Commentary: What’s really behind Connecticut’s increasing graduation rates?

Print More

In Connecticut, official high school graduation rates are on the rise — to 87 percent in 2014 from 79 percent in 2009. This trend holds true for all demographic groups.

But do these numbers tell the whole story?

The main issue is that, in these calculations, there are plenty of students who disappear — neither counted as graduates nor as someone who didn’t graduate. For example, they could’ve been held back a grade or transferred to another school before graduation. The state lists more than 20 different reasons for students to be removed from or added to a cohort.

But when trying to track these details, the state won’t provide this data, even in a way that would protect student privacy. (I’ve asked for it.) And this means there are a lot of unknowns.

At Amistad Academy charter school, 37 students started the ninth grade in 2008. By the end of senior year, there were 15 graduates. Using simple math, that should mean the graduation rate is 41 percent. But according to the state and federal departments of education, the official graduation rate was 71 percent.

How can one calculation get us to 41 percent, but another get us to 71 percent?

In 2011, the federal government began requiring all states to report graduation rates in a uniform way. The method that the U.S. Department of Education adopted was called the “Adjusted Cohort Graduation Rate,” or ACGR. The idea was supposed to be simple:

  1. Count the students who enter the ninth grade and follow the same group over time.
  2. Remove students who transfer out.
  3. Add students who transfer in.
  4. Finally, divide the number of students who graduate in four years by the number of kids remaining in this group. Anyone who doesn’t officially transfer out of the cohort is considered a non-graduate.

So, back to that high school: The data tells us that 18 of the original 37 students were removed from the calculations; they either transferred to another school or left the state. There is no mention of these students in the graduation rate.

After a few students transferred in, the school officially had 21 students in the cohort — 15 of whom graduated.

Comparing the methods
At Amistad Academy charter school
method graduates students graduation rate
Simple method 15 37 41%
New State/Federal method (ACGR) 15 (37 – 18) 71%
A lot of hidden information

The state and federal government say their method is a better of way of counting graduation rates. Up until this point, the state used a variety of methods to determine graduation and “dropout” rates. All of these methods had weaknesses, as does this new one.

But the current way the state and the feds calculate graduation rates hides quite a bit, and we may not accurately know the outcomes for entire racial and ethnic groups. For example, the official method says there were hundreds of Latino students in the class of 2012 who graduated in four years — and hundreds who did not. (See chart.) But this method makes it possible that many more students did not actually earn a diploma.

What we know is that for every 2012 Latino graduate in Hartford, there was another who dropped out and another who disappeared from the original cohort. Where did these Latino high school students go? Did they repeat the ninth grade in the previous year? Did they transfer out of their school or the district?

State data on transfers would help us understand what’s going on here and where we might be able to better support these kids. Even simple charts, like the one above, required three different data sources to construct because the state does not report its graduation rate with these details.

Cause of increased rates: unknown

As National Public Radio recently reported, there are a variety of ideas about why graduation rates are increasing — but we really don’t know. In fact, every state and demographic group in the country is posting increased graduation rates, including Connecticut. So whatever is happening is a national trend.

Plenty of people and groups are taking credit for the graduation rate increase. Gov. Dannel P. Malloy says his education bill, which took effect in 2012, caused the increase. But graduation rates have increased since before Governor Malloy took office in 2011, and in many cases the increases in the graduation rate have appeared to get smaller each year since 2009.

Other unaccounted factors

Finally, there are a number of unanswered questions.

We don’t know how schools and districts have responded to the new graduation rate method. The federal government required the use of graduation rates as a form of rating schools in 2012. Do schools now have an incentive to graduate more students, one way or another, thus distorting this measure?

Reports out of Texas show that thousands of students have been counted as “other leavers” rather than “drop-outs,” thus inflating their graduation rate. Education researcher Julian Vasquez Heilig has frequently raised questions about this clever labeling of students that can distort graduation rates.

We don’t know the impact of “online credit recovery,” which is a new way for students who did not pass a regular high school course the first time to take the same course again — but online. Beginning in the 2010–11 school year, the Connecticut legislature began requiring many low-income school districts with low graduation rates to offer “online credit recovery.” In Connecticut, the students in these districts are often Black, Puerto Rican and Latino. As the Mirror reported in 2012 (Part I and Part II), students can now get these credits from the online course toward a regular high school diploma. Did this make it easier to obtain high school credits towards a diploma through online courses in these districts?

So the next time you read about increased graduation rates, use caution. The graduation rates show as much as they hide.

This essay is based on a post entitled, “What We Don’t Know About CT Graduation Rates: 5 Question originally posted at The Cities, Suburbs, and Schools Project site.

Robert Cotto, Jr. is a member of the Hartford Board of Education and a lecturer in the Educational Studies Program at Trinity College.

We welcome contributors to share data, analysis and perspectives that follow our guidelines.

What do you think?

  • “The main issue is that, in these calculations, there are plenty of students who disappear — neither counted as graduates nor as someone who didn’t graduate. For example, they could’ve been held back a grade or transferred to another school before graduation.”

    This statement doesn’t seem to comport with the linked document which states on P. 11:

    “What happens to students retained in a grade?

    The student’s cohort does not change when he/she is retained in a grade. Therefore, the
    student does not count as a graduate in the four-year rate unless he/she catches up and
    graduates on time.”

    Therefore I’m not really sure why you think its possible someone could disappear from the cohort by repeating 9th grade. Is there something I’m missing or misunderstanding?

    Also I do not think your simple method makes sense because schools would be penalized for the unknown status of students who exit the system. If we’re trying to measure the impact of schools and their efficacy counting students who spent part of their time outside the school could mess things up. I do not think we should count students who transfer in or out in these calculations at all. Maybe there could be some kind of weighted attribution (i.e. a student transfers from Public School A at the conclusion of her sophomore year and graduates on time from Public School B. the numerator and denominator of public school B include 0.5 of a student on each end. Similarly If the student fails to graduate Public School B on time then both Public School A and B have 0.5 added to their denominator).

    Interesting point about online credit recovery. It’d be disconcerting if districts were relaxing their standards to increase graduation rates, but I think that even in the suburban school I went to there is a lot of inertia against holding back a promotion to the next grade. If the teacher and student are willing to work together they are usually able to find a way. The only way I could see enforcing a rigid grade promotion standard would be some kind of standardized exit test for each grade.

  • Robert Cotto Jr

    Hi Matt- You are correct that kids that are retained are removed from cohorts. I wanted to be sure that readers noted that when we compare the “unmatched” to the “matched” cohort of original 9th graders, the difference can be made up of kids that transferred out or stayed back from previous years. The key is that we don’t know and the State won’t provide that data. Thanks for pointing out that this part may have been confusing.

  • Maria Pereira

    Robert, last year Bassick High School in Bridgeport attempted to graduate 26 students through credit recovery that did not meet the requirements to graduate. Interim-Superintendent Fran Rabinowitz was informed, and she had to inform all of their parents and the students that they could not graduate just two days before the graduation ceremony.
    How many times do you think that was done during Paul Vallas’ reign of terror?
    The principal was demoted and has since resigned.

  • readdoctor

    I have always wonder why one of the statisitcal calculations statewide is not how many ninth graders we have have statewide, and how many graduates we have four years later. All states have these numbers. Yes students transfer, lack of enough high schools credits for graduation, and might move out of state. Districts have used fuzzy math to determine HS graduations rates across the country for years. We all remember the Texas Miracle that brought us NCLB. Former Sectretary Rod Paige lost over 15,000 drop outs in Houston when he was superintendent. The only reason people discovered his fuzzy math was parents calling up the local news media saying I have a drop out sitting in my living room right now.
    Fuzzy math hiding drop outs will continue as long as we are not asking people to compare ninth grade cohort size to graduation rates. Doing so will not tell the whole story, but will start us asking the right questions. This data is available, and could be collected as a state as a whole. Doing this as a state would help take in to account transfers with in state as well. It also would help hold the state as a whole accountable. Now that is something I have ever seen in all my years living in Connecticut.
    Every state should be looking at all the data they have on graduation rates. Looking at all the data is not looking for someone to blame, but seeking to understand what is happens to our students.
    Great points here Robert. A great deal of education reform these days is more about fuzzy math rather than trying to understand what is happening to our children in our public schools.