The graphs below give me increasing confidence about that. The first compares 1993 fourth grade enrollment to 2001 graduates, 1994 fourth to 2002 graduates, and so on. The others use sixth, eighth, ninth, and tenth grade enrollment the same way.
If you're wondering why ninth grade shows the smallest percentages, it's because the ninth grade enrollment is always the state's largest. That's partly because of transfers from non-public schools, and heavily because more students repeat grade nine than repeat any other year.
Starting from any grade, graduations look better now than they did a few years ago.
(Notes for readers who love the details: The numbers behind the graphs are below. Earlier posts compared students taking the CATS test to students graduating. The up-side of that method is that it’s nearly impossible to over-count tested students. One down-side is that the CATS data won’t let me reach back to 1993, which is why I used enrollment numbers for this particular post.
Do notice the basic logic of the data. In round numbers, 40,000 graduates out of 50,000 students will always divide out to 80 percent. Any claim that Kentucky public schools have a graduation rate much below the 80 percent mark must show fewer than 40,000 diplomas, more than 50,000 enrollment or both. For recent years, that will be a tough claim to sustain without using the bulging, improbably high, ninth-grade numbers as a starting point.)
Susan,
ReplyDeleteThis is interesting, but what where the rates back in the early 1990s and before? I have heard that we have mostly just recovered back to where we were then.
Also, the KDE just announced a new formula will be used for graduation rates in a year or two. Can that be calculated now?
Anon,
ReplyDeleteKDE has announced both a transition formula and a final method.
The final method will track students from entering grade nine for the first time, right through to graduation, factoring in those who transfer and out. The new Infinite Campus system will make that possible, by using a "unique identifier number." The tracking will start with this fall's freshmen, so we'll have the first figures in 2013 when they graduate.
KDE recently described the transition method by writing "The Averaged Freshman Graduation Rate (AFGR) takes the number of on-time graduates plus the number of graduates that have extended time written into their IEPs divided by the number of first-time 9th-graders who started three years earlier."
I'm uneasy about that for two reasons. First, my understanding has been that we couldn't separate out the first-time 9th graders. Second, when I've read the term "Averaged Freshman Graduation Rate" elsewhere, it's meant that you divide by the average of ninth-graders, the previous year's eighth-grade, and the next year's tenth grade.
My working hunch is that the Department actually plans to do it that three-year-average way. If so, the math can be done now. If not, then I don't understand how the math can be done before 2013.
Anon,
ReplyDeleteHere's a separate answer to your question about older history.
I haven't seen any KDE publications that go back before the 1993 enrollment numbers. At a guess, there probably are some in the state archives.
There may also be numbers available from the federal common core of data, but I'm wary. For years and years, Jefferson County submitted fall numbers that counted transferring students more than once, and I don't know how far back that error goes. The publications I've seen from the National Center for Education Statistics use fall data for enrollment, so I avoid those.
Finally, there are Census numbers on adults 18-24 with a diploma or equivalent:
• 69.5% in the 1980 Census
• 75.2% in the 1990 Census
• 74.9% in the 2000 Census
• 82% in the three-year-average of the 2005, 2006 and 2007 American Community Survey.
Those numbers include public diplomas, nonpublic diplomas, and GEDs, but they still make it hard to explain how the the public school rate was higher in 1980 or 1990 or 2000 than they are today.
The decline you heard about might be that 0.3 drop from 1990 to 2000, but the ACS results have shown us well above the 1990 level at least since 2004. Your source may have been a bit out of date.