Molecular Graphics ManifestoWhat do the students think? |
|
On turning in their assignments, my students complete a survey form, assessing the impact of the graphics tutorial and assignment on their learning of protein structure, as well as on their overall computer skills. I also ask for an accounting of time they devoted to each part of this work, and for projections about future use of the program within and beyond this course. Finally, I ask for comments and suggestions. The following results are based on the most recent offering of the course (fall 2000).
Results (and some wishful thinking)
Grades, which depend most heavily on getting file contents (model views) precisely correct, were uniformly high -- from mid-B to mid-A -- demonstrating that all students can succeed in this project if they simply invest enough time. If I discard the highest and lowest claims for time spent on each task, the approximate average time spent on the tutorial itself (sections 1-6) was 4 hours; on file preparation, 6 hours (perhaps implying that more time should have been spent on the tutorial); and on the research and writing, 5 hours. In my opinion, this total of approximately 15 hours is a reasonable, perhaps even modest, demand of student effort for a project that counts about fifteen percent of the grade in a 3-semester-hour course.
Eighty percent of the students believe that they understand protein structure better than they would have without the use of graphics in the course. Student estimates of the contribution of various course elements to their learning of protein structure were as follows: text: 30%; problems: 22%; classroom: 22%; and graphics: 26%. When asked to estimate these percentages if graphics work were not available, the results were text: 41%; problems: 26%; and classroom: 33%. So students claim that the graphics project reduced about equally their reliance on the text and on classroom activities for their knowledge of proteins. I am particularly heartened by their perception of reduced reliance on the classroom as a source of knowledge. Third- and fourth-year undergraduates should be finding and welcoming ways to become independent learners, and should be learning that the classroom can play at most a guiding role in their mastery of a subject as complex as biochemistry.
Difficulties (and where they really come from)
Student ratings of their own computer skills at the end of the project were as follows: strong, 28%; average, 64%; weak, 7%. Judging from questions that arose and help I provided along the way, I find these ratings highly optimistic. Most of the students' difficulties involved simply finding and starting the program (or installing it on their own computers -- not a required activity), finding and opening a file, and moving among the windows within Deep View. These operations call on general computer skills, not on Deep View specifics. I suspect that many students over-rate themselves as computer users, often simply not knowing how much they are missing. I have learned, but too often I forget, that it is important to give good guidance in these fundamental matters at the outset. Otherwise, students will meet early frustration and direct it inappropriately at a specific program or task. About half of the students indicated that this project helped them to improve their general computer skills, down from previous years when students brought less computer experience into the course. About 80% of the students said "yes" when asked if they planned to use Deep View further, including in other courses.
Overall, student opinion of this part of the course was strongly favorable in the fall of 2000, as it always has been no matter how I have built graphics into my biochemistry course. Opinions appear to vary with the individual student's taste for computer work in general. Prominent among the suggestions from the survey was one that I am adopting in the coming year: to give an initial hands-on demonstration to Deep View instead of a classroom demonstration. This fall, at the time of assigning the tutorial and project, I will hold one meeting in a computer classroom to get students over some of those first bumps. With this experience and my daily use of the program in class, I hope that the less computer-adept students can get off to a smoother start.
Another helpful suggestion I am adopting this year is to ask for one or two additional views of the assigned protein that illustrate some unique aspect of its structure or function, or that simply present their model in a creative way. I hope that this addition will encourage students to play around more with program functions, rather than just slavishly producing views according tp strict specifications. A few of my fall 2000 students, no doubt including the one who made the suggestion, provided some nice additional views of their models.
Here is one of the more thoughtful student comments:
Initially, I was quite apprehensive about this project. I was worried because I didn't understand exactly what was expected and what the description was talking about . I had somewhat of a difficult time loading the program but eventually got it to work. I felt the tutorial was good at explaining how to work the program but the best learning tool was just playing around within the program itself. I completed the first two models [the student means views], which took me quited some time, and after the first couple models it went much more smoothly. As I continued with the models, my understanding of what to do and of my specific protein began to grow. I actually enjoyed the project (especially as I reached the end!) and really got to know my protein. I certainly would not have learned the intricacies of proteins, in general, and would not know how to use this specific program. ... I feel that the project was very worth the time and effort.
Learning Assessment (and other myths)
I am frank to admit that I cannot assess whether students get a better understanding of protein structure with or without the graphics component. Although I certainly have faith that they do, I am highly suspicious of assessment data that supposedly prop up such claims. Judging from my experience with biochemistry classes of modest size (15-30 students) at a public university, the amount of variable noise in overall class performance from year to year easily swamps out any signal induced by changes in teaching methods. Despite the cyclical trendiness of cries for assessment and accountability, I believe that almost all teachers find themselves in this not-really-so-unenviable position: having to trust their hearts in finding better ways to teach. With specific regard to to the use of graphics, it has helped me so much in my own understanding of proteins that I am compelled to believe that it will help students, too. I also believe that students are pretty good judges of just what makes difficult concepts click into focus for them. According to the survey data, they believe that Deep View shows them more than they would see otherwise. While that might not be good enough to satisfy a state legislator or a reviewer for an education journal, it's good enough to guide my future efforts.
Figure: 'Twas the night before deadline... . Click to enlarge image.