Teaching philosophy
I’m an assistant professor in the UTM Biology department, with a large focus on teaching numerical and coding skills to biology undergraduate students. I approach teaching from a highly conceptual standpoint, with exam questions that test understanding through word (or scenario) based problems.
Active learning
One of the most effective form of learning in my opinion is learning through discovery, which is a foundational aspect of the constructivist framework. This is because advancing beyond introductory statistics and coding is a bit like learning how to ride a bike: it requires an understanding of concepts rather than memorization. This is particularly true in my course that heavily focuses on conceptual understanding rather than applying recipes or reciting information. This approach of teaching, however, has drawbacks when teaching in large classrooms at the undergraduate level. The major one is that progression through the topics and exercises becomes more difficult if a student has not been able to follow (due to any set of reasons), and this sometimes becomes unsurmountable. Of course, the goal of the teacher is to find ways to identify shortcuts to understanding and apply them but when the course is highly streamlined, adequate learning is only possible with full student engagement and this is not always achievable. One of my most shocking discovery while teaching at UTM is that only few students are willing to reach out for help once they realize the gulf between their understanding and my expectations has started to grow (and those that do reach out all do much better at the end of the class).
I don’t have a great vision for how do we get better engagement from students but my wish is that students realize professors want to help them succeed. In some ideal world, students are in a class because they want to learn the material presented and not because they are forced to be there. What then can we do if students don’t think biology requires numerical and coding skills? One way is of course to show that all faculty have these skills and that many job opportunities from biology require them. However, I do acquiesce that it may be that we will never fully be able to get students in biology to be enthusiastic about math, stats, or programming. I suppose a better way would be to reinforce from the beginning that biology is becoming a quantitative science.
Student evaluation of teaching
I believe student evaluation of teaching should only be used to assess the resource needs for teaching the desired curriculum, and that specific scores are not useful metrics of effective teaching. Personally, I think teachers should be assessed by senior teachers (though this form of mentership is usually lacking at the University level) with the use of downstream courses to assess how prepared students actually are. The reasons are several-fold.
1) Research has shown that teaching evaluations are mostly uncorrelated with actual learning, and that major contributors of teaching evaluation scores are discriminatory in nature. The reasons for this are not always obvious, but what is clear is that actions following the scores will lead to fewer women teaching at the University level with an even stronger impact on diversity and inclusivity.
2) Undergraduate students are not qualified to assess teaching in large-class settings because they perceive their point of view as universally applicable to the whole student population, and do not fully understand the impact of some actions on learning progress. Furthermore, there is an obvious conflict of interest whereby student evaluation of teaching scores are highly correlated with their grade in the course. This leads to teachers not covering part of the syllabus or the curriculum to decrease the load on students, to instructors teaching to the test because it makes the exams easy, to teachers not teaching 'unpopular' classes despite them being the most suitable for the task, and to pervasive grade inflation.
The troubling use of student evaluation of teaching has massive repercussions for effective learning. First, it is well established that active learning is a superior form of teaching, but that it almost always leads to students believing the teacher is less effective at teaching. Second, it is not an effective metric to gauge whether the students are learning better from year to year (again, it is largely unrelated to actual learning, and in many cases contrary to actual learning). Finally, I teach a 4th year biostats course, which by that point all students have taken a 2nd year and a 3rd year stats course as prerequisite - both with great student evaluations - and very few students have retained the basics so evaluations are not related to future re-use of the material. This view has been shared by essentially all professors I've spoken to requiring these stats requirements for their course.
On giving out solutions to term tests
My policy here has usually been that students can receive their copy of the exam and unlimited subsequent feedback. That is, they can learn from their mistakes by retaking the exam and obtaining a 'putative grade' as a mark of their learning progress. I don't give out the solution unless the student comes to my office and works out with me their reasoning. While this approach is not uncommon at the University level (and maybe only occurs in half the courses due to the type of teaching the professor likes to do), I believe the reasons provided are not always pedagogical in nature (e.g. this prevents recycling of exams). I don't believe these reasons are compelling for this practice.My reasons are simply put, I don't believe knowing the solutions to a problem is helpful when the exams are conceptual in nature. I also believe that the retrospective aspect of looking back at answers and figuring out which questions were less certain is key to actual learning. Here's why: students attempt to learn the test rather than learning the concept. In fact, that desire is so strong that I have been able to find dozens of examples of this behavior from one semester of teaching. The reality is that only knowing the answer to a single question without knowing how to get there strongly sabotages further problem-based learning. That's because you can get the right answer through many incorrect ways and this way of learning reinforces incorrect post-hoc reasoning. For example, in one semester I ask a question on correlation: given a correlation of 1, an x/y coordinate for one point, which following points are possible for a second point? In the first exam, half students choose, within the multiple choice, only the point that falls on a 1:1 line while the answer was 'all of the above'. In the second exam, I ask the same question but with a correlation of -1, and in that case only one point is valid, yet 30% of the class picked 'all of the above'! This type of observation was seen over and over - students pick the familiar answer, not the one that makes sense.
Further, problem solving cannot be learned from solving a single problem: students only learn how to solve that one problem and cannot extend it to slight modifications of this problem! Learning stats and coding is not a recipe or bag of tricks, or formula to apply, yet students insist on trying to use this flawed methods for studying. It is therefore my policy to help students get to the answer, and to provide overall feedback until students can robustly identify what they know and what they don't know. Students complain that they cannot possibly know which questions they got wrong, but in my experience, when faced with this task, students in my office almost always identify the questions they got wrong within a few minutes: they are the questions they are less sure about.
References
Study from PNAS (Deslauriers et al., 2019)Meta-analysis of faculty's teaching effectiveness: Student evaluation of teaching ratings and student learning are not related. (Bob Uttl, Carmela White and Daniela Wong Gonzalez (Studies in Educational Evaluation, 2017))
Eric Mazur interview on teaching methodologies and student evaluations
The Art of Probability for Scientists and Engineers by Richard Hamming