October 26, 2011
DIFFERENCES IN GRADUATION RATES FOR THE CLASS OF 2008
By: Judith Lohman, Assistant Director
You asked for an explanation of the differences between Connecticut high school graduation statistics for the Class of 2008 published by the State Department of Education (SDE) and the Connecticut Coalition for Achievement Now (CONNCan), an education reform advocacy organization.
The differences in the two sets of statistics derive from the different methods used to calculate them.
CONNCan's issue brief Connecticut Graduation Rates, published on October 3, 2011, included statewide and district-by-district graduation rates for the Class of 2008, the most recent class for which final rates are available. CONNCan highlights large differences between the graduation rates included in its report and more favorable graduation data published by the SDE. According to CONNCan, the statewide graduation rate for Class of 2008 was 79.2%, while SDE data shows a 92.1% rate.
CONNCan's report uses graduation rates published in Education Week. Education Week's rates are calculated using the cumulative promotion index (CPI) method. The CPI method is based on the percentage of students in a particular class or “cohort” who are promoted to the next higher grade each year starting in grade nine and graduate with a diploma at the end of four years.
SDE's method, on the other hand, is to estimate the graduation rate for a particular class from a combination of (1) four years of annual school dropout data for that class and (2) self-reported graduation information from students. SDE admits its method contains errors and duplication. Starting with the Class of 2009, SDE will base graduation rate calculations on its new system of tracking individual students and the CPI formula. SDE's preliminary graduation rate for the Class of 2009, derived from the new method, shows a 79.3% graduation rate for that class, which is close to the rate for the Class of 2008 published in CONNCan's report.
GRADUATION RATES REPORTED BY CONNCAN
In its October 2011 report, CONNCan explains that its source for graduation rate data is Editorial Projects in Education (EPE). EPE publishes Education Week among other weekly and monthly education publications, and annual reports on states' educational performance based on analyses of education data. In the 2011 edition of its annual report on graduation rates, called Diplomas Count, EPE presented a new analysis of graduation rates for the Class of 2008 based on school and district data from the U.S. Education Department's Common Core of Data (CCD). The CCD is an annual census of schools and school districts in the U.S.
EPE used CCD data and the CPI formula to calculate graduation rates for the Class of 2008. This formula follows the progress of a high school class through each high school grade to determine overall percentage of the class graduating from high school after four years with a regular diploma.
The calculation starts with the number of students in the class in the 9th grade. It then computes the percentages of students in the initial 9th grade cohort that are promoted from the 9th to the 10th grade, the 10th to the 11th, and the 11th to the 12th, as well as the percentage receiving a regular high school diploma after grade 12. The four promotion percentages are then multiplied together to arrive at the CPI for the class, which is also its graduation rate. The CPI formula excludes students who drop out; repeat a grade in high school; or receive a GED, certificate of attendance, or other non-diploma credential at the end of the 12th grade.
GRADUATION RATES PUBLISHED BY SDE
SDE publishes statewide and district-by-district high school graduation rates on its Connecticut Education Data and Research (CEDAR) website. According to a 2010 press release, these graduation rates are department estimates based on annual school dropout data and graduation information reported by students themselves. The press release admits that the SDE's method is less accurate and contains errors and duplications.
A November 2009 SDE Data Bulletin explains how the department calculates school dropout rates.
The SDE calculates both an annual and a cumulative dropout rate for each high school class using coded data on each student reported by school districts through the statewide public school information system (PSIS). The department counts students as dropouts if, according to the PSIS, they:
1. discontinue schooling;
2. leave their school, do not return, and no record of a transfer to another public or private school is available;
3. are on a class roster from School A to attend School B but never report to School B and no transfer information is available; or
4. leave school to enroll in a non-high-school-credit-earning training program, including GED classes.
SDE calculates the annual dropout rate by dividing the total number of students in a particular high school class that fall into the four categories listed above by the overall public school enrollment in grades 9-12 as of October 1. It calculates a cumulative dropout rate for a class by adding up the numbers of annual dropouts in each year over four years and dividing the total by the number of students in the class who were enrolled in the 9th grade as of October 1 of the class' 9th grade year.
Graduation Rates for the Class of 2009 and Beyond
Starting with the Class of 2009, SDE will use its new individual student tracking system to calculate graduation and dropout rates. It will also use the CPI method described above to calculate graduation rates. In conjunction with the 2010 release announcing this change, SDE issued a preliminary graduation rate for the Class of 2009 of 79.3%. This figure is almost identical to the 79.2% rate for the Class of 2008 calculated by EPE and reported by CONNCan. It is also much lower than the rates SDE reported for previous high school classes.