For faculty advocates, the Bakersfield story is both emotionally disheartening and politically valuable.
It goes like this: In 2009, facing steep budget cuts, California State University at Bakersfield laid off four math instructors and moved its two developmental math courses — courses meant to get weaker students ready for college-level math — to a fully online format. While tutoring services and supervised lab time were made available to students, it was up to them to schedule an appointment. Other than that, they were on their own. The math faculty was consulted, but their protestations “fell on deaf ears,” says Charles Lam, an associate professor of math.
The results were disastrous. In one course, the student pass rate plummeted from 74 percent to 45 percent. In the other, the rate fell from 61 percent to 37 percent.
The California Faculty Association seized on the figures. “CSU Bakersfield Remedial Math Needs Remediation,” the union declared on its website. California faculty reps in recent years have had their hands full fighting those whose plans for dealing with draconian budget cuts include cutting faculty jobs and replacing them, to various degrees, with online teaching technology. Even beyond California, faculty union reps still hold up Bakersfield as a cautionary tale highlighting the limits of technology as a “replacement” for real, live instructors.
What has happened at Bakersfield in the year since, however, is more open to interpretation. Prior to the 2010-11 academic year, the campus revised its developmental math curriculum to include mandatory lab hours under the supervision of teaching assistants, as well as a once-a-week lecture that would cover both math theory and tips on studying and time management. Bakersfield did not, however, re-hire any of the math instructors it had let go, and the amount of time students spent with live instructors at the ready was much smaller than it had been before the experiment.
Pass rates recovered dramatically in both classes. In fact, they exceeded the pre-experiment rates — rising to 85 percent in the first course and 68 percent in the second, according to a university spokesman. Officials say there were no changes to how students were assessed.
On the one hand, faculty advocates can point to the fact that taking mandatory face time with faculty and teaching assistants out of the curriculum caused pass rates to tank, and putting it back in caused pass rates to recover. Furthermore, the first curricular experiment was conceived with minimal faculty input, while the second was deliberated with “more active” participation by the chair of the math department and by Lam, who runs the math tutoring center and was initially angry that the original changes appeared to have been made by fiat. (Lam says it was he who suggested the partial restoration of mandatory, supervised lab hours; he credits the dean with the idea for the 50-minute weekly lectures.)
On the other hand, deans at Bakersfield and advocates of streamlined online courses can point to the fact that under the second iteration of the new model — absent four members of the math faculty and a good chunk of supervised lab time — students performed even better than they had when Bakersfield had been spending money on those resources. After a rough first test run, the university made adjustments and ended up with a model that, according to this limited data sample, serves students at least as well with an ostensibly lower overhead.(Bakersfield officials could not provide Inside Higher Ed with numbers for how much each version of the courses cost.)
“It is much more an accurate description to look at [Bakersfield] as a place where we actually have been able to effectively think outside the box and serve students within the budget constraints than the other way around,” says Kamel Haddad, an associate dean at the university. (Professors point out that the university arrived at a more harmonious balance between technology and human interaction only after heeding the counsel of faculty members such as Lam.) Haddad added that the decision to rework the developmental math curriculum was not a voluntary attempt to do online education on the cheap, which is rarely a successful strategy, but rather a desperation move undertaken to spare entire programs from the budget ax.
Then there is the point of view that says the Bakersfield example proves little for either side. This year’s pass rate was promising for the new model, but not conclusive, says Lam. “I know that there are still faculty who are not happy with the changes, and I think that we’ll have to observe the data for another year to see if the data works out,” he says.
As far as its implications for the prospects of automated teaching software developmental education, Lam says students taking catch-up courses at Bakersfield are unusual in that many of them have in fact taken — and in some cases, passed — the material in high school. That makes them different from developmental math students who never learned it at all, such as those at many community colleges, Lam says. Students who have never previously learned certain theoretical concepts probably need more hand-holding from instructors, he says, whereas a computer program might do for those who simply need a memory jog.
But this is still more speculation than anything, says Lam. Politics is one thing; science, another. And pass rates from single trials of three different models of developmental learning hardly form a basis for scientific certitude. “We are actually perplexed,” he says. “We have done a number of things, but we’re not sure what works.”
For the latest technology news and opinion from Inside Higher Ed, follow @IHEtech on Twitter.