Measuring Graduation Rates

High school graduation rates always come up in the news. People are worried about them. People decry how low they are. People celebrate when they improve. Short of test scores, graduation rates are probably the most used indicator of educational success/failure in this country.

It makes sense to worry about graduation rates, but they're far from perfect. Two of the biggest problems are that it's almost impossible to calculate an exact rate for a school and that everybody calculates graduation rates differently.

I first learned about this a couple years ago when some folks from RAND did a study for Pittsburgh on their graduation rates. All calculations of graduation rates start with the same statistic: the number of new freshman who enroll in a given year. But how do we determine how many of these people have graduated and what the graduation rate should be? Do we check on them after four years? 5? 6? What about students who move out of the district? What about students who move into the district or otherwise enroll after the start of 9th grade? What about people who just stop showing up -- how do we know if they dropped out, moved away, or something else? Can we tell which students are in the 9th grade for the first time and which ones are repeating?

A large part of the difficulty is that there is no national database of students. If a student leaves and enrolls in a different city or state, there's no firm way of knowing that they're still enrolled. Similarly, if a student simply stops showing up at school rather than declaring themselves a dropout it's hard to tell if they've dropped out or enrolled in another district.

Anyway, the point is that it's not possible to compute an exact graduation rate. And the fact that every state and district seem to use different formulas means that it's hard to compare these inexact numbers. But, apparently, the Dept. of Ed. is going to step in and standardize the way that graduation rates are calculated. An article in the NY Times today says that the details of the new formula have not been released, but that all states will be required to use the same one.

The technicalities of which formula a state uses seems so trivial and boring that most people probably don't really give it much thought. But the article details people reacting quite strongly to the announcement. This makes sense because although the change is small, the effects of the change might be huge. The federal govt. is ostensibly holding states accountable for graduation rates, but they all calculate them differently (a recent change in North Carolina's formula led the official rate to drop from 95% to 68%). Sometimes the smallest changes are the ones that matter most.

The Federal Government and Education Research

Harvard is hosting a brief conference for grad students tomorrow (Friday) but they started off with a panel discussion on the state of education research tonight. The panel was highlighted by Grover "Russ" Whitehurst, the head of the Institute for Education Sciences (IES), which basically distributes all of the money the federal government devotes to educational research. He has been hailed by some for reforming education research and making it more rigorous and respectables, and reviled by others for limiting the scope of educational research to certain topics and methodologies.

I won't say that all 90 minutes of the panel was enthralling, but a number of interesting points were made. These include:

-Whitehurst noted that the education research community is not a powerful interest group in political discussions of policies surrounding education research. I've definitely noticed that. When you factor in that teachers and principals don't have much say in this either, it really makes you wonder who is controlling education research.

-There was much discussion of the fact that too much education research is not useful for people in schools, but few solutions offered.

-Whitehurst said that he made the decision to focus research funding on studies of academic achievement not because other things aren't important but because they have a very limited amount of money and want to do one thing well before moving on to others. While I'm not convinced they couldn't throw a few million dollars toward other desirable outcomes of schooling, I'm heartened to hear that it has been considered and would say the decision seems rational.

-In the face of severe criticism from Mica Pollock, a professor at Harvard, Whitehurst emphasized that neither he nor the department believe that only quantitative methods or randomized field trials are worthwhile (although he did seem to imply that RTFs answer more interesting questions). This sentence means nothing to you if you're not in education research but, basically, he refutes claims that many have made that he prioritizes certain types of research at the cost of other types that are more appropriate for certain questions.

Who's Asking What Teachers Think?

This piece in EdWeek caught my eye yesterday: Margaret Spellings is recruiting a team of teachers (5 full-time and 20 part-time) to help refine proposals for changes in education policy at the Dept. of Education starting in the fall. Two thoughts:

1. Wow. Seems like a great idea. Though, to be cynical, anybody could find 25 teachers that agree with their favorite policy proposal if they wanted to (I'll give them the benefit of the doubt and assume that's not what they're doing until I hear otherwise).

2. What momentous changes, exactly, is the Bush administration planning on making in his last few months in office? It's great that they have current teachers to comment on the feasibility of their ideas, but it's hard to believe that they'll have time to implement many of their ideas.