Featured post

Textbook: Writing for Statistics and Data Science

If you are looking for my textbook Writing for Statistics and Data Science here it is for free in the Open Educational Resource Commons. Wri...

Tuesday 3 October 2017

Book review of Improving How Universities Teach Science Part 2: Criticisms and comparison to the ISTLD

As far as pedagogical literature goes, Carl Wieman’s Improving How Universities Teach Science - Lessons from the Science Education Initiative was among my favourites. It keeps both jargon and length to a minimum, as it is barely more than 200 pages without counting the hiring guide at the end. The work and its presentation are strongly evidence-based and informative, and the transformation guide in the appendices provides possible actions that the reader can do to improve their own classes. Most of it is applicable to most of science education.

 I have some criticisms, but please don’t take this as a condemnation of the book or of the SEI in general.

The term ‘transformation’ is vague, despite being used extensively throughout the first two thirds of the book. This is partially forgivable because it has to be vague in order to cover what could be considered a transformation across many different science fields. However there could have been more examples or elaboration or a better definition of the term early on. First concrete examples to show up and clarify what transformation meant are found in the appendix, 170 pages in.

Dismissal of the UBC mathematics Department, and of mathematics education in general.

The metric Wyman used primarily the proportion of faculty that bought in to the program. That is the proportion of Faculty that transformed their courses, because typically faculty transformed all of their courses or none of them. Many departments were considered a success in that 70% or more of the faculty transformed their classes. Those that were under 50% were mostly special cases where they had entered into the Science Education initiative later and hadn't had the opportunity to transform. Among the All-Stars was the UBC statistics Department 88% of their 17 faculty with teaching appointments transform their classes. Among The Faculty of the UBC mathematics Department however only 10% of their 150 + strong Department bought in and transformed their classes. To contrast 1.2 million dollars was spent on the mathematics Department while $300,000 was spent on the statistics Department, so the mathematics people got more in total but the statistics people got more per faculty. It's not the failure to transform the mathematics Department that bothers me but the explanation for it.
Wieman boils down the failure to transform the mathematics department into two factors. First was the culture within that particular department, which was one that did not emphasize undergraduate education and seemed to assume that mathematics was an innate ability that either students had or had not regardless of the amount of effort put in. Before Wieman started attempting to transform this department it had a policy of automatically failing the bottom few percentiles of every introductory calculus class regardless of performance. The second factor Wieman uses to explain the failure is that mathematics is inherently not empirical, which means that a lot of the active learning meant to make Concepts more concrete would not have applied.

Having taught and been taught in both mathematics and statistics departments at multiple institutions myself I don't buy these arguments. From personal experience the most engaging and active classrooms I experienced have spread equally across mathematics and statistics. With in mathematics the most memorable was abstract algebra which by definition is non-empirical. Furthermore, at Simon Fraser University it's the mathematics department that has been leading the way on course transformation.
As for the argument about innate ability, this is an idea that spreads far beyond just university departments. I have no qualification to claim how true or false it is. However it's not a useful assumption, because it makes many things related to teaching quality in mathematics automatically non-actionable.
Finally it seems like a strange argument for a professor of physics to make about mathematics. I would have like to see more investigation and perhaps it's covered in some of his other literature, but then I would have like to see more reference towards that literature if it exists.

Compared to the Institute for the Study of Teaching and Learning in the Disciplines (ISTLD) at SFU, Wieman’s SEI is several times larger in scale and tackles the problem of university teaching entire departments at a time. The SEI works with department chairs directly and with faculty actively through their science education specialists. The ISTLD’s projects are self-contained course improvements, where staff and graduate student research assistants provided literature searches, initial guidance, and loose oversight over the course improvement projects. Both initiates fostered large volumes of published and publicly presented research.
The funding for course improvement projects through the ISTLD was non-competitive; the only requirements to receive a grant were to submit a draft proposal, to attend some workshops on pedagogy and to submit a new proposal guided by these workshops. Grants from the SEI at both UBC and CU was a competitive process, which Wieman used because, in his words, it was the only system familiar to science faculty.

In case you missed it, here is the first part of this book review, which discusses the content more directly.

Book review of Improving How Universities Teach Science, Part 1: Content.

Unlike many other books and literature on the subject, Carl Wieman’s Improving How Universities Teach Science - Lessons from the Science Education Initiative spent most of its pages talking about the administrative issues involving the improvement of university teaching. If you're familiar with recent pedagogical literature this book doesn't come with many surprises. What set it apart to me is the scale of the work that Wieman undertook, and his emphasis on educational improvement being an integrative process across an entire department rather than a set of independent advances.

The Science Education Initiative, or the SEI, model is about changing entire departments in large multi-year, multi-million dollar projects. The initiative focuses on transforming classes by getting faculty to buy into the idea of transforming them, rather than transforming the classes themselves directly.

The content is based on Wieman’s experience developing a science education initiative at both University of British Columbia (UBC) and at Colorado University (CU). It starts with a vision of what an ideal education system would look like any university mostly as an inspiring goal rather than any practical milestone. It continues with the description of how change was enacted in both of these universities. The primary workforce behind these changes was a new staff position called the science education specialist or SES. SES positions typically went to recent science PhD graduates of that had a particular interest in education. These specialists were hired and then trained in modern pedagogy and techniques to foster active learning. These specialists were assigned as consultants or partners to faculty that had requested help in course transformation.

The faculty themselves were induced to help through formal incentives like money for research, or through teaching buy-outs that allowed them more time to work on research, and through informal incentives like considering in teaching assignments and opportunities for co-authorship on scholarly research. Overcoming the already established incentive systems (e.g. publish or perish) that prioritized research over teaching was a common motif throughout this book.

The middle third of the book is reflective, and it’s also the meatiest part; if you’re short on time, read only Chapters 5, 6, and the coda.  Here, Wieman talks about which parts of the initiative worked immediately, which worked after changes, and which never worked and why. He talks about his change from a focus on changing courses to a focus on changing the attitudes of faculty. He talks about the differences in support he had at the different universities and how that affected the success of his program. For example, UBC got twice the financial support and direct leadership support from the dean. He also compares the success rate of different departments within the science faculty. Of particular interest to me are the UBC statistics and the UBC mathematics departments, which obtained radically different results. The statistics department almost unanimously transformed their courses, while the mathematics department almost unanimously didn’t.

Wieman also talks at length about ‘ownership’ of courses, and how faculty feeling that they own certain courses is a roadblock. Calling it a roadblock is partly because of the habit of faculty to keep their lecture notes to themselves on the assumption that they are the only one teaching a particular course. Furthermore, the culture of ownership was perceived to contribute to resistance from faculty to changes to their courses.

Under Wieman's model, course material is to be shared with the whole department so that anyone teaching a particular course has access to all the relevant material that has been made for it by department. Although UBC managed to create a repository for course material, the onus on populating that repository the faculty and there were few people that actually contributed. However where this matters most in the introductory courses even partial sharing was enough because many people tend to teach those courses.

The final third of the book is a set of appendices which include examples of learning activities and strategies in transformed courses, guiding principles for instruction, and several short essays on educational habits and references to much of the other work that Wieman has done. It also includes a hiring guide with sample interview questions for possible Science Education specialists.

The book also includes coda, which is an 8 page executive summary of the first two parts of the book. The coda served as a good review also a nicely packaged chapter that could be shared with decision makers such as deans and faculty chairs. Decision makers are exactly who I would recommend this book to; it has an excellent amount of information for the time and effort it takes to digest.

I had a few other thoughts about this book that were set aside for the sake of flow. You can find them in the second part of this book review.