Framing

Framing is a construct developed in anthropology and linguistics to describe how an individual or group forms a sense of "what is it that's going on here?". We frame an event, utterance, or situation by interpreting it based on previous experience. E.g. when we see someone running like a madman on the street, we may interpret that as a fugitive on the run, and may expect someone else is chasing after him. Students may look at an exam question and quickly associate the same question with a previous exercise problem she has seen before.

Epistemological framing refers to the way learners form a sense of what is taking place with respect to knowledge, e.g. what past experience or knowledge is relevant to complete an assignment. Social framing refers to the way people form a sense of what to expect of each other, and of themselves in a social setting, e.g. what students expect from each other in a group project. Social framing can be observed through people's behaviors. Epistemological framing can be deduced through student learning assessments and their problem solving skills.

Based on the idea of social and epistemological framing, Scherr and Hammer (2009) studied how student interact with each other in physics tutorials. They coded student behaviors based on whether they work alone, discuss with each other, discuss with the TA, or just social, and correlate with student thinking, and their epistemological framing. They show that the behavioral cluster are evidence of student epistemologies. In particular, sitting up, speaking clearly, and gesturing frequently are evidence of novel reasoning and mutually constructed understanding.

Reference:

Scherr, R. and Hammer, D. (2009). Student Behavior and Epistemological Framing: Examples From Collaborative Active-Learning Activities in Physics. Cognition and Instruction, 27(2), pp 147 - 174.

Designing Effective Questions

Good questions that engage students in discussions are essential in peer instruction, whether these questions are posed after a mini lecture (Mazur, 1997) or as the core of in-class instruction (Beatty et al, 2005). Every good question should try to achieve three goals: content goal (deals with the subject material that you want to illuminate, or the what's), process goal (deals with the cognitive skills you want students to exercise, or the how's), and metacognitive goal (deals with the beliefs about learning, thinking, the subject area, etc.).

Beatty et al. propose four tactics in designing good questions. They are listed here in the order that may be appropriate for an one hour lecture where usually four questions can be quite easily incorporated into the lesson:

1. Tactics for directing attention and raising awareness. Focusing student attention and increasing student motivation in learning are important aspects at the beginning of each lesson. Some of the ways to achieve this are to ensure the questions (or invention activities) have all nonessential material removed, provide opportunities for students to compare and contrast different cases, extending a familiar case to something different, setting a trap to show student misconceptions.
   
2. Tactics for promoting articulation discussion. Using unstated assumptions, deliberate ambiguity, questions with multiple possible answers, students can be challenged to discuss and articulate their thoughts, ideas, and to clarify the topic to be further presented.
   
3. Tactics for stimulating cognitive processes. The fundamental rule here is to ask questions that cannot be answered without exercising the desired habits of mind. Some of the methods include asking questions that require students to interpret representations, understand a process or algorithm (rather than just memorizing a formula), having students describe the meaning and to choose from a set of possible ways of solving a problem, comparing and making contrast of different cases, and having students identify the necessary information to continue in their learning.
   
4. Tactics for formative use of response data. By revealing other students' response to a question posed before via a response histogram, a follow up question can be used to drill further down into common student misconceptions and clarify the differences among them. Having students to explain their choice of answers also promote learning and discussion in the classroom.
   

References:

Beatty, I., Gerace, W., Leonard, W., Dufresne, R. (2005). Designing Effective Questions for Classroom Response System Teaching. American Association of Physics Teachers, American Journal of Physics. 74(1), pp 31 - 39.

Mazur, E. (1997). Peer Instruction: A User's Manual. Upper Saddle River, NJ: Prentice-Hall.

Prospective Adaptation

One of the many goals of an educator is to prepare their students to adapt what they have learned in new situations. There can be two types of adaptation: fault-driven adaptation (which are reactions to a difficult situation), and prospective adaptation (which are proactive reformulations of one's knowledge or environment prior to encountering a new problem of situation). Martin and Schwartz (2009) show that graduate students uniformly make prospective adaptations to create meaningful representations of available information much more often than undergraduate students before diagnosing a problem, even though this may cost them some start up time. Undergraduate students who do not have continuous access to reference material tend to create more meaningful representations than students who have continuous access. The long term benefit for the graduate students in creating meaningful representation through prospective adaptation is that they complete a new diagnostic task much quicker than others with no meaningful representation.

In Computer Science, as in many other Science disciplines, students are not usually given the time or opportunity to step back, reflect, and retool one's knowledge. In first year programming, students are taught to solve computing problems through a systematic and methodical way while they may not understand why "hacking" is not suitable. But rather than short circuiting this process of prospective adaptation which the students should work through themselves, perhaps the students should be allowed to hack their code, and then asked to step back to rethink what other ways of problem solving is more appropriate when given a more complex problem. Instead, most often, algorithmic or procedural formulations are provided and students often try to memorize these solutions, hoping that these are sufficient for any new problems they will encounter. A learning goal should be included in each course where students are expected to engage in prospective adaptation to create new representations and ways of integrating new ideas with their knowledge base. This can be in the form of creating concept maps, writing up summaries of new knowledge and its connections to other areas (e.g. through a blog), designing new procedures or methods in solving problems, or engaging in invention activities (which encourage risk taking at little cost).

Reference:

Martin, L. and Schwartz, D. (2009). Prospective Adaptation in the Use of External Representations. Cognition and Instruction, 27(04), pp 370 - 400.