The Retroactive Lowering of the GED Cut Score
On January 20, 2016, the Atlanta Journal-Constitution carried a story about the possibility that the GED Testing Service would retroactively lower the cut score for passage of its revised GED. That same day, EdWeek picked up the story, and on January 26, 2016, GED Testing Service confirmed that the GED cut score would indeed be retroactively lowered from 150 to 145 but that states would be left to decide to follow such advice:
GED Testing Service …recommends that states apply retroactively the 145 passing score to test-takers who have tested since January 1, 2014. When a state approves applying the passing score retroactively, students who earned scores between 145-149 on the new GED test launched in January of 2014 would be eligible for their state’s high school equivalency credential.
“The scoring enhancements are based on an extensive analysis of test-takers’ performance data from the past 18 months, conversations with state policymakers and elected officials, and external validation with experts,” said GED Testing Service President Randy Trask. “This is part of our ongoing commitment to make data-based decisions, and continually improve the efficacy of the GED program.”
Even as GED Testing Service promotes this change as positive because it is based on research showing that GED students were faring better than traditional high school grads on that all-too-oft-heard “college readiness,” what it also reveals is the power of cut scores to drive individuals’ lives.
The Atlanta Journal-Constitution notes that if Georgia follows the GED Testing Service recommendation for states to retroactively issue GEDs based on the lowered passing score of 145, then approximately 1,900 Georgians will be awarded GEDs after being told they failed– and likely adjusting their lives to that failure.
High-stakes cut scores profoundly affect peoples’ lives.
While my view of the value of standardized testing in general has waned sharply over the last few years, the statistician/researcher in me believes that the onus is on testing companies to get it right the first time. That means conducting the longitudinal research before instituting the change.
The GED Testing Service is proud of its retroactive adjustment. It apparently has not offered any public apology to those affected by the initial test score that it now maintains ended up overshooting on so-called “college readiness.”
And interestingly enough, many states penalize schools for students who do not graduate with their four-year cohort and instead earn a GED. But now, we are hearing that when it comes to “college readiness,” those who attend college after earning a GED are doing at least just as well. From the EdWeek article:
Tracking student performance into college, the company noticed that in several states, fewer students who passed the GED needed remedial coursework than those who earned high school diplomas.
So, counting those who take the GED as dropouts (“losers”?) because they don’t graduate with their high school cohort, or penalizing high schools by awarding less “value” to school performance score calculations for students earning a GED, is brought into question as GED “adjusts” its cut score.
Another issue to consider: The GED cut score adjustment is based on a relative outcome– it is contingent on how well traditional high school graduates do in college. Consider this statement in the Atlanta Journal-Constitution article:
“If high school performance starts to improve, we can adjust our cut scores as well, but we want to make sure we are holding adults to the same standards” as those required for traditional high school students, said Randy Trask, GED Testing Service’s president. [Emphasis added.]
Having a relative standard is a problem in modern test-score-centered ed reform. A major example of this is the 95 percent testing requirement retained in the Every Student Succeeds Act (ESSA), the latest revision of the Elementary and Secondary Education Act of 1965 (ESEA). The reason that ESSA must push for 95 percent of students of a given grade level to take standardized tests is so that it has enough test takers to compare subgroups of students to ascertain “achievement gaps.” Those gaps are relative, and as relative, they can only be “measured” if virtually all students of a given grade level take those Title I-mandated tests.
It would be better to set community-invested, fixed goals for all students and investigate what resources the varied communities need in order to best achieve the fixed outcome. Nevertheless, this is difficult to do, if for no other reason than the propensity for politics and profiteering to easily work their way into such noble endeavors.
For now, let’s just end with a simple request for the GED Testing Service:
Get your act together. No more retro GED cut score adjustments.