That isn’t valid, either. The detailed rankings attempted by Education Week do not honor the sampling error in NAEP and create unsupportable images of actual relative performance.

By the way, I found out that Quality Counts ranked proficiency rates, not even the more accurate NAEP scale scores. Very disappointing.

This entry was posted in Education journalism, Education policy, K-12, Richard Innes, Testing/Assessment and tagged , , , . Bookmark the permalink.

Leave a Reply

Your email address will not be published.