Monday, March 28, 2011

Session 4: Responding When Students Don't Get It

Eventually, the session will be archived here. I recommend it. It peters out a bit at the end with the teacher vids but the first part is definitely informative. I'm guessing this is information from the book Checking for Understanding but I didn't confirm that.

Like Guskey, Fisher is worth seeing. He's an engaging speaker and he's able to add a lot more subtlety than what's in his books. I wasn't a huge fan of his work before but now I'm definitely going to take a second look.

Also, I don't know the details here, but at least at one point he was actually in K-12 schools and even teaching classes. A higher ed guy who's actually working in schools? Insane concept. Frey is less interesting as a speaker but I think they realize that. She has smaller speaking parts but they make a good team.

Main idea: When a student doesn't get something, we do the work and jump in. Don't.

Things I found interesting, but not necessarily that I agree with. My quick thoughts are in italics:

Fisher focused specifically on what we do when a student doesn't "get it." In this case, he's talking about either doing a skill or answering a question, not in a global failing-the-class kind of way.

There was a flow chart in his preso, which I think you can download, about what to do at each step if a student was stuck. The flow chart went like this:
  1. Start with a robust question.
  2. Prompt
  3. Cue
  4. Go to direct explanation and modeling.
Robust questions are designed figure out what a student is thinking.  They should uncover what the errors and misconceptions are so we can respond. Fisher identified six types of questions.
  1. Elicitation
  2. Elaboration
  3. Clarifying
  4. Inventive
  5. Divergent
  6. Heuristic 

Our difficulties problems come after the question. We don't have a strategy planned if a student gets something wrong and our instinct is to automatically go to step 4 and explain. If we immediately move to explaining, the student becomes dependent on the teacher.

I like things like this. I don't think of myself as an instinctive teacher. However, I'm very good at identifying certain weaknesses, researching how to improve, and putting that into action.


When we identify an error, we need to prompt, not takeover. A prompt is about getting something going in a kid's brain. The most common prompts are background knowledge or process prompts. For example, prompting the broken rule (PEMDAS) or recalling certain knowledge. We also use reflective prompts, like, "Does that make sense?" He also talked about heuristic prompts. These were prompts for strategies, like "Why don't you make a graph and check?" Fisher stressed the importance of students developing their own strategies that worked for them.

The next step was to cue. Cues say, "Pay attention here." Cues let novices see things through the eyes of the expert. Fisher gave a good example of when you watch Olympic diving. It all looks the same to most of us. An expert can slow it down and point out certain things. When you're an expert, you can pay attention to more things simultaneously.

This is a good point about cues. I know I can fall into that huge trap of expecting students to experience something the same way I do. 

Types of cues:
  1. Visual 
  2. Physical
  3. Gestural
  4. Positional
  5. Verbal
  6. Environmental
Ultimately the idea is to get kids to pay attention to the relevant details. Fisher brought up a good point that we're really good at using cues in our initial teaching, but when a student gets stuck, we usually just tell them additional information.

When all else fails, go for direct explanation. Even then, Fisher had some good advice. First identify the error and explain. Think aloud while you're explaining the error. Finally go back and monitor. Re-assess somehow to make sure they actually get it.

"Telling and leaving" could describe a large portion of times when I say I'm helping a student. Modeling by going through the think aloud process and then going back to monitor are crucial, but often forgotten steps.

He finished with some extended videos. They weren't great but I liked that he acknowledged they weren't perfect. He also acknowledged that this whole process is much easier in small groups. You'll lose the whole class if you go through all this with one student. You have to have them engaged in something else.

He finished by talking about what a huge challenge it is to undo the expectation by a student that they'll just be told the right answer.


Takeaway: I don't know if there are better models out there. I wouldn't argue it's perfect. But I do know that it's better than our standard, "Ask a question-if wrong-answer it ourselves," method. It's not just a poor learning strategy. Any high school teacher with a student who is used to just being told the answer can tell you what kinds of damage we do to a student's feelings of self-efficacy with our standard methods.

When the archive goes up, I recommend you watch at least the first half.

Addendum: The Science Goddess blogged this session as well. She's got her notes and a link to the handouts. 

2 comments:

  1. I was watching this live over the internet. I agree with you when you say, "I wouldn't argue it's perfect. But I do know that it's better than our standard." For me the flaw is this: The focus on when students "don't get it" makes it seem like understanding (or learning) is built by layering correct things on top of each other over time. It creates an obsession with correctness (in each and every moment) that is just as damaging to students' self-efficacy to the standard.

    ReplyDelete
  2. "'Telling and leaving' could describe a large portion of times when I say I'm helping a student. Modeling by going through the think aloud process and then going back to monitor are crucial, but often forgotten steps."

    I agree. I completely do that. If I'm honest, I think I don't go back not because I forget, but because I'm afraid they still won't get it. This way I can delude myself into thinking that I've been really helpful.

    ReplyDelete