Difference between revisions of "Survey01"

Revision as of 19:54, 1 February 2009

My teaching methods are experimental. Quick feedback will allow me to make helpful adjustments. Answer the following questions anonymously on a blank piece of paper, fold and turn in:

• A: Judging from the first the class rate the level from 1 to 10.

1 = Too Easy
5 = Just Right
10 = Too Hard.

Results: 2, 3, 5, 6, 6, 7, (8 or 9), 9

• B: Rate the learning experience with the computer base tutorials.

1 = methods hindered learning
10 = methods facilitated learning

Results: (4 or 5), 5, 6, 6, 7, 7, 8, 8,

• C: Feel free to add any comments

Results: (Paraphrasing and grouping some of the common concerns below)

First of all I want to thank everyone for their thoughtful comments. My teaching methods are experimental and your feedback will help make the class an optimal learning experience for everyone.

Not an easy class for someone without computer background. Take it slow and explain MATLAB. Copy and pasting code gives little insight to how the code works.

Good point. I should have explained before you started Lab B that MATLAB skills are neither a prerequisite for the class nor a learning objective. MATLAB is a tool we will use in class and for the final projects. But we will use MATLAB only to run numerical experiments, by executing code that I will write. To complete the assignments and the projects you need only a very minimal skill set: you need only execute the programs, change parameters, and plot results. I do not expect you to write your own MATLAB code. There was some material about how MATLAB works in the first lab, but I consider it tangential to the course objective, and should have made that clear.

Material in first lab not related to neuroscience.

Another good point. Lab B and C introduce the concept of likelihood, which is the basis for all of the projects. The relevance of likelihood to modeling in neuroscience is that it can be used to estimate parameters of a neural model and to decide which model is best between alternative models based on experimental data. The concept of likelihood is a mathematical one that I feel is best introduced first with the random number generators rand and randn. The concept of likelihood is somewhat subtle and I am hoping that Lab B will allow you to deeply understand the lecture I will give before Lab C. In Lab C we will bring the discussion of likelihood back to Neuroscience with the Hodgkin-Huxley model.

Lecture style learning requested to augment computer based learning.

Point taken. Actually, my plan for Lab B involved both lecture and computer exercise. I was planning to go through the lab together and talk about it as we went along. I did not start that way because MATLAB repeatedly crashed the instructor's station, which led me to tell the class to go ahead with the lab without me. Once I got the computer working I found it difficult and awkward to stop the class and get back to lecturing. I did not know where students were in the lab, and people were undoubtedly going at different speeds. Besides, people were thinking hard about the lab and didn't seem to want to be interrupted. Admittedly, even if the computer had not malfunctioned I imagine I would have had similar problems interleaving lecture with computer exercise. Clearly some adjustments are needed. I plan to start each class with a lecture that wraps up the previous day's lab and introduces the new material. The length of these lectures will vary greatly, depending on the material we have to cover. Personally I feel that "doing" is more instructive than "listening," although I admit that may reflect my own learning style. Specifically I think you will be much better able to understand the lecture on likelihood that goes with Lab C having completed Lab B.

PCs better than Macs.