Replacing lecture with questions

How to cover it all

The first semester I tried to use a classroom “clicker” system I tried to do it all. I committed converting my teaching to the new style, which on this first pass meant using questions during every lecture. I had listened to the physics group, and appreciated the importance of drawing out misconceptions. I also appreciated the importance of having student voices their ideas and leaving them to think about the questions. I refrained from just giving answers.

I wasn’t quite sure how to write good questions though. The physics guys had this collection of clever problems in which they had balls rolling down ramps or weights swinging on ropes and students were supposed to determine the forces or velocities or the like. It wasn’t clear what the corresponding sets of problems were for biology. I tried to write some good questions, but I used a lot of my questions from old quizzes and exams.

In this first try, I didn’t change my lecture content much. My initial, and I think misguided view was that I would use the questions and clickers as a little mini quiz. I would lecture for 15 minutes or so and then ask a question to see if they got it. If they answered correctly, I could move on, if they didn’t, I would see what was wrong, fix it, and then move on.

I quickly ran into time problems. I was lecturing just like I used to, and somehow forgetting that I was already pressed for time trying to deliver all that content when I didn’t ask questions. Now that I was asking questions too, and having the students ask lots of questions, I was in real trouble. I tried to speed up my delivery, and race through the lecture so I could squeeze the questions in. I didn't work very well

I also noticed two important things about the questions I asked. First, I noticed that standard multiple choice, recall-intensive questions fell flat. Students wouldn’t have much to say about why they chose a particular answer. They had either a disgusted, fine-you-got-me-with-a-trick-question look, or a we-get-this-already-can-we-move-on-yet? look. I was losing time with lecture, and students were either getting it quickly, or confused by my questions. Distressingly I wasn’t sure my miniquizzes in the middle of lecture were actually helping students learn anything.

Students did ask some very interesting questions occasionally. They wondered about how systems worked with each other, or what implications were with respect to cancer or development. I loved those questions and happily discussed implications of the topic at hand. The problem was that I then had to return to lecture. The next things in my lecture notes were rarely in line with the interesting ideas we were just discussing. Both the students and I had trouble matching the questions with the flow of the lecture, especially when the discussion headed in an interesting direction.

I also tried to ask some questions that were problems I thought the student would be able figure out readily if they had been paying any attention at all during lecture. I was routinely surprised at the trouble students had with these problems.

Here's an example of a question I asked in the first or second year of teaching with questions. I had assigned the gene structure and gene expression chapters. I had lectured on transcription, RNA processing and translation. I thought this one should be pretty easy after all of this.

******** The sequence of amino acids in a protein is determined by the sequence of bases in the coding sequence of the mRNA.

Where can the coding sequence be found in the DNA?

1. within a single continuous “block” in the middle of the transcribed region 2. within the intron sequences 3. within the regulatory sequences 4. split apart into multiple “blocks” in the transcribed region. 5. two of the above

*********

Students picked 1, 2 and 5 the most, but several also picked 2 and 3. They didn't understand how introns, or exons could be DNA. They weren't quite sure what all the terms meant. They had trouble thinking from the coding sequence "backwards" to the DNA. They really were uncomfortable with the idea that some genes have the coding sequence interrupted by intron sequences, while others do not. They didn't like the ambiguity. "Which is it? What's the right answer?" they asked. I felt I was on to something, but worried that I was causing a lot of confusion. I still had the time problem, the lecture all the content out problem, and now a clear up confusion from the questions I was asking problem.

I reflected on the physics approach. The physics professors would not lecture much, and certainly didn’t try to cover all the old content of the traditional lecture version of the course. They also didn’t ask “quiz” questions that were recall intensive. They would present problems to solve. While working these problems students would have to know the “laws” of physics, which of those laws applies to the problem, and how to apply them. The physics faculty argued convincingly that students only really learn the content by using it. To adopt this approach the challenge was to create a set of biology “problems” that could achieve the same thing.

This presented several challenges. First, how do you not lecture and expect students to be able to solve rich and interesting problems with information you didn’t tell them? Second, how do you design rich and interesting problems that will have enough to them? These questions are going to drive student learning. They must be complex enough so that as students solve the problems, they are forced to use the terms and concepts that are the “content” of that lecture. Though professional biologist will quickly talk about the problems they are trying to solve in their work, biology textbooks don't consistently presented the content that way.

Web-based content "delivery"

The solution to the content delivery without lecture problem was solved with a web site we called the “Class Preparation Page”. We created what now looks like a pretty standard course web site. Each weekly Class Preparation Page contained several things

A list of key concepts

Guidance to prioritize readings and highlight important topics

Phrased as questions and not answers

e.g. What is the difference between cell determination and differentiation? Reading assignments

Specific pages and sections Frequently specified sections of multiple chapters Emphasis placed on particular models, diagrams or figures.

Web links

Sites with relevant and important tools or information

These were occasionally “teaching” sites with simulations or animations, but were more commonly “relevance” sites of groups who worked on the topic for a living.

On line quiz

Duck! An on line quiz focused on feedback for each student response. Taken before class. Reinforced key content required for in-class problem solving. Multiple choice format with a twist.

Instructor feedback provided for each student choice

No credit for right/wrong, only for taking the quiz

Quiz and feedback remains on site permanently, encouraging use of the site for review use by students.

We reasoned that a well directed set of assignments, and supporting materials on a web site could go a long way toward getting students prepared for class. The on line quiz would reinforce student use of the site, emphasize key concepts and vocabulary. Student use of the site has been excellent with well over 90% using it more than once a week. AT face value, the reading and web site use by students was intended to replace the lecture. Though we created the site without using a classroom management system, similar sites can clearly be created using, LON-CAPA, Blackboard, WebCT or a range of other open source or commercial software platforms.

Interestingly, we found that students would do the reading and use the site, but were still significantly challenged by the problems we presented in class.

Question design has evolved considerably in six years. Many lessons were learned and some of them are worth sharing. Rather than presenting our struggle in chronological order, I will talk about where we are now, and then reflect on the struggle to get here where it is appropriate.

Lecture outlines and problems that fit

During standard lecture preparation many of us start with an outline. We list key concepts, terms and content to cover. We collect sets of visual aids that would help present the material. We then fill in the details to deliver a continuous lecture. I suggest a modification to this sequence that starts the same way, but ends up with problem sets rather than a continuous lecture.

Organize the content like any lecture Determine which models of biological systems are related to this content. Consider how the models would be used by a professional biologist.

Why would a professional biologist need to know this? How could a biologist use the models to solve a problem, design an experiment, predict results or interpret data?

Describe a scenario faced by a professional that reflects these activities.

The above activities define a “problem space” that you can describe to students. Working with this information, it is now possible to ask an array of questions that require students to use the models to design experiments, predict outcomes, or interpret data, or modify the models according to results.

Here’s an example. The content at hand fits in the context of the regulation of gene expression characteristic of specific cell types.

Gene Regulation and cell type Concepts

Vocabulary

Transcription factor enhancer / silencer / binding site DNA structure/bending RNA polymerase TATA box Transcription start Regulatory region/promoter Cell type / determination / differentiation

So how can this content be delivered through problem solving?

There significant area of research in molecular biology of how coordinated gene expression occurs. The research involves promoter mapping, characterization of transcription factors and assays of gene expression. Groups investigating these events create models of functional promoters based on transcription binding site, and whether the binding of those sites positively or negatively affect transcription. There are many examples of promoters mapped with binding sites of specific transcription factors defined and effects on transcription characterized. These were problems that research groups have worked for years to solve. Perhaps there was some way to digest these research problems into a more concise problem space for students to learn from.

I decided to create my own gene with a mapped regulatory region. I designated binding sites, and transcription factors that recognized those sites. I also specified the ways in which those proteins affected transcription. I included RNA polymerase stimulating transcription factors and DNA bending proteins that facilitate physical interactions between the transcription factors and RNA polymerase. Finally, I set up a scenario in which I had a set of three genes that had regulatory sequences with some but perhaps not all of the binding sites. This scenario also had three cell types each of which had particular combinations of transcription factors and DNA bending enzymes. The diagrams below help organize this information, and was made available to students to use during problem solving.