More on those dubious graduation rate statistics

— What the Feds are reporting

I wrote on New Year’s Day about a news report that the pending release of more accurate high school graduation rate data worries our educators.

New data from the federal government shows Kentucky’s educators – in fact, all of us – should be worried.

The blue line on this graph shows high school graduation rates calculated by the US Department of Education. The Feds use a carefully researched formula called the Averaged Freshman Graduation Rate (AFGR). These good quality AFGR numbers are compared to the inflated rates (officially confirmed in a 2006 audit) that the Kentucky Department of Education (KDE) has been reporting (shown by the red line).

As you can see, KDE’s numbers run around seven to eleven points higher than the better quality federal data, depending upon the year.

Also note that the most recent AFGR shows Kentucky’s high school graduation rate DROPPED by eight tenths of a point from the previous year although our state educators reported the rate had slightly increased in that time.


Based on other data just released by the US Department of Education, I revised my estimate in the last blog of the number of kids that “fall through the cracks” thanks to the KDE’s inflated graduation rate reporting.

Given the US Department of Education’s estimate that Kentucky’s graduating class of 2007 had 51,173 first time students when it entered ninth grade, and given the different graduation rates reported by this federal agency and the KDE, it looks like around 3,700 students just vanished from the KDE’s calculations for the 2006-07 year.

These kids didn’t graduate. They just vanished from the KDE’s loose data system.

That’s a lot more kids who failed – kids our state educators don’t seem to want to admit even exist.



Officially, the latest Nonacademic Data Report from the KDE says there were only 6,175 high school dropouts in the 2006-07 year. I have heard all sorts of Kentucky government officials from the governor to the chair of the Kentucky Board of Education cite those sorts of figures.

However, since another 3,700 kids apparently “fell through the cracks,” it looks like the true number of Kentucky dropouts was about 60 percent higher. We are losing almost 10,000 kids from each single high school class who don’t make it through the system on time – perhaps not ever.

What is very sad about all of this is that Kentucky’s educators have steadfastly resisted providing the public with more accurate data ever since KERA began. They seemed happy to continue distributing inflated numbers.

It’s not like our educators weren’t being told.

The Kentucky Office of Education Accountability and I separately presented strong evidence that the KDE’s numbers were off in the October 2003 meeting of the legislature’s Program Review and Investigations Committee. That evidence created such serious doubt in legislators’ minds that the committee immediately voted, unanimously, to request an audit of the KDE’s dropout and graduation rate figures.

Next, recognizing that Kentucky’s official graduation rate reporting was lacking, Governor Ernie Fletcher committed the state in 2005 to develop a high quality student tracking system that would enable us to use a very accurate graduation rate calculation that is only available to states with such databases.

In October 2006, the Kentucky Auditor of Public Accounts completed an audit that confirmed considerable error in the KDE’s official dropout and graduation rate data. In response to the audit, the Kentucky Department of Education indicated it was only a few years away from reporting with the high quality data system Governor Fletcher had agreed to create.

Around the same time in 2006, the US Department of Education issued a set of technical reports titled, “User’s Guide to Computing High School Graduation Rates,” Volume 1 and Volume 2. These reports showed the Averaged Freshman Graduation Rate was the best graduation rate formula available for use in states like Kentucky that lacked the high quality student tracking systems needed to generate really accurate graduation rate data. That research also showed the formula in use by Kentucky can produce inflated errors of as much as ten points.

Basically, both the evidence we had problems and the tools needed to do better while we waited for a high quality student tracking system to come on line were all available by the end of 2006.

It’s now more than three years later. The Kentucky Department of Education’s student tracking system is years behind, and we won’t see any graduation rates from that system before 2013, at the earliest.

In the interim, the department is finally going to move to the Averaged Freshman Graduation Rate calculation – in 2011 – but it seems this is happening mostly because the US Department of Education’s patience with the states has run out. Now, the Feds are requiring states to finally get honest.

In any event, this sorry example of foot-dragging is not a success story for KERA. It also raises strong questions about the responsiveness of the KDE to official audits.

I must note that with our new, data-oriented Commissioner of Education Terry Holliday on board, things are looking up in the area of more accurate testing and reporting.

However, KDE’s Dr. Holliday has some tough legacies to overcome. One is the history of test score inflation with our old CATS assessments which ultimately led to the demise of CATS. That legacy is matched by the department’s equally disconcerting history of inaccurate graduation rate reporting.

Holliday has his work cut out to get all of this right, but our kids deserve no less than a decent testing system and accurate dropout and graduation rate reporting that leaves no child behind – or out, either.

Data sources used to create this blog:

AFGR to 2005-06: Digest of Education Statistics 2008, NCES 2009-020, Table 106 On line here.

AFGR for 2006-07: NCES Common Core of Data State Dropout and Completion Data File: School Year 2006–07, NCES 2010-312, Appendix B, Table B-1. On line here.

Kentucky's Reported Graduation Rates:
“BRIEFING PACKET, Nonacademic Data, Dropout, Retention, Transition to Adult Life, Attendance and Graduation Rates, 1993 to 2008, State Totals,” on line here.