Faculty Experiences - Duncan Thomas



Interview Topics

What matters most to you in your teaching?

How are you using technology as a tool to achieve your teaching goals?

How have your students responded to your use of technology?

What new goals do you have for using technology in teaching?


Get student input
Relate theory to practice
Student motivation
Student participation
Test & critique theories
Thinking like a researcher
Work with data in real time


3D graphics
Class web site

Helping Students Understand Empirical Analysis

The goal of the class is to give students a sense of the excitement associated with learning about how the world works using data to test hypotheses. The idea is to provide the students with the tools to enable them to undertake empirical analysis and to provide them with the tools needed to interpret their own and others' empirical studies.

The regression model plays a central role in the class, and I try to give the students a very solid intuitive understanding of the assumptions underlying the model. I then discuss how violations of the assumptions might be detected, what those violations imply for interpretation of the evidence, and what tools we have that enable us to address the concerns raised by these violations. This intuitive understanding is backed up by providing real-world examples that demonstrate these points to students in practical and concrete terms. The examples draw from the world around us so the students can see how the tools apply to things they (might) care about. Moreover, the foundations are laid using sufficiently rigorous arguments so that the key ideas can be further developed by the students without too much extra work in other classes or their work after graduation.

The key to understanding empirical research in the social sciences is doing it and interpreting it. I use in-class demonstration of examples with real data on a laptop computer. The examples are developed with student input in real time. Of course, I have a class plan and have a set of material that I will work through. However, it has been my experience that encouraging student participation reaps tremendous benefits as they really feel they are doing the empirical analysis themselves and, therefore, that they are conducting some research. While this is very hard in a large class (>200 students), it is feasible and reaps tremendous rewards. The students get engaged and interested. To them, it feels like a journey of discovery. And, on occasion it is. This is because students often have suggestions for model specification or interpretation that implies some alternative specification. I am able to pursue that line of inquiry in class with them because everything is done in real time. On several occasions, a student has suggested an approach that introduces concepts that are part of the course syllabus but are not in the plan to teach there and then. By teaching that concept in response to the student question, it has been my experience that the students are much more receptive to the concept and that this interactive approach to teaching is far more effective than a more standard lecture format. I find the students are more engaged, feel they have more to contribute and are more enthusiastic when the class is interactive. And, they therefore learn more.

I do not use multimedia. What I do use is a computer, an overhead projector and a whiteboard. I pick the tool that is most appropriate for what I am teaching. The practice and interpretation of empirical analysis is done on a computer; the introduction of concepts is done with an overhead projector, and mathematical proofs are done on a whiteboard (so that the students can follow as I write on the board and so that I pace myself well).

All class materials are placed on the class web site including all data for the problem sets and the data used in class. My examples are closed to public view and I prefer to keep it that way. I don't want next year's students to print off all the material and never come to class as most students who do that end up with a disappointing grade. Better they don't have the temptation in my view.

Students like the use of technology a lot. I know this because they ask questions, they get interested and they tell me they like it. However, they would like it more if the technology at UCLA were better. The last 15% of the seats in every classroom that I have taught in at UCLA are worthless: no student in those seats can see the computer output. (The software I use is governed by the technical needs of the class so I have no control over its layout.) This is a colossal impediment. Moreover, no classroom I have used allows me to simultaneously use an overhead projector, computer and white board.

Perhaps the most difficult problem for me and for the students is that the students are supposed to have a statistical background, but few do. Integrating the class with a web-based, interactive learning tool which provides remedial help as needed on an individual basis would substantially enhance the quality and quantity of learning. Writing that tool would be a lot of work but would have tremendous value to the students I think. I would also like to add a visualization component to the class drawing on three-dimensional graphics more effectively and using multiple views of the same data to get a fuller grasp of the regression tool box.

Written Interview, May 9, 2003