USC Education Professor Calls it ‘Crazy’ to Hide Student Growth Data from the Public

USC Education Professor Morgan Polikoff calls it “crazy” to keep student growth data from the public.

Morgan Polikoff is an associate professor of education policy at the USC Rossier School of Education and the author of a new report that argues strongly for California to adopt a student growth model to evaluate schools, something he says the state could do easily. This would give parents a better idea of how well each school is helping students progress academically. Polikoff also believes that LAUSD should share this information with the public. Growth data was a crucial part of the School Performance Framework that LAUSD had originally promised parents it would roll out this month but that Board Member Jackie Goldberg (BD5) is attempting to stop. The board will vote on Goldberg’s resolution to dismantle the SPF on Nov. 5 and will also decide whether to greenlight the release of student growth data. We spoke to Polikoff about his report and what might be keeping the powers that be from sharing this critical information.

Speak UP: You write that 48 states have some kind of model for measuring student growth, meaning how a group of students is doing this year compared to how that same group of students did last year. California is supposed to be really forward thinking. But we’re one of two states—the other is Kansas—that does not have this. Why?

Morgan Polikoff: I don’t think there is an accepted answer for why we don’t have one. The two likeliest reasons are one, that the previous governor was really not a believer in educational data, in general, and test scores, in particular, and so he didn’t want to invest in this. And two, there has been significant opposition to a growth model from the CTA [California Teachers Association] because of perceptions that it is a slippery slope towards some kind of teacher accountability.  

Speak UP: Your report makes a strong argument for using a student level growth model to evaluate and support California’s schools. Why isn’t a fixed snapshot good enough?

Polikoff: A fixed snapshot tells what the average performance is in a school. But it doesn’t tell you how good a job a school is doing at raising performance. What is a school doing during the year in principle? Well, it’s taking kids who come in at a certain point. It’s educating them. And then at the end of the year, it’s seeing how well it did. And when you think about it that way, it’s intuitively clear that what is the measure of school performance is not the average level at the end of the year, rather, how much students learn during the year, which is measured by growth. We also know that just looking at the average performance levels doesn’t work because schools differ so dramatically in the kinds of kids they serve and where those kids come in at. It doesn’t make any sense to compare the average performance levels of a kid from the most disadvantaged neighborhood in South Central with a kid from the most affluent neighborhood in the Valley. Those are different kids and different schools in different contexts. But comparing growth accounts for those differences because you’re comparing each kid to themself, and you’re really asking, how much are kids at this school learning on average?

Eleven education and community organizations are calling on LAUSD to share student growth data with the public.

Speak UP: The National PTA along with the Data Quality Campaign came out with a brief in September titled, Parents Deserve Clear Information About Student Growth in Schools.  So I assume growth data could be a valuable tool for parents to evaluate school options. How might parents use this information?

Polikoff: They might use it in the same way they used to use API or they currently use Great Schools. These are different ways of rating school performance. But as I just said, API or the way they rate schools on Great Schools, those are based typically on just average performance levels. So what you are telling parents is, the best schools are the ones with the highest average performance levels. Well, shock of all shocks, those are the schools that are in the most affluent areas. If you instead present growth data and you explain to people what it represents—it represents how much kids are actually learning in these schools on average—well, first of all, you’re going to see that certainly some of those affluent schools that are highly rated on status are also going to have high growth. But actually a lot of them won’t, and a lot of schools are going to have high growth that you might not think of. It really changes the way people think about school performance away from a system that just says the most affluent schools are the best schools, to a system that actually says the schools that are doing the best job of raising kids’ achievement, those are the best schools.

Speak UP: So growth and performance, they don’t always parallel each other?

Polikoff: No, definitely not. In fact, the correlation is pretty low. Schools that are high on average performance level, on average they might be slightly better on growth, but the relationship is pretty weak.

Speak UP: You write that California could adopt a growth model “very easily, and at close to zero cost.” Really? How?

Polikoff: It’s just a different way of analyzing the test data that the state already calculates. So the CDE [California Department of Education] would be the ones, or some contractor of the CDE, would be the ones who run the growth model. And this is not rocket science. My third year PhD student could set up this model in 5 to 10 minutes.

Speak UP: So they have the numbers to feed in?

Polikoff: Absolutely. It’s the same numbers that already exist. There’s no additional testing required or anything like that.

Speak UP: We hear the state may be working on a growth model, but it won't go into effect until late 2020 at the earliest. Given what you said about the ease and speed with which California could do this, why would it take so long?

Polikoff: The generous answer is, because the state really wants to do its due diligence. And there are a bunch of different kinds of growth models, and the state could be making decisions about which kind of growth model they want to use. In fact, I was on a task force [for California] that just wrapped up where we talked for months about which growth model we should choose and what were the strengths and limitations. The cynical view is that they are kicking the can as long as possible and like Lucy, may pull the football away at any moment because the politics are still…If CTA were to be really opposed and were to push back, it’s not super clear to me that there’s a strong constituency to get this thing through. Until it actually happens, I will not believe it’s going to happen.

Speak UP: The fact that there was a task force, we shouldn’t get too hopeful about that? We shouldn’t take it as a good sign this is coming?

Polikoff: I just don’t take anything for granted when it comes to data and this state. I mean, I think there is clearly some openness to it. But the task force itself was super frustrating for me and for other people in the room who believe, as I do, that this shouldn’t be an especially difficult decision because 48 other states have already made it. And as I say in the brief, there’s really only a couple approaches that are used in the vast majority of states.

Speak UP: What made it frustrating?

Polikoff: The fact that it felt like it was taking way longer than it should, and it wasn’t clear, even at the end of this fourth meeting…You would think if it was a task force, that what we would do is meet a few times, talk through some things, then write up a recommendation. That’s not what we did. We just talked for four meetings, and then it ended. And it wasn’t clear how they were going to take the work we did and feed it to the state. There were people from the CDE in the room for all our meetings, and they were listening. It just wasn’t really clear what we were there to accomplish. I was the main research person in the room. There were school district people. There were advocates. And a lot of the four meetings we spent on talking through the different kinds of growth models and what are the strengths and weaknesses of each. That’s not a lesson I particularly needed.

Speak UP parents called on the LAUSD board to give parents student growth data.

Speak UP: Let’s be really generous and say maybe the hang up is that they can’t decide on the model. Would any of the generally accepted growth models be something you support and think would be of value to parents and others?

Polikoff: I think either of the two that I call out in the report would be dramatically better than what we have. I think there are some choices that they could make that would be not better. So I don’t want them to make a bad choice. But if they just went with the student growth percentile, which is again, one I mention by name, if they just went with that tomorrow, that would be great. That would be a dramatic improvement over the change measure, even though I think that has some flaws and isn’t the best choice. But still, what we have now is really not good.

Speak UP: And when you talk about what we have now, you’re talking about where we see, for example, what third graders at a particular school did this year as compared to what a different group of third graders did at that same school last year?

Polikoff: Yes.

Speak UP: Let’s talk about LAUSD. The district was planning to launch its own School Performance Framework weighted heavily toward growth scores, but the teachers union has been pushing back on that. Could you talk a little more about why teachers unions are or might be opposed to this?

Polikoff: The most obvious answer is that there is fear that growth models, of which value-added models are one type, could be used for teacher accountability, although as far as I can tell, there is actually a zero percent chance that would happen in California because there’s just no interest in it. More generally, that growth models could be used for accountability not at the individual teacher but at the school level. And in general, teachers are opposed to external accountability.   

Speak UP: LAUSD was supposed to release the growth data right about now. The work has already been done. But then Jackie Goldberg put forth a resolution to halt that. Do you think LAUSD should let parents and the public see this information?

Polikoff: Absolutely. There’s no question. It’s crazy. The data exist. There’s no reason not to make them public. They could be useful to lots of different people, including parents, but also educators. As far as I can tell there’s no real downside. The resolution was awful and really disappointing. It’s hard to come up with a rationale that I think holds any water for the decision that was made.

— Interview conducted by Leslee Komaiko