Teaching, Research and Scholarship Part 2

Today’s post will focus on the oft-cited Hattie & Marsh (1996) article: “The Relationship Between Research and Teaching: A Meta-Analysis” in Review of Educational Research available here in JSTOR. I briefly mentioned some of the results in Part 1 of this series, where I also set the stage on the broader issues related to today’s post.

The article begins with a literature review citing the three main relationships between teaching and research: Positive, Negative and Zero. The dominant models proposed in each of these areas are briefly summarized. I will highlight one model from each group. If you’re interested in the other models, please read the paper!

Positive Relationship: The “Conventional Wisdom” model stems from many survey results whereby a large number of academics claim that teaching and research enrich each other. A teacher who stops doing research will eventually repeat old-hat and lose touch with the students and with the reality we live in. Being an active researcher is key for teaching advanced classes (requiring advanced knowledge) and in stimulating student attitudes towards the discovery of new knowledge. Interestingly, while active research is claimed to benefit teaching, the reverse claim – that someone actively interested in teaching will be a better researcher – is not widely held. Could being aware and abreast of advances in one’s field without directly partaking in the “discovery” process be sufficient to teach well at the university level? (More on this in a future post!) Regardless of the answer, it is clear that in recent years, a sizable portion of undergraduate instructors in U.S. baccalaureate degree granting institutions do not engage in “discovery” scholarship. Certainly in the sciences, they have no lab space or resources to do so.

Negative Relationship: The “Scarcity” model claims that time and commitment pull in different directions for teaching and research activities. To excel in one requires taking away from the other given that both activities are labor-intensive. Dimensions of conflict include curricular versus scholarly interests (particular in institutions with graduate programs), and disciplinary versus institutional identification of the faculty. Studies show that time on research is negatively correlated with time on teaching. The relationship however is more complicated. While time and commitment spent on research correlates with research “productivity” (usually measured in terms of publications), there is no clear correlation between time spent on teaching on either research productivity or teaching “effectiveness” (often measured by teaching evaluations). On the other hand, those with a higher commitment to teaching do tend to receive higher teaching evaluations. A quote sums this up: “Teaching and research, far from being complementary activities, appear to be either completely unrelated or to be in conflict with one another. The most productive researchers have the least favorable attitudes towards teaching, while the least productive are the most committed to teaching.”

Zero Relationship: The “Different Enterprises” model argues that research (i.e. the discovery of knowledge) is not driven by curricular considerations but rather “by an interest structure based on academic careers.” Furthermore research activity “relates more to the discovery of knowledge usually by normative means within various disciplines”, whereas teaching relates to imparting this knowledge in a way that leads to student learning. “Researchers are valued for what they discover, teachers for what they enable students to discover.” While the qualities required in teaching and research may be different according to this model, a single individual may exhibit the qualities required to excel in both, but it is not a given. Thus, if we unrealistically divided people into four stark categories and ignored the grey areas for the moment, you would find good teachers who are good researchers, poor teachers who are poor researchers, good teachers who are poor researchers, and good researchers who are poor teachers – and you might find the categories equally represented in academia (well, hopefully not too many in the last group).

The actual study is a comprehensive meta-analysis that aims to systematically review the different models with a common statistical framework. Some context to the data: (1) 34% from liberal arts colleges and 62% from research universities, (2) about half the of studies were across all departments, and a quarter from social sciences, with the natural sciences and humanities at 17% and 10%, (3) 80% used teaching evaluations as the main indicator for teaching effectiveness, with 19% peer evaluations, (4) 63% used number of publications as the main indicator for research productivity, although citations, grants, and weighted sums of book chapters were also used.

The reader should consult the article for the actual statistical methods. I’ll jump directly to highlighting some of the results.

·      The overall relationship was slightly positive but very, very close to zero (weighted average of 0.06).

·      There was a negative relationship in more recent studies.

·      More extreme positive or negative relationships in individual studies all came from much smaller sample sizes, and three most significant outliers used self-ratings of teaching and research.

·      Variation was larger in liberal arts colleges compared to research universities.

·      Relationship is very close to zero in the natural sciences and the humanities and consistent across sub-disciplines. There is more variation in the social sciences (negative in commerce and law, zero in psychology and anthropology, and positive in education).

·      Time on teaching was negatively correlated with time on research, but this did not factor in teaching load (number of courses taught during the term). Therefore one cannot conclude that less total time spent on teaching leads to poorer teaching outcomes.

·      There is small but positive correlation that more active researchers are better at presentation aspects of teaching, although there is little correlation with facilitation aspects.

Some have interpreted the Hattie and Marsh study as a reason to provide differential funding streams and have instructors or even institutions specialize more in one and less in the other. The authors however clearly state in their conclusion the opposite: “The goal should not be publish or perish, or teach or impeach, but we beseech you to both publish and teach effectively.” They, and many others, think that the main problem is how to bring teaching and research together effectively such that the “relationship between the two attributes can increase, [and] it is hoped, positively.”

To conclude this post, here’s Figure 1 from the article – one of several “mixed” models that try to go beyond the simpler models mentioned above. This is Marsh’s Compensatory Model. Further commentary can be found in the paper.