CS8801 Graduate Teaching Workshop - Instructor's Postmortem
for more info or questions about this document contact Idris Hsi at idris@cc.gatech.edu
Design Heuristics and General Implementation
General
-
15 hours of class
-
60+ students in class
-
Course was integrated with 7001 to provide a better transition point for
another instructor to take over. This was also to accomodate Gregory
Abowd's desire to have the entire PhD population take the class.
-
I divided classes into thirds. 1st 3rd consisted of courses directly
necessary for TAs immediate use. 2nd 3rd consisted of supplemental
courses that would help TA with the subtle aspects of being a good teaching
assistant. Last 3rd consists of courses that may be necessary for advanced
TAs or for instructors. The rationale was to hold those classes which
would be relevant at those times where the information might be needed
- just-in-time learning. For example, we moved Difficult Issues to
the middle third because most TAs don't encounter those problems until
about that point in the term.
-
General objective was to raise student awareness of issues related to teaching.
Alterations from last year
-
Reduced number of guest lecturers from 4-5 to just Kurt Eiselt and myself.
This was directly in response to student comments about the redundant content
delivered by the lecturers as being boring and unnecessary by the last
one. (I still think it would have been useful to have had at least
one or two speakers talking about very specfic topics.)
-
Added or re-implemented the following lectures
-
Time Management (taught last year to 7100 class)
-
Teaching for Diverse Populations (last taught by Stephanie Rey in 1997)
-
Designing and Presenting lectures split into two days (in previous years
it was a 1.5 hour lecture)
-
Syllabus Design and Course Planning
-
Removed the following lectures
-
Role of the Teaching Assistant
-
Assigned more assessment materials other than just attendance and the course
postmortem. However, these were not thought out from the outset but
midway through the course - something that can be corrected on a next pass
by another instructor.
-
I think some of the original design heuristics (to have fun learning about
teaching and learning) have fallen by the wayside as the class sizes increased.
That was my fault.
Suggestions for Future Design
-
Redistribute the courses with homeworks throughout the term. The
homeworks fell on the last few weeks of the semester where students already
feel overwhelmed by work from their regular classes.
-
Uncouple the workshop from 7001. This will allow the instructor to
have weekly sessions that meet 2-3 days a week consistently so that the
class end halfway through the semester the way it did last year.
Part of the resentment (see Analysis
of Students) on the part of students came from the perceived length
of the course that seemed to drag on forever (15 hours is really only 5
weeks of 3 hours a week - this could be finished very quickly). Not
having to deal with the restrictions of 7001 will also make it easier to
schedule guest lecturers and help students plan their schedule - especially
those Masters students who aren't taking 7001 - by having a consistent
day and time of the week to meet.
Back to Top
Analysis of Lectures/Workshop sessions
In this section, I will do an analysis of each session in the course -
specifically what the workshop was trying to achieve. What worked
or what went wrong and what the students thought of that session.
The workshops are:
Objectives of Session:
-
Hand out Class questionnaire
-
Go over the basic responsibilities of TAing at the College
-
Present objectives of course
Analysis of Day:
-
The copier somehow managed to mess up the syllabus handout. Need
to double-check before handing out to students. Very bad start..
-
Handed out
Homework 1 before the lecture on Developing
Grading Criteria. This was a bad idea.
Suggestions for Improvement:
-
If a graduate student is running this course, it is vitally important that
either the Associate Dean, the Director of Student Services, or, at the
very least, a faculty member be on hand to explain the reasons why the
students are there and why they are taking the course. I don't believe
these points were made clearly enough as questions during the Course Postmortem
indicated.
-
The other purpose of the session is to find out which students should not
be taking the class. Based on the questionnaires, you can filter out those
teaching assistants that have had extensive teaching experience and/or
training. The College needs to develop a set of criteria that will
help determine which students belong and to act accordingly based on the
information given by the students.
Handouts:
Objectives of Session:
-
Review the basic responsibilities of a TA in respect to grading.
-
Review some heuristics to preserve consistency and fairness in grading.
-
Review some heuristics for addressing student complaints about grading.
Analysis of Day:
-
Needed to make clear the following points but didn't. 1. That grading
criteria, above all else, must be consistent and fair. 2. Well designed
grading criteria can have many parameters, as long as it provides sufficient
guidance for measuring a students knowledge of a subject within the design
parameters of the assessment criteria being graded.
-
What was missing from this session and the homework was feedback to the
TAs about their particular grading criteria. On the whole, they were
serviceable.
Suggestions for Improvement:
-
I think that the tough part is assigning grades and maybe this session
could turn into a series of small group exercises or mini grading sessions
looking at sample programs, exams, projects, or whatever was available
along with small discusions.
-
The most common mistakes associated with grading and grading criteria that
I have experienced or seen over the years are the following and should
be addressed in any future implementation of the workshop:
-
Omission - no grading criteria provided prior to assigning the assessment
materials.
-
Poor granularity - grades were assigned in large chunks that are hard to
quantify. For example - Part 1 (50%), Part II (40%) without explaining
further what the details of how they would be assigned (for example, what
does it mean to only get 43%/50% in Part I).
-
Poor description - the TA provided no explanation for why points were taking
off or how the question/project/essay should have been done.
-
High variability - the TA did not maintain a consistent pattern of grading
through the student papers (very hard to do).
-
Giving no or poor feedback to the students - students can only interpret
or learn from -1, -5, and comments like "bad". TAs need to practice
giving useful and positive feedback to students.
Handouts:
-
Grading Guidelines in PDF,
RTF
The Philosophy of Grading - taught by Kurt
Eiselt
Objectives of Session:
-
Challenge their thinking about what grades are for.
-
Present some rationales from the instructor's perspective about absolute
vs. relative grading scales and curves.
-
Reinforce the notions of Consistency and Fairness.
Analysis of Day:
-
The basic premise of the lecture was that students come to universities
and similar institutions to be assessed and graded, not to learn, be taught,
or even to do research. It presented a very cynical viewpoint.
-
Kurt also presented some grading heuristics, articulated from the instructor's
point of view - including how to treat student complaints about grades,
how to view curves of grading distribution (The Wild Kingdom model of student
fitness and the bottom curve of the distribution is a great story).
-
He also presented some contrasts (Abowd and Karloff were used here) of
professors and their relative approaches to teaching and grading.
-
Kurt is very good at stirring up controversy. This was a good class
discussion about motivations.
-
Several students didn't see the purpose of this lecture. As with
many of the other sessions, this was addressed to those students who never
really thought about why grades exist or what their responsibilities are,
as instructors, relative to the goals of the institution. The main
point this session was designed to hammer home is that they have a responsibility
to assess students fairly and consistently against a set of measurable
standards because they are essentially certifying these students as being
competent in a skill or area of knowledge. Mediocre or poor efforts
in this regard will either certify students who haven't accomplished these
goals or erroneously punish those students who are being held up to a set
of poorly designed standards. Both cases not only harm the students
in the long run but undermine the credibility of the system.
Suggestions for Improvement
-
None. Have Kurt write up a handout or a brief essay to supplement
these pages so that future instructors will know what he talked about.
-
Some students have talked about integrating this with the Grading Strategies
lecture. I would really prefer to keep the philosophy and practice
as separate sessions but that's just me. If for nothing else, it
allows more time for the practice portion for class activities, assuming
the population of the class can be accomodated.
Handouts:
Objectives of Session:
-
To reinforce the notions of listening before talking and the importance
of active listening.
-
Create an icebreaking situation that was low key and helped students to
meet one another.
Analysis of Day:
-
Too much chaos in getting students to pair up in a random fashion which
I engineered on purpose. Anyhow, I'm very certain that you cannot
do this exercise quickly past 40 students. I'm sure the chaos was
probably a good story later on but made the class look chaotic which probably
reflected badly on me and, indeed, some of the students said so in feedback.
-
I think the objectives were lost in the shuffle, so to speak.
Suggestions for Improvement:
-
The purpose of the workshop was to give people a sense of what it's like
to do nothing but listen to a single person (as opposed to being distracted,
thinking of a response while the other person is trying to finish, interrupting,
and so on). Usually the activity is followed up by a discussion of
the usual bits of frustration, surprise, and "learned interesting things"
comments which did not happen. Teaching assistants sometimes forget
that half of the learning process is allowing students to articulate their
understanding or misunderstanding of a situation. They are sometimes
overeager to tell the student what should be done. If someone can think
of another way of achieving these goals of teaching active listening to
a class this size, more power to them. Maybe skits by senior TAs?
-
If the class remains at its current size, then this session cannot be implemented
in this fashion. This is unfortunate because it's one of the workshops
that many people enjoyed in the past.
Handouts:
Learning Strategies
Objectives of Session:
-
Review some basic theories of education and learning theory
-
Establish foundation for understanding the range of strategies that other
people use to learn to help Teaching Assistants understand students who
come to them with problems with the class.
Analysis of Day:
-
The first fundamental problem with this lecture is that it argues against
lectures as the ideal way to teach, in spite of institutional entrenchment.
It makes for a very ironic session.
-
This class has been taught by Wendy
Newstetter for the last 3 years. Unfortunately, given the schedule
and the timing of the course, the class ended up on Friday at 5 pm. I chose
not to ask her to give the lecture figuring that after listening to her
for 3 years that I could give the lecture myself. This was sheer
folly. Wendy brings to her lectures about 10-15 more years of experience
with implementing and studying education. She would have done a much
better job at managing this audience (see Analysis
on Students).
-
The other problem with the lecture is that it really requires an hour and
a half to deliver. The last time Wendy gave the lecture, she was
rushing through her last slides and it was all good content. The
time frame of this class may be insufficient to discuss this material in
any depth.
-
Students who have never had any classes on education might this very useful.
Students who have had classes on education find this class very general
and ineffective.
-
Several students said through feedback that several found this topic to
be very nebulous, void of hard experimental evidence, empty of real content,
and not worthy of attention (is this how bad professors start?).
Suggestions for Improvement:
-
Definitely have a professor or teaching assistant with a strong background
in learning or education teach this class.
-
I think as a definite secondary goal, it should be emphasized that teaching
is a difficult subject to learn, research, or understand.
-
This is another one of those courses that might be flagged for removal
but I can't stress enough how important it is that students develop some
sense of alternative learning heuristics or even some of the cognitive
methods used in learning. One of the hard lessons that I have tried
to learn over the years that I have spent teaching is trying to formulate
alternative or multiple methods for teaching any topic or explaining things
like my thesis. Very often, we get caught up in a particular viewpoint
and force others (square peg/round hole) to try to see it from there.
Teaching becomes a lot easier when you can teach to someone else's strengths.
-
Ideally, have Wendy teach this class.
Handouts:
Objectives of Session:
-
Show students how problems can be presented in different (and hopefully
interesting) ways to assist learners who prefer a particular style of learning.
-
Provide them with another social setting with a chance to meet other students
while doing a creative activity.
Analysis of Day:
-
Overall, it went well. There were some very funny examples created
by the student groups as usual - the high point of this kind of session.
-
To my surprise, some groups actually refused to speak when I called on
them. Again, the large class size granted them anonymity and I hadn't
thought ahead to know which people to call on from belonging to particular
groups. In the student feedback section, one student commented that
this made me look weak as an instructor. I can't say I disagree with
this assessment.
-
This is another day which was better implemented in the smaller 30-40 person
classes.
Suggestions for Improvement
-
Unfortunately, this is the kind of session which was a lot of fun to do
as a small to medium size class. It will not work at this size.
-
It may be better to turn this into a shorter team or group assignment and
have teams develop several examples, either multiple modalities using one
problem or some mixture of problems with matching modalities.
Handouts:
Motivating the Students
Objectives of Session:
-
To generate a discussion on what makes students motivated to work or learn
material.
-
To present a view (slightly cynical but mostly true) of what motivates
a majority of the students at Georgia Tech and to challenge those assumptions.
-
To have a discussion on the mission of teaching and how it could be achieved.
Analysis of Day:
-
This class has been taught by Gregory
Abowd, George Riley, and Greg
Turk.
-
Covered the spectrum of motivation techniques from fear (threat of not
being able to finish a degree) to challenging the students with creative
and interesting projects (setting achievable but difficult goals - helping
students to meet high expectations). Discussed the employment motivations
of the students and some of the problems that we're seeing in the increased
enrollment at Georgia Tech in particular - for example, more students who
don't understand the implications of computer science or who don't care
about the material. Revisited the use of real world examples to help
motivate topics - "This is how it can be used in your job" examples.
As a side note, Jessica used this quite a bit in her animation classes
by showing cartoons and movies that used those techniques. Unfortunately,
not all of us have the luxury of having that quality of media to motivate
the material.
-
This day went badly and was next to useless. Again, it suffered partly
from large class syndrome but mostly from a lack of interest on the part
of the students and the inability of the instructor to carry the point
across in a coherent manner.
-
And as badly as it was going, it didn't help that halfway through the session,
Dr. Abowd said that teaching assistants don't really have the responsiblity
for motivating the students and that it really belonged to the instructor.
Whatever momentum the class had at that point ground to a halt.
-
In response to what Dr. Abowd said (and now what a few students are echoing),
I think that from a practical standpoint and the way that teaching assistants
are used at the College that this, for the most part, is true. However,
from an idealistic standpoint, you want every point of teaching to be coherent.
It doesn't help that the instructor is trying to motivate students when
the teaching assistants present the attitude that they could care less
about what the students learn or whether the students consider the material
to be important or not. I think it undermines the class.
Suggestions for Improvement
-
Definitely have this taught by an guest lecturer or remove this from the
syllabus. This may be a good time to have a panel session of faculty
or TAs (if you can get them to show up) to answer questions or to present
viewpoints on good teaching.
Handouts:
Objectives of Session:
-
Present a range of ideas related to Time and Stress management including
their interrelationships
-
Familiarize students with techniques garnered from other graduate students
for surviving graduate school in a relatively stress free manner.
Analysis of Day:
-
Went okay even though I had to prepare for it due to a last minute cancellation
by Kurt.
-
Many of students did not see the relevance of Time and Stress Management
to being a teaching assistant. The more interested students thought
it would have been better suited to being a 7001 course.
Suggestions for Improvement
-
I don't have any accept maybe to make this one voluntary or to have students
keep a journal for one week tracking where all their time goes (also a
useful activity).
-
Whatever you do, don't go to the Counseling Center for an expert to come
and teach this session. When I did this 4 years ago, they sent us
someone who taught a relaxation exercise (which was almost interesting)
but did not deliver a terribly useful session.
Handouts:
Difficult TA Issues - taught by Kurt
Eiselt
Objectives of Session:
-
Discuss some of the more difficult problems that a TA might face.
-
Familiarize TAs with the campus resources ready to support them.
Analysis of Day:
-
Kurt has been teaching this class for the last 4 years and he's a lot of
fun as a guest speaker.
-
He usually discusses the following issues:
-
Academic Dishonesty
-
How to Deal with Belligerent Students
-
Disabled Students
-
Difficulties with the Course Instructor
-
Mentally distraught students (depressed, suicidal, etc.)
-
This was a good day. Kurt's a good speaker and has a range of interesting
stories.
Suggestions for Improvement
-
Generate a handout that could go into a TA manual that covers these procedures
and resources that Kurt discusses.
Handouts:
Objectives of Session:
-
To raise the questions of "What is culture?" and "How do we react to people
different from ourselves?"
-
Discuss some useful heuristics for communicating with, teaching, and interacting
with people from different cultures.
-
Share some perceptions about what America and American education is like
and contrast them with some perceptions that Americans have of other students
Analysis of Day:
-
The discussion was not bad. The session could have been much more
effective (see Suggestions).
-
I wasn't sure how to approach this class and spent about 2-3 weeks doing
research on the subject. In the past, we got the Assistant Dean of
Minority Affairs to come teach a class on diversity issues. It was
quite interesting but only for a small population and definitely geared
towards undergraduates. For example, she pointed out that many people
had a different way of counting to 10 on their fingers (take a sample some
time, you'll be surprised) which was kind of fun. Also, through an
icebreaker activity, she pointed out how many different hats we wear besides
our nationality (middle child, from the Northeast, adopted, etc.). I was
hoping to generate a more intelligent discussion on the topic because the
problem at this level is not recognizing diversity but learning how to
adjust teaching and communication to leverage off of those differences.
Even at the College, the students have the tendancy to collect by nationalities,
rarely interacting outside of them in some extreme cases.
-
Don't use irony as a lecture technique to make a point. It's easily
misunderstood. I'm not sure how much leeway there is for spontaneity
but I'm beginning to believe that ad libbing, especially in the context
of teaching, can be dangerous. This includes discussing material
only half thought-out. In particular, there was one incident which backfired
horribly. Towards the end of this session, I tried to make an ironic
point by saying something slightly thoughtless, directed at the TA for
7001 who happened to be from Canada, saying, more or less, "Well, Canadians
are just like us anyhow." He protested (appropriately) and I apologized.
Unfortunately, it was spontaneously driven and it didn't occur to me to
check with him first (the better way to go) and it should have been an
opener, rather than a closer. Nevertheless, a number of students
took me seriously and remarked in feedback that they lost faith in me as
an instructor as a result of that. Ironically, however, I believe
that one comment served to make the point of cultural sensitivity in teaching
better than anything I said during lecture but the personal cost was too
high.
Suggestions for Improvement
-
I tried to get a panel of graduate students to help me teach this course
from a wide variety of perspectives. Unfortunately, everyone I asked
(except 1 person) was busy. This would have been invaluable to motivate
a discussion of the issues of diversity in research.
-
I still think this is a terribly important topic, even if the issues are
minimized at the graduate level. Georgia Tech has a lot of funny
intercultural dynamics that can impact the classroom. For example,
even the way that groups form shows some interesting social patterns.
It just may be beyond the scope of what this course is trying to achieve.
Handouts:
Designing Assignments and Exams
Objectives of Session:
-
To generate discussion about how assessment materials work and to analyze
the efficacy of certain types (projects, group projects, programming assignments,
exams, open book vs. take-home, etc).
Analysis of Day:
-
Pretty dismal. Not much to say here. I think the only effective
part of this day was the design of the homework which probably did a better
job of articulating the principles of good assignment and exam design embedded
in the instructions then anything that was said during lecture.
-
Assigned Homework 2
Suggestions for Improvement:
-
Remove from the syllabus. Teaching Assistants who actually care about
teaching enough to take this task on will muddle through somehow.
Handouts:
Objectives of Session:
-
Review the good and bad heuristics that go into designing talks and lectures.
-
Reinforce the concept of "Tell them what you're going to tell them, tell
them, tell them what you told them."
-
Introduce the idea that a talk should have a flow similar to telling a
story.
Analysis of Day:
Handouts:
Objectives of Session:
-
Review good heuristics for delivering the lecture including how to design
visuals, pace the talk, and field questions
Analysis of Day:
-
This class has also been taught by Ashwin
Ram and Spencer Rugaber.
In previous years, Spencer had students come up and give 1 minute elevator
speeches on topics like "Is computer science closer to engineering or science?".
He also spent the lecture giving a "bad" lecture on purpose so that students
could point out what he did wrong. Absolutely brilliant.
-
I intended the session to take no longer than 30-40 minutes. For
some reason, this session generated more discussion and commentary than
others and pushed it to an hour. Who knew?
-
As a final point of irony, I had designed both talks to be useful on multiple
levels to the graduate students taking the class so I expanded the two
classes from lectures to the range of talks that graduate students may
have to give during their careers. Nevertheless, in spite of these
efforts, I still had one student ask "What relevance does this topic have
to us as Teaching Assistants?" Fortunately, Gregory answered it by
saying "Whenever I leave town, the first person I ask to teach class is
the TA."
Handouts:
Objectives of Session:
-
Familiarize students with basic details of what should go into a syllabus
and how to double-check for flaws in course design
-
Stimulate some critical thinking about how courses are designed and implemented.
-
Review design of the workshop in preparation for the course postmortem
Analysis of Day:
-
A student pointed out that the syllabus for this particular workshop didn't
have any of the deliverables on it (true) and implied that it was stupid
for me to discuss how a course should be planned when I wasn't even following
my own heuristics. The class went downhill from there.
-
In a smaller workshop, it was easier to convey how the workshop was a work
in progress and the students would have been much more instrumental in
its formation. Again, this is an interesting topic that requires
extensive discussion but fails with this large population.
Suggestions for Improvement:
-
This class should definitely be marked for removal from a future curriculum.
It's fairly far removed from a teaching assistant's basic responsibilities.
Those TAs who are given this level of responsbility will probably be able
to muddle through.
Handouts:
Objectives of Session:
-
Discuss the importance of leaving a record behind for future teaching assistants
and instructors.
-
Discuss the implementation and success of the GTA Workshop.
-
Discuss how students can improve their own teaching skills
-
Students handed in Homework
4, assigned in the last workshop.
Analysis of Day:
-
The copy machine broke just as I tried to copy my handout for the day.
In a way, a nice closure to the way the semester started.
-
Pretty much a failure from a discussion point of view. As should
be evident by now, the class is much too large to sustain a meaningful
discussion.
-
One saving point was Gregory Abowd's presence as students asked implementation
questions (why was the class 1 pass/fail unit, why did it have to be 15
hours, what are the Institute's requirements exactly, why did we have to
be there in the first place). These questions probably should have
been better addressed in the very first class.
-
By this point in the semester, I was pretty demoralized and my sense was
the class was fairly bitter about the whole thing. I completely abandoned
the third objective and would suggest that it be dropped from the course.
As an aside, I will add that I firmly believe that those students who intend
to become professors and don't have any interest in teaching will probably
not improve themselves and those students who care about teaching who will
work on improving themselves to the best of their ability, regardless of
whether they have taken this class.
Suggestions for Improvement:
-
Improving Your Teaching was a separate course. It has been taught
by David Shook from CETL and twice by Richard
Fujimoto. Dr. Fujimoto did a very nice job of presenting good
teaching perspectives from a senior professor's point of view. If
the session is kept, I strongly suggest a guest speaker in a similar vein
- ideally, one who has one the Gus Baird Teaching Award that the College
gives each year to a top notch instructor.
-
To make the class more productive, it would have helped to look at some
examples of postmortems. Unfortunately, the class size above 60 people
prohibits the second objective.
Handouts:
Back to Top
Analysis of Assessment Materials
Attendance
-
Students were required to sign a sign-up sheet during class that also asked
them for some bit of trivia. An example is here.
Problems
-
Initially, I passed out one sheet but this took too long to circulate.
If instrumented again for populations of larger than 30, you'll need to
hand out 2 copies and have one start from the back and go to the front.
This actually helped lighten the mood of the class somewhat and was one
of the more successful ideas.
Suggestions
-
Requiring mandatory attendance is very draconian and not a method you want
to use to inspire teaching assistants to go out and do a good job.
Unfortunately, because the primary mission of the class is to certify students
as teaching assistants, this is almost unavoidable. To make this
less onerous, I think it would be good to split the workshop into mandatory
and optional lectures, requiring the students to attend N number of optional
lectures and to sign up for them at the beginning of class. It would
reduce the time burden on the students and, possibly, the effort required
on the part of the instructor. I'll mention this point again in the
Suggestions section.
Homework 1 - Grading Criteria
The assignment is here.
Support material is here.
Problems
-
For some reason, one or two of the links to the homework that I had researched
that week went down right after assigning the homework. Go figure.
-
I assigned this before the class on Grading Criteria. The idea was
to motivate a discussion on Grading Criteria. It would have been
better to have assigned it afterwards or even during the class like I did
last year as part of a class activity.
-
Regardless, it would have been good to post answers or to give feedback
on the homework. I chose not to because it didn't make sense to "grade"
an assignment that was assigned before the class. This resulted in
two problems. The first was it undermined credibility in the effectiveness
of the course. The second that any feedback would have been okay
and it didn't leave students with a sense that they were on the right track
or not. Bad error.
Suggestions
-
I still think the homework makes sense. I would try and pick better
examples somehow (and make sure that the links persist before assigning
the homeworks). I think it would be more effective as a class exercise
and discussion session if time permits.
-
Grading criteria is one of those nebulous topics that cannot be taught
as an absolute method but as a series of working heuristics. The
lesson that I failed to convey in both the lecture and the assignment is
that it doesn't matter how the grades are assigned - as long as they are
assigned fairly, consistently, and in a manner that hopefully measures,
as accurately as possible, how well the students have achieved the learning
objectives.
Homework 2 - Designing Assignments
The assignment is here.
Problems
-
Note: Support material was not available during the lecture or at the time
the homework was assigned. The assignment was meant to be reflect
and implement assignment. Most of the useful material of the session
is actually embedded in the assignment itself.
-
It is unclear whether this homework has enough relevance to a TAs daily
duties to be considered relevant or interesting.
-
Not all students had classes that had a mixture of exams and homeworks.
This led to some confusion as to which two assignments to pick. There
is a phrase in the assignment that suggests that they take one from the
web but it was not made clear.
-
The writing of the assignment had the following errors and ambiguities:
-
Part II - not clear that the students were to deliver an assignment that
was in a form that could be handed out to students.
-
Part II - some of the bullet points that were to be explained were things
like "Name of assignment" and thus, trivial. This should have been
written more clearly.
Suggestions
-
This may be better implemented as a group or pair project.
-
One of the points that was not made in the homework, except by ironic implementation,
is to avoid ambiguity in assignment design. A good addition to this
homework might be to have students look for details that may be ambiguous..
Homework 3 - Designing Talks
The assignment is here.
Support material is here.
Problems
-
This assignment was really only useful to those people who hadn't designed
talks before or who were working on poor heuristics.
Suggestions
-
Good idea in principle, bad idea in execution. The better way to
have people design talks and give them is to have them to it for this workshop.
This may not be possible given the size of the class.
Homework 4 - Course Postmortem
The assignment is here.
Support material is here.
-
Not much to say for Problems or Suggestions. I think the assignment
did what it was supposed to and gave me reasonable feedback. Leave
it alone (modify Part III as needed relative to what you want to learn
from the class though.
Back to Top
Analysis of Students and
Student Feedback
-
The original implementation of this course had 34 PhD students. The
next year had 42. The next year had about 53. This year had
64. The class has almost doubled in size from the original version.
-
This is the first year where all first year PhD students were required
to attend, independent of current funding sources. There were also
several 2nd year PhD students, some of whom were denied TA funding because
they had not taken the class.
-
I found about 7-8 students to be actively hostile or disruptive during
the lecture, mostly among the 2nd year students. They clearly didn't
want to be there from Day 1 and were looking for ways to justify their
perception of the uselessness of this course. I believe that some
of this was motivated by stories from unhappy students from last year.
It's possible that a professor or someone older would have been able to
mute their effect. This won't be a problem in future years now that
all first year students have taken this class. I made the decision
not to call them out and hold them up to scrutiny by the class. My
sense was that the class was already fairly unhappy with the current draconian
implementation and that this would simply make the environment more hostile.
Maybe this was a mistake because in a sense, I was failing my responsibilities
to the other students.
-
There was also a number of experienced students who were teaching assistants
or had good teaching experience prior to coming to this class. Again,
in a smaller discussion group, the class could have benefitted from their
experience and input. In a large class, the dynamics were difficult
to manage. So, again, they were unhappy because they were revisiting
material that they had already been taught or had experienced themselves.
-
Lastly, there were a group of students who have never been teaching assistants.
Since the class was mostly designed for them, they were the ones who found
it the most useful.
-
In short, I would characterize the audience as indifferent tending towards
hostile with increasing resentment towards the end of the term that was
only heightened by the last three homeworks which unfortunately fell during
the latter half of the semester.
Student Feedback
-
There were several students who remarked that I didn't have the credentials
or authority to teach the topics that I was attempting to teach.
I'm not sure what to say to this. I've been teaching and TAing for
7 and a half years now. I've performed nearly every activity and
responsbility given to TAs and instructors. I've been researching,
obeserving, and studying teaching issues for the last 5 years. Short
of receiving external credentials in an education degree, I don't know
what I could have done to better prepare myself. However, as I said
Back to Top
Lessons Learned
-
You can't run a class with students who don't want to be there except by
Draconian methods. At these class sizes, the probability that you
will have a critical mass of unhappy students increases (and statistically
speaking, I received 1.5 nasty Postmortems last year, this year I had 4,
if those are sufficient data points to be used as proof). When a
sufficient number of students are unhappy, the sessions, designed as they
are, become useless and ineffectual - which only serves to contribute to
the unhappiness.
-
In general, the class ran as about as well as could be expected given the
diverse populations. Strangely enough, for every student who said
something like "fewer assignments" or "too much chaos" you had a student
who said "I thought there should have been more little assignments." or
"I liked the chaos." With this kind of distribution of interest at
these population sizes, it will be impossible to make everyone happy.
At best, you'll be able to get the majority of them. At worst, you'll
only reach 10% of your population. In the worst case, you're mainly
hoping that your actively unhappy minority don't infect the rest of the
class.
-
Because the course is about teaching, the instructor(s) have very little
room for error. In a normal class, if the instructor delivers a bad
lecture the students would have to do more reading or research to reach
an understanding of a material. In a class on teaching, if the instructor
teaches poorly or makes an error (like getting an assignment back late),
it undermines the class's legitimacy as well as the confidence in the instructor.
-
Draconian approach to attendance and assessment contributed greatly to
student resentment and hostility. I'm not sure whether an instructor
with a PhD or an official instructor title could have muted this.
There is still a fundamental problem here that relates directly to the
external mission of this class. The Institute requires all departments
to show that they are training TAs. This course is not a high priority
in the lives of first year graduate students (and it shouldn't be).
It's also not of interest to roughly 60% of the attending body. At
the same time, it was my responsibility to certify these students as being
ready for the job of being a teaching assistant based on some criteria
set by Kurt Eiselt and myself over the years. It has been my experience
as a TA that not putting strong requirements on assignments means that
students will choose not to do the work. I strongly dislike the fact
that the only way to do this, as with most other mandatory training programs,
is to institute Draconian measures and certifying students who jumped through
a sufficient number of hoops. Personally, I'd prefer that students
just attended on their own but the Institute certification and the demands
of graduate school are in conflict here.
Back to Top
Suggestions for Future Design and Implementation
-
Future implementations of the course should remove the inherent schizophrenia
designed into the course of TA training and Future Instructor training.
In the original design, the class size was roughly 30+ students.
At that size, there was room for discussion and debate into how education
is conducted currently and some of them were very interesting. Many
of the sessions used this semester are simply holdovers from the first
year. But with the Institute now strongly enforcing the training
of TAs and with our growing graduate student populations, the class has
simply become to large to approach these courses in a thoughtful manner.
TA training for the pure purpose of preparing TAs to teach at the College
of Computing is necessarily an activity that has to address a large population
of students. Courses/seminars designed to prepare people for teaching
in the future (a fairly small number to begin with) have to be small with
a high amount of feedback - difficult to do for populations larger than
20 as I discovered when grading the 2 homeworks which did require a lot
of feedback. This leads to the next bullet point.
-
Break up the workshop back into those courses related directly to the training
of TAs for their jobs. While there are a set of secondary courses
that could be used to prepare those PhD students for academic life, they
may be best left to students who want to voluntarily take these courses
- possibly moved to CETL in conjunction with their new professor training
program. Alternatively, it would be good to split the workshop into
mandatory and optional lectures, requiring the students to attend N number
of optional lectures and to sign up for them at the beginning of class.
It would reduce the time burden on the students and, possibly, the effort
required on the part of the instructor.
-
Build the fun back into it somehow. This goal may not be possible
with the number of students that the course has to handle but teaching
is supposed to be a fun thing, not an onerous one. Games or skits
involving multiple TAs would be a good idea, provided you could find those
TAs to help out.
-
Many of the sessions would have benefitted from being more practice-based
or having more practical activities - grading real assignments for example.
As I recall, we had more opportunities to do this in the first implementation
but these have fallen steadily by the wayside with the size of the class.
-
Give the responsibility of the workshop back to a professor/TA pair or
possibly a committee. If the workshop is going to be TA-driven again,
I strongly recommend that the TA report to an active professor for mentoring
and feedback or that they split teaching and grading duties. 60+
people is too much work for one person.
-
If the College decides to strip TA training down to a few days, it should
discuss, at minimum, the following:
-
Basic responsibilities of TAs to the instructor, the class, and the students.
-
Professional responsibilities
-
How to handle the difficult issues of academic misconduct, disabled students,
conflicts with students, conflicts with the instructor, and so on.
-
College and Campus resources in support of TAing and the procedures for
accessing them.
-
A general review of grading heuristics and methods - especially for the
projects and programming assignments that are commonly seen at the College.
-
The College absolutely needs to supply TAs with a pamphlet of contact information
and basic duties and procedures. However, unless this pamphlet is
updated yearly with the latest information, it will fall into disuse.
There are documents embedded in the class web page which will help support
this endeavor.
Back to Top
References
These reference materials should be available from the College of Computing's
Student Services. They were purchased 5 years ago to help assist
in the planning and design of this course.
-
Davis, B., Tools for Teaching, Jossey-Bass, Inc.: San Francisco,
1993.
-
Eschenbach, B., Parrott, R. and Arneson, I., “Teaching in a Diverse Community:
Multicultural Awareness”, in Teaching Assistant Development Program,
College of Engineering, Cornell University, 1993.
-
Hewlett-Packard, Managing Diversity / Affirmative Action, Corporate
Training and Development
-
Streicher, R., Graduate Teaching Assistant Handbook, Center for Teaching
Development, University of California, San Diego, 1996.