(Guest Post by Matthew Ladner)
I’ve been digging into the very impressive NAEP scores of Colorado charter schools. NAEP is the gold-standard of academic testing data, but it does make use of sampling whereby a representative sample of students in each state takes the NAEP rather than every single student. Charter school students in Colorado are now more numerous than say left handed kids with blonde hair and blue eyes, and each NAEP example involves a separate sample of students. So a single sky-high score on one of the NAEP exams for charter school students could easily be the product of a favorable sample, solid achievement across all four tests instills confidence.
Still it is good to examine other testing data that does not involve sampling. The Colorado Department of Education released a 2016 State of Charter Schools Triennial Report that contains state testing data comparisons for charter and district schools. This data addresses a different question than what we have examined in the previous two posts. Previous posts have demonstrated that Colorado charter schools show top-notch performance across a variety of subgroups when compared to high-performing states. The data below uses state data to address how Colorado charter students scored compared to Colorado district students.
These data have further limitations. The report notes them as “preliminary” and we do not have access to the raw data- only the percentages of students meeting for exceeding grade level benchmarks. The Arizona Charter School Association analysed state data and determined that the state data displayed an even larger gap between charter and districts than the NAEP gaps. Without analysis however, we can’t make comparisons between state and NAEP data. Note also that judging by NAEP that Colorado’s district system is relatively high performing itself compared to other states.
With those caveats in mind, the state data shows a consistent advantage for charter students in math:
And in English Language Arts:
The report contains some breakdowns by student subgroups (family income, ethnicity) on both the current and previous state exam. These comparisons broadly favor the charter schools. The report also presents aggregate data on schools by type under the state’s accountability framework:
For those squinting at their Ipads, the last four columns basically show that a smaller percentage of Colorado charters fell into the “Does Not Meet/Approaching” category (40.52% to 52.07%) than districts, and a larger percentage made either “Meets/Exceeds” (59.48% for charters, 48.93% for districts).
The overall conclusion of these data is not that Colorado charter schools are “good” while Colorado district schools are “bad” whatsoever. Good or bad for whom? The good news for Coloradans is that you have a system of independent public schools operating at scale and producing on average world-class academic results.
Keep it up- in the concluding post I’ll show how Colorado’s need for highly skilled/educated workers is set to grow over time.
Jay P. Greene's Blog 