By Doug Ward

Gauging the effectiveness of teaching solely on student evaluations has always been a one-dimensional “solution” to a complex issue. It is an approach built on convenience and routine rather than on a true evaluation of an instructor’s effectiveness.

And yet many universities routinely base promotion and tenure decisions on those evaluations, or, rather, a component of those evaluations in the form of a single number on a five-point scale. Those who rank above the mean for a department get a thumbs-up; those below the mean get a thumbs-down. It’s a system that bestows teaching with all the gravitas of a rounding error.

A new meta-analysis of research into student course evaluations confirms this weakness, underscoring the urgency for change. The authors of that study argue that student evaluations of teaching are not only a questionable tool but that there is no correlation between evaluations and student learning.

That’s right. None.

“Despite more than 75 years of sustained effort, there is presently no evidence supporting the widespread belief that students learn more from professors who receive higher SET ratings,” the authors of the study write, using SET for student evaluations of teaching.Macro shot of pencil writing/sketching on the checkered, blank page.

The study, titled “Meta-analysis of faculty’s teaching effectiveness: Student evaluation of teaching ratings and student learning are not related,” has been accepted for publication in Studies in Educational Evaluation. It was written by Bob Uttl, Carmela A. White, and Daniela Wong Gonzalez of Mount Royal University in Calgary, Alberta.

As part of their analysis, they challenge the validity of a seminal 1981 study that is often held up as evidence of the importance of teaching evaluations. That study and subsequent studies, they say, suffered from small sample sizes and “multiple methodological flaws that render their conclusions unwarranted.”

Course evaluations, they say, provide little more than a score for student perceptions, arguing that if student learning is important, we need other methods for evaluating teaching.

Their findings fall in line with a 2014 study by the statisticians Philip B. Stark and Richard Freishtat of the University of California, Berkeley. That study argues that course evaluations are fraught with statistical problems and “pernicious distortions that result from using SET scores as a proxy for teaching quality and effectiveness.” Among those distortions: low response rates, and failure to account for factors such as size and format of class, and academic discipline.

This is all damning evidence, especially because universities rely heavy on student evaluations in making decisions about instruction, and about instructors’ careers. It is especially problematic for the growing number of adjunct instructors, who are often rehired – or not – based solely on student evaluations; and for graduate teaching assistants, who are often shoved into classes with little pedagogical instruction and forced to make decisions about their teaching solely through the lens of end-of-semester evaluations.

All this points to the need for swift and substantial change in the way we evaluate teaching and learning. That does not mean we should abandon student evaluations of courses, though. Students deserve to be heard, and their observations can help instructors and administrators spot problem areas in courses.

The non-profit organization IDEA makes a strong case for using student evaluations of teaching, and has been one of its staunchest proponents. IDEA has created a proprietary system for course evaluations, one that it says accounts for the many biases that creep into most surveys, so its defense of course evaluations must be viewed with that in mind.

Nonetheless, it makes a strong case. In a paper for IDEA earlier this year, Stephen L. Benton and Kenneth R. Ryalls make a point-by-point rebuttal to criticisms of student evaluations of teaching, saying that “students are qualified to provide useful, reliable feedback on teacher effectiveness.” They acknowledge faculty frustration with the current system, saying that course evaluations are often poorly constructed, created in ways that ask students to make judgments they are not qualified to make, and “overemphasized in summative decisions about teaching effectiveness.”

“Those institutions who employ an instrument designed by a committee decades ago, or worse yet allow each department to develop its own tool, are at risk of making decisions based on questionable data,” they write.

So what can we do? I suggest two immediate steps:

Expand the evaluation system. This means de-emphasizing student evaluations in making decisions about teaching effectiveness. No department should rely solely on these evaluations for making decisions. Rather, all departments should rely on range of factors that provide a more nuanced measurement of faculty teaching. I’ve written previously about CTE’s development of a rubric for evaluating teaching, and that rubric can be a good first step in making the evaluation system fairer and more substantial. The goal with that rubric is to help departments identify a variety of means for judging teachers – including student evaluations – and to give them flexibility in the types of discipline-specific evidence they use. It is a framework for thinking about teaching, not a rigid measurement tool.

Revisit student evaluations of teaching. As I said, students’ opinions about courses and instructors deserve to be heard. If we are going to poll students about their courses, though, we should use a system that helps filter out biases and that provides valid, meaningful data. The IDEA model is just one way of doing that. Changing the current system will require an investment of time and money. It will also require the will to overcome years of entrenched thinking.

The problems in student evaluations of teaching are simply a visible component of a much larger problem. At the root of all this is a university system that fails to value effective and innovative teaching, and that rewards departments for increasing the number students rather than improving student learning. If the university system hopes to survive, it simply must give teaching the credit it deserves in the promotion and tenure process. Moving beyond reliance on course evaluations would be a solid first step.


Doug Ward is the associate director of the Center for Teaching Excellence and an associate professor of journalism. You can follow him on Twitter @kuediting.

By Doug Ward

The spread of evidence-based teaching practices highlights a growing paradox: Even as instructors work to evaluate student learning in creative, multidimensional ways, they themselves are generally judged only through student evaluations.

Students should have a voice. As Stephen Benton and William Cashin write in a broad review of research, student evaluations can help faculty members improve their courses and help administrators spot potential problems in the classroom.optical illusion box

The drawback is that too many departments use only student evaluations to judge the effectiveness of instructors, even as they submit faculty research through a multilayered evaluation process internally and externally. Student evaluations are the only university-mandated form of gauging instructors’ teaching, and many departments measure faculty members against a department mean. Those above the mean are generally viewed favorably and those below the mean are seen as a problem. That approach fails to account for the weaknesses in evaluations. For instance, Benton and Cashin and others have found:

  • Students tend to give higher scores to instructors in classes they are motivated to take, and in which they do well.
  • Instructors who teach large courses and entry-level courses tend to receive lower evaluations than those who teach smaller numbers of students and upper-level courses.
  • Evaluation scores tend to be higher in some disciplines (especially humanities) than in others (like STEM).
  • Evaluation scores sometimes drop in the first few semesters of a course redesigned for active learning.
  • Students have little experience in judging their own learning. As the Stanford professor Carl Wieman writes: “It is impossible for a student (or anyone else) to judge the effectiveness of an instructional practice except by comparing it with others that they have already experienced.”
  • Overemphasis on student evaluations often generates cynicism among faculty members about administrators’ belief in the importance of high-quality teaching.

Looked at through that lens, we have not only a need but an obligation to move beyond student evaluations in gauging the effectiveness of teaching. We simply must add dimension and nuance to the process, much as we already do with evaluation of research.

So how do we do that?

At CTE, we have developed a rubric to help departments integrate information from faculty members, peers, and students. Student evaluations are a part of the mix, but only a part. Rather, we have tried to help departments draw on the many facets of teaching into a format that provides a richer, fairer evaluation of instructor effectiveness without adding onerous time burdens to evaluators.

For the most part, this approach uses the types of materials that faculty members already submit and that departments gather independently: syllabi and course schedules; teaching statements; readings, worksheets and other course materials; assignments, projects, test results and other evidence of student learning; faculty reflections on student learning; peer evaluations from team teaching and class visits; and formal discussions about the faculty member’s approach to teaching.

Departments then use the rubric to evaluate that body of work, rewarding faculty members who engage in such approaches as:

  • experimenting with innovative teaching techniques
  • aligning course content with learning goals
  • making effective use of class time
  • using research-based teaching practices
  • engaging students in hands-on learning rather than simply delivering information to them
  • revising course content and design based on evidence and reflection
  • mentoring students, and providing evidence of student learning
  • sharing their work through presentations, scholarship, committee work and other venues

Departments can easily adapt the rubric to fit particular disciplinary expectations and to weight areas most meaningful to their discipline. We have already received feedback from many faculty members around the university. We’ve also asked a few departments to test the rubric as they evaluate faculty members for promotion and tenure, third-year review, and post-tenure review, and we plan to test it more broadly in the fall.

We will continue to refine the rubric based on the feedback we receive. Like teaching itself, it will be a constant work in progress. We see it as an important step toward making innovative teaching more visible, though, and toward making teaching a more credible and meaningful part of the promotion and tenure process. If you’d like to be part of that, let us know.

****

This article also appears in Teaching Matters, a publication of the Center for Teaching Excellence.


Doug Ward is the associate director of the Center for Teaching Excellence and an associate professor of journalism. You can follow him on Twitter @kuediting.

By Doug Ward

Women teach a sizable majority of online courses at KU, even though men make up a sizable majority of the university’s faculty.

Data provided by Laura Diede, the associate director at the Center for Online and Distance Learning, shows that of 171 online courses that CODL worked with in the 2014-15 school year, 60 percent were taught by women.

That’s especially interesting when you consider that of 1,649 faculty members on the Lawrence campus that fiscal year, only 42 percent were women.

I’ve not been able to find comparable data for online courses nationally, so I have no way to know whether the dominance of women in online teaching is unusual or not. In general, faculty members have been highly skeptical of online courses. In a recent survey by Inside HigherEd, more than half of faculty members said they didn’t think online courses could achieve the same level of learning as in-person classes. That attitude makes many long-time faculty members resistant to, if not hostile toward, online teaching.

Still, Diede and I still puzzled over the high percentage of women teaching online courses at KU. We came up with three possibilities:

  • Women are more willing than men to try new approaches to teaching.
  • Women prefer the flexibility that online teaching provides.
  • Men, who are more likely to have tenure, are more likely to refuse to teach online courses.

That last possibility seems the most likely, although there may be factors we hadn’t thought about. Whatever the case, it will be interesting to see whether this trend grows as the number of online courses grows.

Female instructors score slightly higher in online course evaluationsbar chart comparing course evaluations for men and women in online courses

The data that Diede provided about online courses also showed another interesting facet of online teaching at KU: Female instructors score slightly higher than their male counterparts on student evaluations.

This runs counter to a widely publicized recent study (and widespread perceptions) that argued that student evaluations are inherently biased against female instructors. I’m not going to wade into that debate here other than to say that student evaluations of teaching are problematic on many levels. The reliance on them as the sole measure of teaching quality benefits no one.

Fear and loathing about good teachers

The observation below is from Richard M. Felder, a professor emeritus at North Carolina State University. It was reprinted this week on Tomorrow’s Professor. It’s something I wonder about frequently and have talked about repeatedly with colleagues who value high-quality teaching:

“Some departments I know, including mine, have in the past hired faculty members who were exciting and innovative teachers and who didn’t do research. Some departments I know, again including mine, have hired former professionals with decades of practical experience who also didn’t do research. Both groups of faculty members did beautifully, teaching core courses brilliantly and serving as supportive advisors, mentors, and role models to the undergraduates who planned to go into business or industry after graduation. Professors like that are the ones students remember fondly years later, and endow scholarships and student lounges and sometimes buildings in honor of. And yet the thought of bringing one or two of them into a 20-person department faculty instead of hiring yet another research scholar who looks pretty much like the other 18 or 19 already there is unthinkable to many administrators and professors. Why is that?”


Doug Ward is the associate director of the Center for Teaching Excellence and an associate professor of journalism. You can follow him on Twitter @kuediting.