At a recent conference
put on by CALDER — a research group affiliated with the American Institutes for Research — top scholars from the University of Virginia, Duke, Northwestern and many others presented new studies on a host of education topics. All of the studies were interesting and policy relevant; here are summaries of five particularly interesting ones.
When teachers cheat, students suffer in the long run.
Everyone who follows education remembers the Atlanta cheating scandal, in which several teachers went to jail
. But what happened to the students whose test scores were manipulated? Research
shows that those students’ scores on later tests were lower, with enduring effects on English exams. There’s also evidence that cheating led to an uptick in high school dropout rates among students whose tests were fudged. The researchers suggest that the students were harmed because they didn’t get the remediation or extra support services they needed since their true academic standing was masked by the phony scores.
In other words, cheating on high-stakes exams doesn’t just corrupt accountability systems; it also inflicts lasting harm on students.
Should states have stopped test-based teacher evaluation during the transition to Common Core? Maybe not.
paused negative consequences for teachers based on test scores during the move from the old standards to the Common Core. The argument for doing so was intuitive: it wouldn’t be fair to hold teachers accountable for results on tests and standards that were new and unfamiliar. But a new study
of five states suggests that even during transition years, the reliability1
of value-added ratings
— an estimate of teacher impact on students’ test score growth — generally held constant. Although in a couple cases this wasn’t true, even then the researchers show (not too surprisingly) that teachers’ past value-added rating helped predict their future scores.
This suggests that states may have been hasty in hitting the breaks on test-based teacher evaluation. On the other hand, moving slowly may have been smart politics, given the backlash
set off in many states against testing and the Common Core.
Teachers are (literally) paying for states' pension debt.
Many states’ pension systems are badly underfunded and current teachers are carrying that burden, according to research
from late last year. The study finds that to meet pension obligations — underfunded, the authors say, largely because of unrealistic assumptions about the funds’ return on investment — states are using contributions from younger teachers. In Missouri, for example, 29 cents for every dollar of teacher salary goes to pensions, but working teachers only see 19 cents of that. The remaining 10 percent is for paying pension debt. Similar trends apply across the country.
The authors refer to this as a “pension tax” on current teachers because it’s money that could, for instance, go towards higher salaries to help recruit and retain teachers. The researchers suggest moving towards a system that ties the value of teachers’ retirement benefit to their own contributions and investment returns.
Still, there’s no easy answer to the current problem: Someone — current teachers, retired teachers, or taxpayers — has to pay the long-term, unfunded pension obligations.
More teaching assistants = higher student achievement.
Not all adults in schools
are teachers of course, but researchers know very little about how staff like teaching assistants and health care providers affect student outcomes. That’s why this new study
is so important: it shows strong evidence that teaching assistants have significant positive impacts on student achievement and attendance in North Carolina schools. The paper also finds that small class sizes have small positive effects on achievement, as well as suggestive evidence that health providers — largely speech pathologists, psychologists, and social workers — boost test scores. The paper contributes to the long-running debate
about whether more spending improves student outcomes by showing that the Great Recession budget cuts that reduced school staff harmed test scores. “Cutbacks in resources of this type have clear adverse consequences for students,” the authors write.
Some places are laying off teachers all wrong.
As the previous study showed, layoffs of school staff are painful for schools and harmful to students. But insofar as staff reductions are necessary, are states and districts at least doing all they can to minimize the damage? Not really, according to a pair
focusing on Los Angeles and Washington state. In both places, when layoffs are likely, teachers who might lose their job are informed in the middle of the school year. Once budgets are finalized, though, many of those teachers remain in their schools. But the damage, in some ways, is still done: Teachers who receive layoff notices become less effective perhaps because their morale suffers or they are distracted searching for another job. They are also more likely to leave their school.
This research — paired with other evidence
suggesting that basing layoffs largely on seniority harms student achievement — suggests significant scope for improving how job cuts are handled. The researchers recommend considering alternatives to last in, first out layoffs — which in some places are mandated by state law or the local teachers contract — and delaying layoff notice timelines until school districts have a more accurate picture of how many teachers will really be let go2
1. Reliability here refers to the correlation between a teacher’s value-added rating in one year and that same teacher’s rating the following year. (return to story)
2. Interestingly, these suggestions partially echo recommendations from the non-partisan California’s Legislative Analyst’s Office. (return to story)