The information in the Florida Department of Education’s Tax Credit Scholarship Accountability Document: How can it be made more accessible?

If you haven’t yet read the Orlando Sentinel article “Private schools’ curriculum downplays slavery, says humans and dinosaurs lived together”, you should.  You should read it especially if you are a school choice supporter or (like me) a school choice agnostic – because you really need to think about the challenging questions the article raises and whether there should be academic boundaries on tax credit scholarship schools.  This post is an (admittedly technical) exploration one of the many questions raised by the Sentinel article.  I will be exploring other such questions in future posts.

When it comes to academic accountability, Florida’s Tax Credit Scholarship Program relies entirely on parent choice.  (Although if someone would like to correct me on that, please do.)  That means that the parents whose students are in that program or who are considering entering the program need easy access to information about content of the curriculum and the academic achievement of students.

The information about academic achievement comes from nationally norm-referenced tests given to every tax credit scholarship student.  (Tax credit scholarship students do not take the state’s homegrown tests that are taken by students in all public schools – traditional and charter)  When a particular tax credit scholarship student takes a test at a school two years in a row, the measurement of a learning gain is possible.

As far as I am aware, the FLDOE report titled “Evaluation of the Florida Tax Credit Scholarship Program Participation, Compliance and Test Scores in 2015-16” is the best source for academic achievement information on individual schools.  The 2015-16 report was released in June 2017 (which means a 2016-17 version is coming out soon) and was composed by five researchers from Florida State University’s Learning Systems Institute.  One of the authors, Carolyn Herrington, is one of my heroes.  I have confidence that the report was written with integrity.  You can trust what it says.

Information on learning gains in math and reading for the 283 tax credit scholarship schools that provided both 2014-15 and 2015-16 testing data for 30 or more students (that is, information on testing data for both years for the same 30 or more students so that learning gains could be determined) is located in the appendix of the report.  In total, there were (according to the report) “1,330 private schools that had participating students in grades 3 through 10 during the 2015-16 school year”.  Most had too few students who had attended in 2014-15 and 2015-16 to provide 30 measured learning gains.

Is the present report format the best way to provide this information to parents?  I don’t think so.  I have little or no creativity, but even I can imagine a web site in which a parent can key in the name of a tax credit scholarship school and come up with the learning gain information (or the fact that there is no learning gain information available) and information about the curriculum (or curricula) used at the school.  Or key in a zip code and get a list of tax credit scholarship schools within 20 miles.  Or something like that.

Perhaps there was something in this spring’s school choice legislation (HB 7055) that addresses this information accessibility issue.  If so, I haven’t been able to track it down.  If you know of such, please let me know.

The report is linked here:


A page from the report’s appendix showing learning gains for ten schools (the names of which all start with the word “Saint”) is shown below.  The explanation of the table from the report is here:

Appendix Table: Average gain scores in 2015-16 and three-year moving average of gain scores from 2013-14 to 2015-16 for schools with 30 or more students with gain scores in 2015-16.  Notes: Cells report average gain scores. We shade cells where the difference between an individual school’s three year moving average gain score is statistically significant from the national average (at the 95 percent confidence interval).

Example:  In the table section below, three schools have learning gains (averaged over three school years – 13-14, 14-15 and 15-16) that differ significantly from the national rate.  The school listed on the top line of the page, Saint Agatha School in Miami, has learning gains in both reading and math that are significantly below the national average gains.  Saint Andrew Catholic School in Orlando, which was visited by President Trump and Secretary of Education DeVos last year, has a math learning gain that is significantly above the national average gain.  The reading gain is not different from the national rate by a statistically significant amount.  Saint Bartholomew School in Miramar has an average learning gain in math that is below the national average gain by a statistically significant amount.


In general, the ability of the analysis to discern learning gains with statistical precision is hindered by the small number of students for which learning gains are available, as can be seen in the table section above.

This entry was posted in Uncategorized. Bookmark the permalink.