Students self-assess which exercises they attempted – how can I make this more homogeneous?

Question

Background

I am organising and conducting the exercises for a graduate course. The exercises mainly exist to give the students the opportunity to learn by doing. However, the students also need to participate to some extent to be admitted to a final exam. Exercises are given out as homework one week before the class. During the class, students present their solutions to the exercises, which are discussed by other students and me. We evaluate this as follows (which is mostly standard at my university):

• At the beginning of each exercise class, a paper containing student names and the numbers of the exercises is passed around and each student checks those those tasks for which they feel that they can present a reasonable attempt to solve it. This attempt does not need to be complete or correct; they should just be able to show what they tried and, if they failed, elaborate where they are stuck.

• During the exercise class, the student presenting a given task is selected at random from those who checked it.

• Tasks are so small that most of them cannot be reasonably divided into subtasks. (This should reduce the uncertainty how to report when only half of a task was attempted.)

• To be admitted to the final exam, a student has to have checked at least half of all tasks. There is no benefit for students who have checked more exercises.

Mainly, the system seems to be working well: I haven’t spotted an overly optimistic self-report yet and I currently expect that 0.3 students will fail the exercise criterion.

Actual Question

My biggest issue with this system is that there is a huge variation amongst students regarding how optimistic or pessimistic their self-assessment is. In particular the system can make life more difficult for pessimistic students. For example it happened that a student who (legitimately) checked a task and was chosen to present her attempt was stuck at some point, and another student volunteered to help out even though he did not check the task.

Hence I am asking: Is there anything I can do to effect less variability in these self-reports? In particular, I would like to flatten the pessimistic side of the spectrum.

And just because somebody is bound to remark that life is harsh and the pessimistic students need to be more optimistic anyway: Yes, but I may be able to help them learn this.

Answer 1

If the problem (being too pessimistic) is not too common, let the student check the exercises not only in the usual way, but also with "?".

Announce at the beginning that if students are totally unsure whether or not the may check an exercise, they may check it with "?". Then, whenever a student checked an exercise with "?", let them present this exercise (with no negative consequence whatsoever) and then say, if the attempt was resonable enough to warrant a checkmark. Announce when explaining "?", that students with "?" will have to present. (This has also the advantage that students see a not-so-perfect solution which is often more helpful and more intuitive then a perfect solution.) However, this only works if exercises do not get too often marked with "?" (as otherwise, this system could be gamed).

Some additional thoughts:

• In my experience with similar systems, the "pessimistic" students are often the excellent ones, that is, the ones that pass the course anyway.

• A lot of students have exactly the number of checkmarks they need to pass the course (and not more). I do think (but without proof) that if students are just at the border between passing and failing, they will do enough work to pass and get more "optimistic".

• Announce in the first lesson as clearly as possible when an exercise can be checked and also, when it cannot be checked. It is often useful to give examples when it can not be checked as students may compare their situation to the example.

Answer 2

Would it be possible to get students to "grade" or "mark" other students submissions - blindly of course?

This may well help with those who are too pessimistic on themselves.

The rule would be that it is "constructive criticism" only and marks have to be justified :

laid out nicely +5

or

confusing layout +1 ...

Edit, you could also consider "recognising" the "best" for each assignment - anonymously i.e. this is the best this week etc... might add some competition.. or the prize is a bar of chocolate...

Answer 3

I'm not sure what your field is and what class you are teaching. But, if it was a physics class, I'd prepare problems so that there are a few intermediate results before the final one.

For example suppose you need to solve a problem in which you seek x. To get x, you need to find x1, x2, x3. Then ask the class if they found x1 -- maybe write it on the paper and show it to you. People who answered correctly can check the task.

Answer 4

I wish you had given us the actual wording of the question on the sheet that gets passed around. I have a feeling this is a key part of the problem. If you can give us that wording, I'll see if that helps me improve my answer.

Here are some ideas for you to consider. I don't mean, with this list, that you necessarily have to do all of these things. The numbers are intended only to make discussion easier.

1. Collect some data and look to see if there's a gender correlation with timidity/pessimism. Knowledge is power. If there is a gender correlation, it will be helpful to be explicitly aware of this. (I'm not going to write specific suggestions based on an assumption that there is a gender correlation. If there is, I'd encourage you to write a separate question.)

2. Ask the students for feedback and suggestions. Because you have a problem with timidity in your section, I suggest asking via a short written questionnaire, not via a group discussion.

3. There should be multiple ways that students can demonstrate knowledge, skills and effort.

4. One of my fondest memories of being a student was my freshman calculus class, which had chalkboards on all four walls. The last 20 minutes of class, the teacher had ALL the students in the class write up a different homework problem on the board. That experience was fun! Like when an only child of two or three begins to experience "parallel play" in a play group. There was a bit of opportunity for interaction with one's neighbors, and the barrier between instructor and student came down. No one felt put on the spot, because we were all writing our work on the board.

5. Timid students may benefit from being told in advance which problem they will be presenting to a group. It might be even more helpful to let them choose the problem they want to present. Timid students may be hesitating to check off some of the problems they did look at at home, if they feel nervous that they might be asked to present certain problems that they feel unsure about.

6. It might be beneficial to change the way the question is worded on your sheet. I wish I had access to the actual wording you're using. Here is a draft wording that might give good results in English:

   "For each problem that you looked at outside class, please put a check mark under 'attempted'. For each problem you would be comfortable presenting to the group, put a check mark under 'willing to present'."

7. Ask someone with excellent observation skills to come in as a fly on the wall to observe the behavior of the more confident students, and your own behavior. There are subtle things that can have a big effect on timid students. Depending on what exactly is going on, it might be helpful to do some work in subgroups.