A recent working paper by Jan Feld (Victoria University of Wellington), Nicolás Salamanca (University of Melbourne), and Ulf Zölitz (University of Zurich) provides an initial answer to this question (see also their article in The Conversation a couple of weeks ago, describing the research). They used data on over 12,000 students in a Dutch university where, importantly, students are randomly assigned to a tutorial teacher, who may be at any level between undergraduate student and professor. They then looked at whether it made a difference for student's grade in that course, their grade in subsequent courses, course evaluations, job satisfaction after graduation, and earnings after graduation. They found that:
Instructors’ academic rank is, overall, unrelated to students’ academic outcomes. The most effective instructors—postdocs—add less than 1 percent of a standard deviation more to students’ grades than student instructors. For all other instructor types, we can rule out differences between instructor types as small as 1 percent of a standard deviation in grades. Instructors’ academic rank is also unrelated to students’ grades in follow-on courses, where our results are also precisely estimated.
Looking at nonacademic outcomes, we find that instructors with higher academic rank add more value to students’ course evaluations. However, these differences are also small. Students taught by a full professor, for example, evaluate the course only 4 percent of a standard deviation more positively than students taught by a student instructor. Finally, using matched survey data on university graduates, we find no systematic relationship between instructor academic rank and students’ job satisfaction and earnings after graduation.Where the effects are statistically significant, the size of the effects are tiny. In other words, it really makes no difference whether students are taught in tutorials by undergraduate students, professors, or someone in-between. At least, it doesn't make a difference to the outcomes they measure, which are arguably most of the outcomes that we would care about, starting with whether the students get better grades in that class.
Feld et al. then go on to argue that universities would be better off if they moved to using undergraduate students as tutorial teachers, because they are paid less and therefore it would be cheaper. I don't find that argument necessarily persuasive, since it requires that the marginal cost of a professor is more than the marginal cost of an undergraduate. You could argue that the marginal cost of a salaried staff member is zero, since you don't have to pay them any extra to stand in front of a class. However, that would ignore the real opportunity cost of having a professor teaching a tutorial group. That opportunity cost isn't the cost of the professor's salary, but the quality research time (and research outputs) they would give up by teaching the tutorial. Chances are that the value of more high-quality research to the university exceeds the professor's hourly salary by some margin, so Feld et al.'s argument that tutorials are better taught by undergraduates (or other students) is probably correct, but not quite in the way they argue.
Either way, it provides support for the approach we use in economics at Waikato, where we use senior undergraduates as tutors (at least in our first and second year classes).
Read more:
No comments:
Post a Comment