play in driving their perceptions of their peers' consumption of alcohol. The potential screen of anonymity provided by clickers obviously can serve a useful purpose when asking students to make predictions based on their own personal behaviors. It also could prove useful anytime you feel students might not want their predictions shared publicly, either because they want something kept private or because they feel they might be embarrassed by making a wildly incorrect prediction in front of their peers (or in front of you).
Prediction–Exposure–Feedback
Even without formal testing or the use of clickers, you can always ask students to make informal, in-class predictions about any course material to which they are about to be exposed. This could happen in almost any discipline, in any type of class. Scientists know full well how prediction plays a role in the scientific method—in the form of the hypothesis—and likely already ask students to engage in predictive activities in their use of laboratory experiments and reports. But outside of the laboratory, and in other disciplines, instructors can still follow this same basic approach. Get into the habit of asking students to make predictions about new content based on their knowledge from earlier in the semester, from their previous courses, or from their own general knowledge. How Learning Works gaves two quick examples of this: “Before asking students to read an article from the 1970s, you might ask them what was going on historically at the time that might have informed the author's perspective. Or when presenting students with a design problem, you might ask them how a famous designer, whose work they know, might have approached the problem.” In these kinds of questions, again, you are requiring students “not only to draw on prior knowledge but also to use it to reason about new knowledge” (Ambrose, Bridges, DiPietro, Lovett, and Norman 2010, p. 33). Ideally, you will both ask for the prediction and give them the opportunity to explain why they made it; doing so will require them to examine their thinking and might help them recognize their knowledge and skill gaps. Even more ideally, after you give them the answer you might ask them to explain why their predictions did or did not hold true.
Earlier in this chapter, I described for you three experiments on learning designed to demonstrate the power of predictions. Imagine that I was describing those experiments for students in a psychology or education course. I describe the setup of the first experiment—with one group of students trying to memorize the word pairs for 13 seconds, and a second group making their guess and then seeing the correct answer for five seconds—and then stop and ask the students to predict the results. I could do this in a generic way, and just invite some students to raise their hands and speculate, or I could offer specific options and invite students to select the one they think would be correct:
Memorizers performed 50% better on the test.
Memorizers performed 10% better on the test.
The two groups performed about equally.
Predicters performed 10% better on the test.
Predicters performed 50% better on the test.
(If all is going well in this chapter, you will remember that the answer is number four.) Students could select the answer they think was correct using polling or simply by raising their hands. After I record the student votes, I might invite a few students to explain their answers, and then show the results. After those results have been revealed, I can use them to make the two points that matter the most: prediction enhanced learning, and it did so by a small but significant margin. Once students have seen the results of this initial experiment, the next two experiments I tell them about will help confirm the knowledge that they have now lodged more firmly in their brains.
If your teaching routine includes describing surveys or experiments for your students, consider varying the typical pattern of describing the setup and then showing the results. Instead describe the setup and ask students to predict the results. After you have shown them the results, invite them to reflect upon the accuracy or inaccuracy of their predictions: Why did they get it wrong or right? What did they learn when the results were revealed?
Closing Predictions
Predictions can close a class as easily as they can open them, but in the case of closing predictions you are pointing students toward the material that they will be reading or studying for homework. Many textbooks include prediction-style questions at the beginning of a chapter; I suspect that few students read or think about those questions unless they are specifically required to do so. If you have spent a class period finishing up Chapter Five in your textbook, you might end class by asking students to answer one of the prediction questions at the beginning of Chapter Six. When they are sitting down to read Chapter Six that evening or the next day, and spot that question again, they will recall the predictions they made in class, which should (re)-activate their prior knowledge.
You can push this activity one step further by asking students to revisit a closing prediction question in the opening minutes of the next class and reflect on whether or not they got it right and why they did (or didn't). To continue with the example from the previous section, let's assume I had presented the three experiments on prediction and learning in my class period on Monday. But I had reserved a fourth case, one that introduces a new wrinkle to the research, for the end of class. In the closing minutes of the period I describe the setup of this last experiment and ask them to predict the results. But this time I send them away without revealing what happened, and leave them with a teaser: “You'll find the results of this experiment in your reading for Wednesday's class.” When students return on Wednesday, I can ask a few students to remind me what they had predicted on Monday, to describe the results they read about, and to account for the difference: What principle or new idea explains the discrepancy between the first three experiments and the fourth one?
PRINCIPLES
Asking students to make predictions requires a very small investment of time, which makes predicting an ideal small teaching activity. The following principles can help guide the creation of prediction activities in your classroom.
Stay Conceptual. Remember that part of the reason predictions work is that they require students to draw up whatever knowledge they might have that will assist them in making their prediction. If you ask them questions that are so specific that they have no prior knowledge to activate, you won't see this benefit. It seems unlikely, for example, that asking students to predict the meaning of a word in a language with which they are totally unfamiliar, in a different alphabet, will offer much learning benefit. Focus prediction activities on the major conceptual material that will maximize their learning in the course.
Provide Fast Feedback. Close the loop on every prediction your students make by providing feedback as immediately as possible. Predictions made at the opening of a class session should be addressed within that class session. Those made at the end, even if they will be answered by the reading or studying that they will do for the next session, should still be reviewed at the opening of that next session. Predictions made in online environments should provide feedback within the same session they are made. Remember that you don't want wrong predictions hanging around in students' heads for very long; the more immediate the correction, the better.
Induce Reflection. As Daniel Willingham has argued, “Memory is the residue of thought” (Willingham 2014, p. 54). In other words, we remember what we spend a little time thinking about. Prediction provides an excellent spur for thought, in that you can ask students to think about why they made their prediction, what actually happened (if the prediction leads to direct observation),
