TA Training: "Sticky Teaching"
In their book Made to Stick: Why Some Ideas Survive and Others Die, Chip and Dan Heath explore why people easily remember some ideas — including wrong ideas — and forget others. This and similar works have prompted educators to explore "sticky teaching," that is, how can we as teachers present material that students remember, and remember correctly?
This posting is based on a presentation by David Langley of the University's Center for Educational Innovation.
The Made to Stick book, and Prof. Langley's presentation, present six specific suggestions whose first letters spell "SUCCESS" (well, almost — they actually spell "SUCCES"). The suggestions are useful if you are giving a lecture, presenting material in a recitation, giving a research presentation at a conference, etc. The suggestions will not always all be relevant, but even employing only one or two of them will often be useful. Moreover, the suggestions are not necessarily novel; instead, the importance of the sticky teaching approach is it focuses on specific techniques which are particularly effective.
Here are the six suggestions:
- Simple. What is the core idea or ideas of your presentation? Present your main point(s) in a compact form (a short phrase, a simple diagram, etc.). For example, if you are giving an introduction to how computers store different types of data, you might use "It is all 0's and 1's." If you are introducing students to the idea of algorithmic complexity you might use "does it scale?" or might have a graph comparing logarithmic, linear, quadratic, and exponential growth. Then build your presentation around the core idea(s).
Think of places in computer science where we use short, simple messages to explain complicated concepts. One example that comes to mind is "Go To Statement Considered Harmful," which was the title of a very influential letter Edsger Dijkstra wrote that was published in the Communications of the ACM. To a computer scientist this phrase is rich in connotations, bringing to mind the evolution of programming languages, or guidelines for writing well-structured code.
Prof. Langley recommends the following steps for implementing the Simple suggestion: identify your core message, state it in eight words or less, and use it to guide construction of your presentation. - Unexpected. People often remember ideas better when they are surprising or different. There are a number of ways to do this, for example present a counterintuitive concept, or pose a compelling problem and then present the idea as a solution to it. Many key ideas in computer science and computer engineering are jarring on some level. Think for example about using the diagonalization argument to prove that the set of real numbers is uncountable; or think about the fact that computers can store so many different sorts of data: numbers, text, audio, images, video, etc. We are used to these ideas, and so we often discount how surprising they can be to someone who is exposed to them for the first time.
Prof. Langley recommends the following steps for implementing the Unexpected suggestion: vary your teaching routine, use a counterintuitive finding, show where there is a "knowledge gap" by using a compelling story or question, or relate the idea in an unexpected way to the specific background of your audience. - Concrete. As computer scientists we are comfortable with abstractions. Abstraction is one of the key ideas in computer science. Abstractions are important, and useful, and necessary. However, most students need concrete examples to understand abstract ideas. For example, if you are explaining binary search, don't just give code or pseudocode or a general description, but give a specific example of applying the algorithm to names of pets, or Canadian cities, or building materials, or … .
Prof. Langley recommends the following steps for implementing the Concrete suggestion: use images; include demonstrations; give specific names to companies, people, etc. in examples and problems; and use metaphors to explain abstract ideas. - Credible. When I grade student papers, one comment I often find myself writing is "support this claim — why should a reader believe it?" Similarly, to make important points in a presentation memorable it helps to explain why a listener should believe them. What evidence do we have? Can we support the points with data, provide an illustrative example, provide a logical argument, an in-class exploratory exercise, etc.? For example, why should a student believe merge sort is more efficient than selection sort for large lists?
Prof. Langley recommends the following steps for implementing the Credible suggestion: use your own experience and expertise, use numbers but make them understandable, provide opportunities for students to "test drive" the ideas, and teach students to think critically about the ideas. - Emotional. This suggestion might better be named "Relevance," "Caring," or "Personal Connection." Students are more motivated to learn material when they have a personal stake in it. How does the idea relate to the students' personal lives? How does it tie together what they are seeing in class and what they do outside of class? For example, compare using a straightforward explanation of how computers store numbers with an explanation that involves music: music is important to many students and is part of their daily lives, music can be stored as sequences of numbers (which can in turn be represented within the computer in certain ways), and issues such as numerical precision and range are important in music quality.
Prof. Langley recommends the following steps for implementing the Emotional suggestion: learn what your students care about, link your presentation to what they care about, and be enthusiastic and tell the audience why the idea is important to you. - Stories. The narrative form is very appealing to humans. Integrating key points into a story, or integrating a story into a presentation of a key point will provide another means to make the point(s) memorable. For example, one story I often use when discussing software development is the space shuttle software. Unfortunately there have been space shuttle disasters; however, these were all material failures. The space shuttle software is an example of how meticulous development practices can result in software with very few errors. See the article "They Write the Right Stuff" if you are interested in learning more about the space shuttle software development.
Prof. Langley recommends the following steps for implementing the Stories suggestion: make sure anecdotes, stories, etc. focus on deepening student understanding; practice the stories in advance to make sure the stories illustrate the learning points well; and use compelling language and strong images when relating the story.
Some things to think about: take any topic associated with the class you are TAing. Suppose you need to lead a recitation (or even give a lecture) on that topic. Figure out how to use at least two (and preferably more) of the sticky teaching suggestions in your presentation.