High School Graduation Rates in Kentucky
High School Graduation Rates in Kentucky
Highly Accurate Data Not Yet Available
Although the Kentucky Education Reform Act of 1990 has now been in force for nearly two decades, the state still cannot provide high accuracy graduation rate data.
The current graduation rate calculation officially in use by the Kentucky Department of Education was officially audited in late 2006 by the Kentucky Auditor of Public Accounts and found to have significant errors. Dropouts were being under-reported and graduation rates were inflated. In addition, the audit disclosed that students who dropped out over the summer between school terms were not being captured in the student tracking software then in use.
The Kentucky Department of Education is working on a high quality student tracking system, but there have been missteps along the way and the current program, called “Infinite Campus,” is just now becoming operational (Spring 2009) in all schools around the state. This system must be in use for four years before it can generate the first, high quality graduation rate information for the state. According to Appendix B in the May 19, 2009 Nonacademic Data Report from the department, at present the first set of high quality graduation rate data won’t be available until the 2013-14 reporting year.
In the interim, Kentucky must rely on estimation formulas to report high school graduation rates, and this Wiki item examines how various commonly encountered formulas perform.
What Is a Successful Graduation?
The No Child Left Behind Act of 2001 stipulates that the graduation rate used for federal school accountability shall only credit students who graduate with a regular high school diploma in the standard number of years. Students who take more time, or who get a certificate rather than a diploma, are not to count. This requirement is not consistently met with the calculations discussed in the next section.
Graduation Rate Formulas Cited in Kentucky
There are actually a wide number of different formulas that are in current use to estimate graduation rates for Kentucky. Below are some of the more frequently encountered formulas
Freshman Graduation Rate
This is one of the simplest formulas. It is the ratio of graduations to the 9th grade enrollment of the class.
The formula generally reports rates about two percent lower than actual according to an extensive study described in Volume 1 and Volume 2 of a National Center for Education Statistics report titled "User’s Guide to Computing High School Graduation Rates." However, despite this limitation, the Freshman Graduation Rate has the most extensive set of data over time and can be calculated even for years well before KERA began in Kentucky.
The Bluegrass Institute calculates this formula in a slightly different way from the recommendations in the User's Guide, using the total number of diplomas and certificates awarded each year (rather than only the on-time diplomas) as published by the Kentucky Department of Education in the annual Nonacademic Report under the “Transition to Adult Life” section as “Grads.” The 9th grade enrollment comes from the Department of Education’s fall membership counts which are contained in annual “Growth Factor Report, Ethnic Membership By District And Grade” Excel spreadsheets.”
The Prichard Committee for Academic Excellence has also used this type of formula. For example, one of the Appendix items in their November 1999 report, "Gaining Ground: Hard Work and High Expectations for Kentucky Schools" was written by Thomas G. Mortenson (not available on line) and uses this formula, which Mortenson calls a "Cohort Survival Rate."
A criticism of the Freshman Graduation Rate formula is that it “double counts” 9th grade students who have been retained from the previous year’s class in the denominator. Thus, the rate returned is somewhat lower than the true rate. Another criticism is that most practical implementations of the formula count all diploma and certificate awards. The second criticism tends to offset the error from the first criticism.
The Freshman Graduation Rate formula as the Bluegrass Institute calculates it can also be calculated over a long period of time. That is not true of some other formulas described below, which require separate calculation of on-time versus more than standard time diploma awards. Such data is only available for Kentucky for the Class of 2003 and later.
The Freshman Graduation Rate using only NCLB eligible graduates can only be calculated in Kentucky for the Class of 2003 and later.
NCES Averaged Freshman Graduation Rate (AFGR) Formula Using KY Data
The AFGR has been extensively researched by the US Department of Education in several states which already have high quality student tracking systems and high accuracy graduation rate information, which allows careful research on formula accuracy. That is discussed in Volume 1 and Volume 2 of the "User’s Guide to Computing High School Graduation Rates."
The numerator of the formula used here is the sum of the number of regular high school diplomas awarded to students who complete high school in the standard number of years (4) plus the number of regular diplomas awarded to students with disabilities within the time listed in those students Individual Education Plans (IEP). Students who take more than four years to complete their diploma work and students who get certificates rather than diplomas are not counted. The data source for the numerator is found in an Excel spreadsheet titled “Graduation Rates by District and High School (Five year trend, 2008).”
The denominator is an average of the membership of the class in grades 8, 9 and 10, adjusted for those students who are severely disabled and not formally assigned to a specific grade. The source of this data is the “Growth Factor Report, Ethnic Membership By District And Grade” Excel spreadsheets.”
NCES AFGR From Digest of Ed. Stats. 2008
The National Center for Education Statistics (NCES) provides a calculation of the AFGR through 2006 in the “Digest of Education Statistics 2008.”
Data source for the calculation is the NCES Common Core of Data, a large database of school information including graduation and enrollment figures.
There is a small problem with the NCES calculation. Comparisons of the NCES graduation data to that provided by Kentucky in its own reports indicate that in recent years the NCES graduation counts include all students who get a diploma or certificate in a given year regardless of time required. That makes the NCES AFGR calculations for Kentucky somewhat higher than the true, on-time graduation count should be for NCLB purposes. It is unknown if similar problems impact the NCES calculations for other states.
Transitional AFGR with KDE Formula
This is the new formula that the Kentucky Department of Education intends to use until it can get the high quality Infinite Campus reports on line. Read Averaged Freshman Graduation Rates in this Wiki for more information and examples.
The figures reported here were calculated by the Bluegrass Institute. The numerator is the number of on-time diplomas plus the IEP diplomas awarded within the time specified. The denominator, which is supposed to be first time freshmen students, is calculated by subtracting the number of students who were retained (held back) from the previous year’s ninth grade class from the ninth grade enrollment when the graduating class was in ninth grade.
Prichard Committee's Grads/Grade 8 Membership
This calculation is very similar to the Freshman Graduation Rate calculation except that it uses the number of eighth grade students who were tested in the state’s CATS assessment program instead of ninth grade enrollment for the denominator. See the Prichard Committee's May 19, 2009 Blog item "Graduation improves, mysteries continue" for more.
The intent of this calculation is to overcome the issue of double-counting ninth grade students who were retained from the previous year. However, the formula has other short-comings. It fails to account for the fact that a number of students who attend private schools from Kindergarten to grade eight in the state then transfer to public schools in grade nine. The formula also counts more than four year graduates as a success, which NCLB specifically prohibits.
KDE Officially Reported Rates
Since the inception of NCLB, the Kentucky Department of Education has used a fairly involved formula to report graduation rates. A full description of this formula is found in Appendix B of the May 19, 2009 Nonacademic Data Report.
The main shortcoming of the formula is that it is critically reliant on accurate dropout data, and as previously mentioned, that data isn’t reliable in Kentucky.
In fact, dropout data across the nation has proved to be highly unreliable unless it is collected with a high quality student tracking system, which Kentucky is just now bringing on line.
The list of formulas above is not exhaustive. An interesting further discussion on this topic is found in “Calculating High School Graduation Rates” from Strive.
Comparison of Results – Graduation Rate Calculations for Kentucky
This graph shows the graduation rates as determined by several formulas in common use today.
In reviewing this graph, keep in mind that the NCES AFGR rate, shown in light blue, and the Freshman Graduation Rate, in dark blue, have the most extensive research behind them. That research shows the AFGR generally reports figures only about one point or so higher than the true NCLB rates while the Freshman Graduation Rate reports results about two points below the true value.
Also notice that the AFGR calculated by the NCES from the Common Core of Data has been getting somewhat larger over time in comparison to the AFGR the Bluegrass Institute independently calculated using data directly reported by the Kentucky Department of Education. Again, the reason for this seems to be that Kentucky may not have reported graduation counts to the US Department of Education in a consistent manner. In the early years of NCLB, the Common Core of Data graduation counts appear to only include on-time and IEP-on-time regular diplomas awarded. More recently, the count seems to include all diploma and certificate awards, which isn’t consistent with NCLB graduation rate requirements.
The Transitional AFGR calculated by the Bluegrass Institute seems to track the NCES AFGR as calculated by NCES fairly well, especially in 2005 and in the last year of available NCES data for 2006.
Clearly, the Prichard calculation and the official Kentucky Department of Education calculation provide the largest graduation rates. Because the official Kentucky calculations have been audited and are known to be inflated, they should be considered unreasonable.
In any event, it looks like somewhere between one in five to one in four students in Kentucky isn’t graduating on time, if at all. That is an expensive problem that we must learn to address better.