ANS 109-TAT-year-four-lessons

Subject 20.109, Laboratory Fundamentals in Biological Engineering, Spring 2011

Spring 2011 was a tough but rewarding semester. The sleep deprivation associated with having a baby taught me both what I’m capable of and what I value, along with pointing up my limits of course. I continued to expend much and well-appreciated effort on written feedback to students, and also enjoyed the one-on-one interactions that happened in person. Many – though not all – aspects of the class are either staying the same or improving. The biggest area I want to develop now is reducing unnecessary stress while retaining a full and varied learning experience.

In the lab: pedagogy of experiment design, hits and misses

The second run of the first module was on balance better than the first. Though from our perspective changes were minimal, students appreciated having greater autonomy in selecting their expe rimental conditions. Unfortunately, there were rampant technical problems — perhaps even with the plasmid stocks — that resulted in confusing data. This spring we can shoot for both elements (sufficient independence and technical robustness) being in place.

The second module, a brand new one, was ambitious and not fully formed. The students accepted this trade-off for the most part, and enjoyed above all the chance to follow through on their own designs and see how they worked. It’s clear, though, that more time should have been allotted for design, in concert with moving the computational work earlier in the module. Most frustratingly, I broke my own rule by not having a positive control that all students ran in parallel with their unique designs. Although I did develop a construct with the intended effect, it was deemed insufficiently dramatic to use; nevertheless, it may have been better than nothing. Now that we’ve run this experiment once, both the lab work and lectures (to better synergize with lab) could readily be tweaked.

I finally implemented qPCR in the third module, which turned out well. The data are not perfect, but certainly more believable than before, and the technique was a nice one to teach. There are some theoretical gaps I would like to fill in more next time, but for a first run the content wasn’t bad. The ELISA data was also improved, but the DMMB assay that I pre-ran failed in the students’ hands. I think too few cells were used.

Developing lecture technique

As usual, students found the contextualizing pre-lab lectures I gave quite helpful as a complement to the formal lectures (given by colleagues for the first two modules). However, I was both surprised and pleased to get specific compliments about my formal lectures this year. Several students found the pre-lecture review periods, in which they addressed questions in small groups before having a class-wide discussion, useful. Several also enjoyed the mid-lecture discussions of current topics (e.g., bioengineering-related ethical conundrums) enjoyable and refreshing. Given the challenges I faced this spring, it was unexpected to be praised for my somewhat unpolished lectures. However, in a case of can’t-please-everybody, one student really enjoyed the detailed discussion of techniques and what’s behind them (versus performing them blindly in a UROP) while another student found the lectures boring and too narrow.

Improving writing instruction

Some instruction is done by me and some by writing-specific faculty.

For my part, I thought I developed the perfect methods section exercise, and was surprised to see students still break “the rules” afterward. I know by now that it’s unreasonable to expect immediate implementation of abstract writing guidelines, but even knowledge transfer from similar examples seems to be slower than I thought. Figuring out the best way to order, repeat, and reinforce technical writing concepts will be key in future. On a positive note, I was quite happy with the homework assignments that I designed for the second module: the drafted Results and Introduction sections were a mix of fully written (first paragraph) and outlined (remaining paragraphs) text. I thought this gave the best of both worlds: writing without outlining can lead to rushed and ill thought-out content, while merely outlining can give a false sense of how much remains to be done.

I continue to seek the best use of the writing faculty. I selected more examples from student reports to be incorporated in the first writing lecture, and the ensuing discussion seemed helpful but limited. Specifically, I should have picked bad examples rather than only good ones, because good ones alone are harder to learn from. One student explicitly asked for even more “examination of unquestionably excellent examples,” not just for reports but for presentations.

I also experimented with having the students meet with the writing instructors rather than receive written feedback. This worked well for some students; however, the approach would surely help many more if the instructors read an excerpt of each student’s report (selected by that student as a focal point) in advance. Extensive written feedback and personal meetings would be best of course, if resources could allow.

A significant structural issue in training our students as writers is how difficult it frequently is to decouple the writing and technical content. One student wondered if rubrics from the writing faculty, similar to my expectations rubrics for the oral presentations, might be helpful. Others suggested that the writing faculty be intimately familiar with their experiments, again a good idea but potentially unrealistic given time/resource limitations.

Whether from me or others, students were most interested in having more instruction on abstracts and discussion sections. The former is doable, while the latter is difficult given the small window between final results and report due date.

Reducing stress and compression

Mid-way through, students still seemed excited about the class. Fully 22/28 filled out mid-term evaluations, and the comments were largely positive. The final evaluations were only completed by 13 students, and both the subject and I personally took an evaluative hit compared to the two prior years. While some of this change may be a fluke – given my sleep deprivation, the mess of running a complicated and brand-new module, or even changes in our student population – several comments nevertheless really got me thinking. Throughout, the students valued the individual feedback they received and the pre-lab lecture content. However, even in the mid-term evaluations there were signs of the demoralization (of an appreciable fraction of the students, in my view) to come.

General environment

In the mid-term evaluation, one student said that during lab “[Agi being] very particular about timing and smaller details… induces avoidable stress and creates a less comfortable environment.” To minimize such in-lab rush and stress, one student made the very useful comment that I assign preparatory calculations as homework more often, rather than only on the one or two most hectic days of the semester. This approach might be counterproductive in part, if the goal is to have students develop the maturity to independently do advance work, but I can probably err on the side of assigning more.

Homework assignments

Related to being too particular, a student remarked in final evaluations that “[s/he] really disliked the quantization of the homework” and suggested assigning an effort grade instead. S/he added, “Since much of the homework is re-graded as part of the major assignment, it doesn't make sense to penalize heavily for minor errors on daily homework.” Whether or not I think I am marking down “heavily” on such homeworks, I have to agree with the above logic. A different student argued that “Although the course is designed for beginners, many experienced people are taking this class. This does not make it fair [and] raises the professor's expectations of other students in lab reports.”

So I’m brought to ask the age-old educational question: am I teaching these students, or am I merely sorting them? Or less dramatically, is there a path for every student, no matter their background (within limits) and with requisite effort, to achieve mastery (also known as an “A”) in my class? We have always claimed that the homework is low stakes, but I would like to better align our rhetoric and reality. Homework is the best place for students from varying high school backgrounds to reach something like an even playing field. I will experiment next time with early homework assignments being more effort-based, and later ones more quality-based. I hope that this alone will reduce pressure and demotivation and thus increase learning. For final reports, I don’t think that the quality we ask for is too high for any student admitted to MIT, so those standards will be unchanged.

On a smaller note, a few students found that homework assignments unrelated (directly) to major assessments felt like busy-work. Such assignments are rare and generally limited to essential concepts that aren’t directly covered by the labs and reports that semester. Still, it seems clear that I need to either better motivate or cut these assignments. The overall cost-benefit may be poor if they contribute to an environment of overwork and checking off boxes rather than deeply engaging with material.

Major assignments

One student remarked about the lab report and scientific writing in general, that “as with all assignments at MIT there is never much time to really delve deep into it.” I can’t change the culture alone but I don’t need to epitomize it. Small changes to low stakes assignments will be a good start. Eventual restructuring of draft and/or final report due dates may be possible, but is not an unmitigated good.

Quizzes

A few students thought the quizzes were not optimally helpful for further learning, but “just one more thing” to get stressed out about. Even students who did well said they were unlikely to go back on their own to review missed questions. At the very least, answer keys (which I have used sporadically) are needed to close the loop, and some students would prefer going over them explicitly in class. Although I’m unlikely to cut the quizzes outright, I want to keep in mind that they are perceived by some as contributing to the general stressful environment without adding much of value.

Participation grade

I asked students directly whether the participation self-assessments were helpful or not. On balance, most students aren’t gaining enough from the exercise to justify the time and psychological cost of “just one more thing to do” at the end of each module. Instead, I will stick with requiring a short reflection about journal club, and also introduce one or two similar opportunities on days with relatively little homework.

Lab notebooks

I experimented with the lab notebook format, requiring something more akin to what I do: cite a protocol (rather than write or paste it in) and write down only the unique aspects for that day (along with contextual material such as objectives). Surprisingly, this change did not end up making lab notebooks less of a time sink for many students. Some wrote out all of the procedures by hand anyway, perhaps a practice they have picked up elsewhere. I think I'll try this method for one more semester and keep an eye on how I present it and how the students implement it.

Final module assignment

Students were allowed to work in pairs for this assignment, and little seemed to be lost. While I like seeing an explicit trajectory of individual writing improvement, there is also a lot to be learned from collaboration. I’ll keep this change for now.

Big picture learning

It was illuminating to revisit the 2010 replies to questions about how comfortable the students felt reading journal articles and designing experiments, both in absolute terms and compared to their prior experience. In 2010 the responses were simply more uniformly enthusiastic and positive than in 2011. Although this outcome may reflect stochastic fluctuations and not a trend, it’s important not to get complacent about how we teach these key skills.

Minor catch-all comments

The wiki

Most students prefer the wiki to a Stellar site, and appreciate the depth of material available. However, a few note that some sections are inefficiently set up and require a lot of page switching. Some of my extra time this fall may be spent finally reorganizing the wiki template. I should look into accessibility issues while I’m at it.

Office hours

My concern about being overwhelmed at office hours proved unfounded; students did not come in the droves that they had the year prior. Apparently, having students meet me as a recitation instructor in 20.110 the preceding fall significantly increases my perceived approachability in 20.109 the subsequent spring.

Several students on both mid-term and final evaluations requested more regular office hours, perhaps run by the TA or an undergraduate TA. All the teaching staff is stretched quiteItalic text thin already of course, while the problem with an additional undergraduate TA is that she may not have taken the same modules as are run that semester. The most efficient solution for now may be a short but regular Monday office hour run by me or perhaps alternately with the TA.

Sundry content

It was nice to hear that some students enjoy the thermodynamics-focused discussions of PCR, etc., during the pre-lab lectures.

One student wanted to hear more about how to design an experiment from scratch, how one chooses between different protocols, etc. This material is primarily beyond the scope of our class, but merits a mention or two in lectures.

One student requested even more technical guidance/tips on lab minutiae (such as reminders to mix samples before measuring), a comment to keep in mind.

Students continued to benefit tremendously from journal club. One or two students even requested having it more often, but not in a grade-critical form. Given time limitations, additional journal clubs would have to be an enrichment activity.

Teaching Assistant (TA) Training, Summer 2011

In summary, most students found the training quite complete (except for no snacks on the second day!) and several said it was useful or even fun. We have a good foundation now to tweak for next time.

Things that were done well with minor suggestions for improvement

The microteaching session was listed as a positive feature by almost everyone. The opportunity to first critique a “practiced lecturer” &emdash; i.e., Forest &emdash; was also deemed valuable.

Some folks suggested having a longer time to teach (and/or shorter problems to teach) as well as the option to choose a topic directly relevant to one’s assigned class. Now that we have a few years of experience with microteaching, the problems do seem ripe for revising.

One person suggested using a written evalution form during microteaching to allow for more candid peer feedback. Obviously there are pros and cons to this approach: it’s important to get realistic feedback in order to improve, but a stream of deconextualized negative comments could be demoralizing.

The second-most-lauded feature of the training &emdash; mentioned by many session attendees &emdash; was meeting with former TAs. Those trainees whose TA mentor had not taught their exact class assignment would have liked contact information (at the very least) for a previous TA of said class.

The first day lectures were considered useful overall but also redundant or belabored at times; some of that information can perhaps go in a handout.

Things that were missing or can otherwise be substantially improved

Most of the ideas below reflect a single comment rather than critical mass opinion, but seem worth noting.

A few topics were raised for expanded discussion: more on how to handle “troublesome” students; more about the teacher-student relationship, including outside of the classroom; tips on framing recitation structure. One person requested a checklist of good TA qualities as a quick reference after which to model one’s teaching. Another requested summary information from previous TAs on “what takes the most time and how they handle that.”

The talk on diversity and teaching was a new addition this year, and needs major revision. The gestalt comment was that it was interesting but too theoretical and narrow in scope. Because diversity can be seen as a “fluffy” topic, I attempted to put together a rigorous and data-driven talk. However, the real test of its success is its usefulness, that is, the ability of TAs to put the information into practice. In future I would like to leave sufficient time for discussing classroom thought experiments. I will also broaden the topics rather than focusing almost exclusively on issues relevant to marginalized groups. The impact of diverse learning styles, personalities, high school cultures, etc., should also be covered.

One person requested time to practice boardwork and using chalk. A potential response is to simply leave 56-614 or (likely) another space open to folks that evening. Another option is to combine boardwork with a get-to-know-your-group activity. The current (reflective questions) activity was longer than needed according to one trainee.