SESSION 5 TEACHING Joan Francioni, University of Southwestern Louisiana Virginia Lo, University of Oregon Barbara Ryder, Rutgers University Hi. I'm Joan Francioni and along with Ginnie Lo and Barbara Ryder I'm going to talk about the teaching component of academics. You've heard a lot this morning about things that you have to do, research-wise and service-wise in an academic position, but that fact that you're in an academic position distinguishes you from a purely research position in that there is this teaching component. Let me start off by talking about what the benefits of teaching are. Independent of whatever benefits there are for you, the teacher, the primary reason that you are teaching is to educate the students. It's important to keep in mind that there is this serious purpose for your teaching. For you yourself, there are a number of benefits, but I'm going to talk about three right here. One is just the pleasure and satisfaction you get out of teaching. It can be quite a rewarding experience to take people who don't understand a concept and teach them and move a whole class forward. In addition to that, you will undoubtedly have a much deeper understanding of anything you have to teach. The deeper understanding of the material comes via the course prep that you have to do and also from interacting with students, as they will constantly be asking you questions and challenging things that you've said. The third benefit to teaching, which is relevant to what was talked about earlier today about being connected in the research community, is that you develop your own communication skills. You get a lot of practice talking to people and explaining things in a clear and organized way. (You also get a very good feel for what fifty minutes is, which is I think why some of us are running over our time slots today.) --Laughter As with most things in life, everything comes at a cost. Your real cost for teaching is the enormous amount of time and energy it takes. And you can put as much time and energy as you possibly have into teaching and you will still feel as if you could put more. You've heard earlier that you will have this same conflict and pull on your time from the research side, so I remind you that you must find a balance that works for you. You can minimize the cost and maximize the benefits of teaching by teaching courses of interest to you -- ones that will motivate you to put in the time and energy that you need and that will reward you with a deeper understanding of the subject. But I think it's really important to take note of the following: that teaching is usually not rewarded externally, commensurate with the effort that it requires. Therefore, if you can not find some internal satisfaction from teaching then you really shouldn't be in this type of job, okay? For the rest of my talk I'm going to discuss how students learn so that you can understand where they're coming from, and I'll talk about a few specific teaching techniques that you can use to help. Following that, Ginnie's going to talk about grading and test assignments and Barbara's going to talk about some of the special problems that come up. In general, there have been a number of psychological and educational studies done on how people learn. In particular, it is believed that students view the world based on certain general structures, i.e., there is a framework that the student tries to fit everything into. In addition, these structures change over time in a fairly well understood pattern. Where your students are and what structure they're dealing with at that point in their lives will affect how they interpret and how they perceive things that you tell them. Also, students won't always move willingly from one structure to the next. Thus it's important to know where they are and to always be pushing them to move forward. [The following structures are adapted from the book: Forms of Intellectual and Ethical Development in the College Years by William G. Perry; Holt, Rinehart and Winston, Inc; 1968.] The first structure is when people see the world in a binary type of mode. Things are either right and therefore good or they are wrong and therefore bad. Students in this structure feel that as the teacher, you're the one who knows the right answers so your job is to teach those answers to the students and then it's the students' job to remember them. As an example, you may be giving a lecture where you're going off on an interesting tangent to motivate the class and you see a hand go up. You think -- "Great, they're interested in this." -- but the question is "Is this going to be on the test?" Now, Barbara says that she always answers "Yes." -- Laughter. Nonetheless, the question can be demoralizing because it seems the students are not really interested in what you're talking about. But if you can remember that the reason you're getting that question is because the student is working within this first structure, then it helps you to deal with the question and the student more effectively. This, by the way, is typical for your first-year students. In their second year of college, students usually move into the next structure: perceiving diversity of opinion. As an example, a student in your class hears you say one thing on a certain topic when a teacher in a previous class said something different. The students start to catch on that different people tell them different things but they don't have the depth of understanding yet to realize that things are different in different contexts. Thus they usually attribute hearing the different opinions to either (1) the teacher is unqualified and doesn't know what's going on, whereby they become antagonistic with you, or (2) they think "Oh, I get it. They're just testing us. This is an exercise and we have to find out the right answer." Notice that they will assume there is "the" right answer. There aren't two. It is just their job to figure out what "it" is. When you know that this is how students are -- where they are, where they're coming from -- it can help you understand their reaction to your methods. The third structure is where students start accepting that all this diversity of opinion and general uncertainty is legitimate; it actually exists and will continue to exist. As an example of the students' behavior from within this structure, consider the test question: "Which is the best sorting algorithm to use in the following situation?" In class, you had covered a few sorts and talked about how they work. You want the students to look at the pros and cons of the different sorting algorithms in relation to how each would fit with the example scenario you've posed and to make a decision. But, that's not what they're likely to do while in this structure. They do know they're supposed to justify their answer because they're going to be graded on "good expression," but they don't really understand what the grading standards are. So they try to write everything they can think of about sorts. (And if you only leave them a small amount of space for their answer, they'll just write smaller. --Laughter) A better strategy is to try and structure your questions, as well as your general dealings with the students, to try and force them to address the differences. Eventually students will move into this fourth structure where they start perceiving all knowledge and values (including the authority's) as contextual and relativistic. For many people, this is when college life really becomes exciting -- when you start understanding that everybody has these different opinions that are based on where they're coming from. In particular, this is when your students finally start understanding, not just that they're supposed to say "It depends," but why that is often the correct answer. Given that background on where students are coming from, I'll now talk a little bit about learning strategies that people use. A natural tendency in teaching is to try and emulate teachers that you thought were really good. One reason you felt they were so good is because the style that they used in teaching worked well for you to learn things. What you have to recognize as a teacher is that students will have a variety of learning strategies and they will not all learn best in the same way as you do. I remember as a student I never liked to read the book that much. I preferred to get the information out of the lecture. Therefore when I first started teaching I never gave reading assignments. I felt students should come to class and follow the lectures since that was the best way for them to learn. If a student asked about what part the book we were on, I would tell them to look in the index for some of the key words I was using and read those sections if they wanted to! ---Laughter-- Later on I found out that certain people have a more visual orientation and learn better if they can read about a topic before or after hearing about it in a lecture, whereas others learn better by hearing it presented. As other examples, some people like a more competitive class situation than others; some students do well in groups projects and some don't; etc. The key thing here is to recognize that students in your classes will have different learning strategies and therefore, it is most effective if you can provide a variety of learning situations for them. Try and remember different methods and styles used in classes you took as a student; things that didn't work great for you but might work great for somebody else. Lets look now at some general teaching strategies and then I will go into more specifics on course content. Again, this is based on basic theory of pedagogy. Most people know, I think, that you learn by doing and therefore it's much better if your students are active listeners and viewers then if everybody just sits there and takes notes. But even though we know this is true in theory, that doesn't mean you can always implement these ideas in the kind of class situation that you end up being in. Nonetheless, you can do things to encourage students to be active participants on some level. For example, in situations where you don't have enough time to have a total interactive class you can still ask rhetorical questions and let the students think of an answer before you go ahead and give it. Another teaching strategy is to include repetition of concepts but via different experiences. If students don't understand something, just saying it to them the same way over and over won't help. You have to come at it from different ways. Concepts should also be reinforced, and positive reinforcement is better than negative reinforcement. There is a tendency in certain situations to give negative reinforcement -- in particular in grading. For example, a 10-point test question would usually be graded as -3 rather than +7. This is because most people find it easier to think in these terms. But you can compensate for this by putting a few check marks or "good" remarks here and there when appropriate. It can be very effective in teaching to provide opportunities for students to generalize and discriminate among concepts -- to take things that are in different situations and find what's in common and what's different among them. In addition, you want to motivate your students to be active learners. Mary Vernon made a statement this morning that in your own research you should work on problems that you think are important. She said that so that you will motivate yourself, and that exact same comment goes for your students. If you can convince them that they are looking for answers to questions that they think are important, then they will. No matter how good a lecturer you are, of course, if the student does not want to learn, the student won't learn. Thus, you must also try to provide different conditions in which they can become motivated. Sometimes a student is motivated but is also dealing with conflicts and frustrations in understanding the material. Try to recognize this and provide avenues for them to get around it. Finally, a very important teaching strategy is to provide an atmosphere of cooperation between you and the students. If you have the attitude that you're doing your job of preparing the lectures and it's up the students to learn the material if they want to or not, then it makes for a tense class. Rather it is more conducive for the students to learn when they have the feeling like you're interested in them learning the material. More like the tests are there for you to know whether the class is catching on or not, instead of as a way to punish the students who haven't caught on yet. You can't always be "on their side" so to speak, but you sometimes can and you should make an effort to do so when it is feasible. All right, so let's look at a few specifics on general course content. For one, make the course objectives known to your students up front. It is also useful to reiterate them a few times during the semester. You are trying to motivate the students and you are trying to let them know that there are important reasons for studying this material besides just "I need to take this course because it is a required course and I need to pass it to get my degree so I can finally get out and start making some money." --Laughter-- That shouldn't be their only motivation. For example, in operating systems your course objectives might include: to understand concurrency in parallel programming. You could then discuss why parallel programming is important in scientific applications, let's say. This helps provide different motivations for students instead of just expecting the students to want to learn because "operating systems stuff is really cool and you should like it just for it's own sake." Another specific suggestion is to use multiple sources for your lecture notes. But here you have a couple of conflicting objectives. You want to provide the student with different ways of looking at things but you must also remember that they don't know the subject ahead of time. So if your text book uses certain definitions and notations, and then you use different definitions or notations in class from another book that you preped from, that's going to totally confuse them. They don't really understand it enough to see how the two different notations really mean the same thing and to extrapolate the essence out of that. The goal is to give them different examples than the text book has and present the material in different ways but still be consistent with the text book. It can be very helpful to the students if you can connect topics in one class to other C.S. classes. In particular, this helps students to generalize certain concepts. So, for example, if you are teaching concurrency in operation systems, you can also discuss how that same issue comes up in database systems. It is also useful to incorporate references and experiences outside of the department. Making analogies to real world concepts is one way, like talking about lines at the bank or the grocery store when you're lecturing about queueing theory. Another is to discuss applications of computer science topics. I have found that many undergraduates, in particular, get started in the program and just keep going from course to course without really knowing much about what you can do as a computer scientist. Sometimes they don't catch on until they are seniors and start thinking about what jobs they can get. It helps if, in your classes, you take a little time now and then to tell them about different applications of what you're covering. This also helps them feel the excitement of the field and can help reinforce certain issues. Another strategy here is to arrange for guest lectures in other departments if possible. This tends to give students a very different perspective on the topic and again can tie in some things that they might be learning in another class in another department. Some examples would be to invite a psychologist into an A.I. class, a linguist into a Programming Languages class, or an economist into a Software Engineering class. On the first example, one of the applicants to this workshop mentioned a situation where a psychologist was invited into an A.I. class. The psychologist was doing work on neurobiology and so brought in this bucket and pulled out a human brain. --Laughter-- A concrete thing that you can do to really help your students is to make effective use of the board. Again I would stress that you should remember that the students don't understand the topic when you are first trying to teach it to them. So when you are going along explaining and you just kind of scribble little things on the board -- that's fine if they already know it and the board notes are just to jog their memory. But when these students go back and look at their notes to try and understand something that they don't know to begin with, such notes don't help them at all. A few simple hints are to try and write legibly; to always put key concepts on the board; to make an effort to be somewhat structured as you are moving across the board and avoid always working on one section and constantly erasing just that part of the board; and use correct spelling -- if you can't think of how to spell a word, pick another one. --Laughter-- And lastly, try to be an as organized as possible. A little unorganization can really hurt. Say you are rushed and you you don't have time to make a up a thorough assignment, so you just kind of throw something together. This will usually come back to haunt you, especially if you have large numbers of students because they will all be at your door or sending you email trying to clarify what is going on. And you will end up spending much more time "fixing" the assignment than it would have taken to set it up right in the beginning. On another topic, let's talk about how you can meet the needs of so many different kinds of students. Students come in all shapes and sizes -- from the at-risk to the exceptional, male and female, multicultural. And, in particular, as you finish graduate school your peers are a very select group of computer scientists. So then you go in and start teaching undergraduate students and you think to yourself "Why are they so stupid?" --Laughter--- They aren't really stupid. They just aren't Ph.D.' students and they don't know all this already and they are working in these different structures so that they can't understand everything you are saying for what it really is. You have to work on trying to meet all of these different needs because your job as an educator is not just to pull the exceptional students through the program. It's not always easy to do, mind you, but you should at least recognize that and be prepared to work on it. I can offer a few specific suggestions that might help you along the way. The first one is to distinguish what you can do from what is fair to do to help out students. Usually enough time and effort will help out most students, but is that really fair to you, the student, and the rest of the class? You need to consider this question so that you know how far you should go to help someone out. It usually helps you to know how well you're getting across to the students when you can get feedback from the class as a whole. In some classes you'll have the brightest students answering as well as asking all the questions; in others it will be the slower students giving most of the feedback. But you need feedback from the whole group so as to be able to judge where they are and to adjust how quickly you are moving or not. Find out what's going on in your own university about tutoring options available to students. Sometimes there are classes for English and writing and if you have students who write poorly then you can send them off to a class like that. Keeping regular office hours -- and being in during your office hours -- helps out a lot. This helps the students to know when they can get help, and it also allows you to not be on call 24 hours a day. Try to be approachable to your students. Research on sex bias in teaching evaluations indicates that just smiling now and then makes students feel much more comfortable with a teacher. Giving a warm smile now and then makes students feel like you are cooperating with them. On the other hand, although you might think a sarcastic comment is funny, it is almost always taken the wrong way, at least by somebody in the class. And you are going to end up hurting somebody's feelings. And last, but not least, always try to maintain your professionalism and grading standards. We're going to switch gears now and Ginnie is going to talk about making up tests and assignments and the general grading aspects of teaching. Hi, I'm Ginnie Lo from the University of Oregon and I got my PhD from the University of Illinois. My area of research is parallel and distributed computing and I just got tenure at the University of Oregon of which I am very happy about. --Applause. I did it in a slightly nontraditional path because of strong commitments to family, both my kids and my parents who were elderly, and if someone would like to hear more about taking "another" route, I can talk to them later on. I will say it took longer than usual but it worked out very well for me. Right now I'm going to talk about assignments and tests and give some suggestions for writing good homework assignments and good tests. This is actually something that can be quite enjoyable and that takes a lot of creativity. It's something I enjoy very much as part of my teaching responsibilities. I also want to give you some hints about how to do this efficiently without too much of a time commitment because for most of you you will be teaching in the context of a university whose criteria for tenure is mostly research productivity. And then I want to give you some general suggestions for managing time with respect to teaching, not just for working on homeworks and tests. My first slide is a little bit more oriented towards writing a good homework assignment problem or a good exam question. Two suggestions I have are that (1) you want to promote cooperative problem solving among your students as well as individual problem solving. So my homework assignments always have a mixture of group problems and individual problems. This is not just programming assignments -- it's even written assignments. For the group assignments I help form the groups or I let them form their own and then help adjust them a little bit. They work on the problem together with open discussion and they turn in one solution and each individual member of the group gets the same grade. For the group grade on that problem I don't try to distinguish who gets what; I just take the group product and grade it. I've gotten lots of feedback from students who have said they really enjoyed those group problems. They tend to be more open ended types of problems with some creativity or with some debatable points in the problems so that they have a chance to interact and discuss their viewpoints. That's one guideline and then the other one is that when you write questions, especially on an exam, write a variety of types of questions. You want to write some questions that are computational and some that are essay because I think communication skills are extremely important and they have sort of been undervalued in our field. You may want to write a variety of questions as far as how they test the information. Basic questions are just kind of drill type questions where the answer comes back in a very objective form -- sort of a straight forward "spit back what you have learned" type of question. I try to write a lot of questions that integrate so that it's not a form of a question they will have seen in homework. Rather I try to take two things they have learned and put them together to see if they can assimilate what they have learned and apply it to something else. And I write open ended questions and those are usually on a take home exam so that they can show me how they can take the material beyond just the minimum requirements of the class. I like to think of grading such that if someone gets a C in my class it's not because they sort of have a half foggy notion of what the material was, but that they have attained a basic level of competence. Of course, that's what our definition of A, B, and C is, but when it comes out in the numbers sometimes getting a 70 in the class means someone just two-thirds understood everything. I want them to have a real competent understanding at some level of a certain subset of the course material. I also usually have my exams include both in-class problems and take-home problems, and the take-home problems are, of course, all individual on the exams. I do that because I want to optimize their chance to perform well and the time pressure of a 50 minute class is just sometimes not amicable to doing a good job. It certainly is not a good environment for doing an open ended question or a longer type of question. So here is just one example of a question I made up for my operating systems class. I asked "Suppose you are to design an operating system for a computer that has no clock and no interrupt hardware. Discuss how your operating system could support the services listed below." When I first used this question, that was the whole question on the exam [without enumerating specific services] and I found that with such an open ended question I got so many answers going off in so many directions that it was very hard to grade. I ended up refining it with guidelines later and that's going to be one of my hints for managing your time in writing and grading exams: Focus it so that you get an answer going in the direction that you want it to go. So these are some suggestions for writing exams that are easier to grade and that focus the student's attention. The other thing is to write questions that are not too hard to answer nor too hard to grade. I always tell students to write down their assumptions if there is something that they don't understand on an exam question. I often provide a templet for the problem. Here is a question that was on one of my early exams when I first started teaching as a new assistant professor. I wrote "Define internal and external fragmentations that can occur due to memory management schemes." Again I got answers going all over the place -- some people left out some of the memory management schemes all together and others addressed all of them, so it was hard to grade uniformly. Later on I made the question have an answer format with boxes that guided them to what I wanted to come out. Also with these boxes I could really grade the answers quickly. Now you do loose something when you structure the questions that tightly. The students don't have to organize their answer and decide what are the important issues in the answer. This, then goes back to writing exams with a variety of questions. You want some of the questions designed for the ease of grading but you still want to leave some where you are telling them that their structuring of the answer is part of a good answer. I have found that as I grade exams, I learn how to improve the questions. I recycle my exam questions to homework questions in following years so I find ways to improve on the exams. I give the same homework assignment again and I refine it each year. Of course, it's very useful if you can write out all the solutions yourself and then have the teaching assistant take the exam and debug it. Here are some general guidelines for saving time. When I first started teaching I contacted a number of faculty at other universities who had taught the same courses and asked for their resources. I was able to utilize a lot of those and they were very helpful. I recycle old questions and reuse them year after year. You need to try to cope with the time demands of teaching. Lecture preparations are very time consuming so you just need to cut it at some point. One way to deal with that is to ask to teach the same courses in successive years -- that's very very helpful and, more than just the time savings, you also you end up learning the course material very well. Ask to teach courses that support your research. Don't try to be too ambitious. People who are learning can't learn too much material at once so it turns out you can actually go at a slower pace then you initially think you can. It helps both you and the student to not cover too much. Utilize the resources of your teaching assistant. Sometime you feel like you ought to do it all. For grading, I try to give some of it to my assistants. But it's also important to try and grade some of the questions so that you know how the students are doing. If necessary, be aggressive -- if the department isn't giving you enough resources, go ask them for another grader or whatever. I guess the last hint I can give is to learn to let go. You want to do an excellent job but sometimes the time doesn't permit it. You don't have to be perfect. And even though in grading I want to do it really thoroughly for all the students, I sometime just grade a little bit more superficially and accept it. I'm Barbara Ryder and I teach in the Computer Science department at Rutgers University which is the state university of New Jersey. By way of introduction, I finished a math degree in '71 and went into industrial work at Bell Laboratories and subsequently after having two children went back to graduate school and graduated in '82. I have been an academic at Rutgers since then and I am now an associate professor there. We are trying to concentrate on issues of teaching in undergraduate courses. My job is to try to tell you about teaching in the situation that many of us at large public universities find ourselves. This is teaching courses of between 75 and 100 people in a big lecture for three hours a week where the students break up into smaller recitation sections of 30 some odd people that are taught by graduate students. I think when you have a large lecture class the most important thing is to make clear to the students at the beginning what their obligations are and what your expectations are. The first day they come to class there should be a syllabus with a tentative outline of the sequence of topics you are going to cover: how many exams you are going to have; when they are; any programming assignments that are multiweek assignments. This way the students know what their obligation in this whole situation is. At many schools there is sort of a "shopping period" in the first week of class and your syllabus helps the students to decide how they are going to balance this class and its obligations verses other classes that they are taking. It's also important, as Joan pointed out, not to always rely only on the text book for materials. It's good to start the semester out on the right foot by having your extra reserve materials in the library ready for the beginning of class. Also, students can go and look at the material and get an idea of the kinds of topics that you are going to cover. Now in a large class, no matter how clear and succinct you are at the board, if you are trying to write down a program of more than four lines, you probably need to have a hand out. It's very very easy for people to get confused and have difficulty with examples that are multiboard examples, particularly in a large lecture class. My approach has been to develop a suite of examples. You don't walk in Day 1 and have this but as you begin teaching a course and you ask colleagues for their suggestions when they teach similar courses, you can develop your own suite of examples. This next point is a little controversial: Encourage effective class participation. My colleagues and I don't always agree on this one. I think you can teach a large lecture class and also allow and encourage questions from the students. In general having questions from the students is the only way you can tell that you are getting your points across. You will have eye contact with only the first three rows in the class. That means the vast majority of the students are out there and you can't really tell whether the points you are trying to make are being understood. (And generally the same kids sit in the first three rows.) There are several ways to try to set up a classroom atmosphere that encourages questions. For one, you have to be patient enough to wait for the students to get up the courage to shout out an answer. This gets better later on in the term after the students are used to you, but they tend to be very shy initially. When you do get an answer, it's very important to avoid replies that make students look unknowledgeable or foolish. This is what students worry about in these large classes. They're willing to go on the edge if there's in a group of 10 doing homework, but not when there are a hundred people. Similarly, when you get a question from the floor, try not to bite off the student's head if the question is inappropriate. If you can manage to make your class nonconfrontational, you will get participation. And finally, do motivating examples. I teach by example and, in general, I think most of us do. It's important not to just lecture about ideas but to show people ideas in the context of some kind of example. Two things that I put under logistics but that could have been under problems just as well are grading and project courses. Assigning grades is really very difficult to do. It's difficult to do because we know it means a lot to the students, it's difficult to be fair, and it's difficult in the beginning to know how to be consistent when you haven't done it before. What you have to try to do is set some standards for yourself and try to keep to them. The most important standard is defining what is a passing grade in your class. You will have to decide at the end of the term, when you've totaled up all the tests and the projects and the homeworks and everything else, what does it take to pass the class. That's really the most important decision; not who gets the A's and who gets the B's. It matters to people's cumulative average, but the important decision is who passes the class. And you should ONLY consider a student's performance in the one particular class, not whether or not getting a D or an F in this class will get a student kicked out of school or cause the loss of a scholarship or anything else. Many of us have rules of thumb that we use, but this is not something that people take lightly so be ready. Project courses. There are a lot of aspects of courses that deal with computer projects that make them somewhat different from a pure lecture class. Let me just highlight one of these. The important thing in a project class is not to give a monolithic project to a group of undergraduates and tell them it is due at the end of the term. That is death. Uniformly, you will have people who will never finish. Uniformly, you will have people who don't even know how to get started. So in a project course, what you really have to do is break the project into logical pieces and give people some guidance as to how to start designing, particularly, the beginning pieces. You will find that when you give realistic timetables, you can do some kind of coordination with what's expected of the students in their other courses. Projects are also a good mechanism for students to work on their communication skills. Having them document their projects or give oral presentations, aids them in terms of developing skills that industry looks for when they get out. Let's spend a little time on problems that you may have to deal with. One is handling questions in class. I guess the thing that's most important in handling questions in front of any size class is knowing that saying, "I don't know." is not a cop-out. The important thing that you learn after you get over the fear of facing a hundred bodies and trying to present some material and to interest them in the material, is that the students will ask questions that you may not have an on-the-fly answer for. The appropriate response is to say "I'm not sure." or "I don't know." and immediately to come back in the next lecture with an answer. So in that way it's not like you're showing vulnerability; you don't get negative points for being unprepared. There are questions that come out of left field that you can't always handle. Also, make sure you understand the question and make sure all the people in the class understand the question before you start answering it. Another problem is handling your own mistakes. Just as in the other situation, I think quick feedback is the key to handling something you have said that is not true or was said by mistake. Many of us make little glitches in our teaching and the important thing is to be honest with people because they know you've made the mistake. You're not fooling anybody and so the best recovery is to be honest about it. Let me talk a little on cheating because this is really a large problem in many of our lives. Cheating exists. I don't think there's any way to get rid of it but there are, shall we say, strategies to try to control the opportunity students have for cheating. These are some of the things that I do. I give all open book exams. It's true I teach upper division courses, juniors and seniors, but I give open book exams. Well that cuts out cheat sheets and all that kind of stuff and it also, unfortunately, puts a burden on me to make harder questions because you can't give regurgitation questions on open book exams. So you have to come up with these integrative questions that Ginnie talked about that make people think and that are tough enough. Cheating on homeworks? Well one thing you can do is to always model one or two of the problems on your exams from the homeworks. Then you know the people that didn't really do their homework, no matter what they got as a homework grade. You make it hurt them later as it were; crime doesn't pay is kind of my philosophy on this. On programming assignments, this is much harder. In the days of text processors it's fairly easy to change variables and to change indentation. When you or a TA is looking at a programming assignment and there's a very bizarre programming error that comes up in two assignments out of the thirty or forty, it certainly becomes obvious that in the least, these people talked about the design of their program. Your recourse in such cases should be decided ahead of time and told to the students up front. What you decide to do finally should be based on your personal feelings as well as any departmental or university policies. We'll take some questions now and then add a couple of wrap-up comments at the end. This is on cheating and is a slightly long-winded question, but one thing I like to do is give solutions. And in context let me say I'm working at a sophomore level; algorithms and data structures. I like to give out solutions and I have a lot of students who really say this helps. And I give out solutions when the homeworks are due. I don't take late assignments. That makes it easy because the solutions are right there. I also like to reuse questions. I've made some questions that I like and so I reuse these questions. The answers are noticeable better the second time around and it's clear the students got the solutions from someone else. Related to that, I ask students not to collaborate on homeworks but they do anyhow. At times it's obvious. Yes, I do sort of check them on the exams but I have this problem -- if you weight the homeworks enough, they spend a lot of time on it but then if you weight them too much then the cheaters get by. If you weight them too little the people who work hard and are really bad exam takers have trouble How does one combine all these things? I don't think there's an easy answer to the problem of cheating. And I also don't think that as a university professor I should be a policewoman, frankly. I don't want to make it easy for people to get away with not doing the work by themselves and I don't want people to steal other peoples' answers. But I'm only willing to go so far in the pursuit of crime at the university. You've raised some reasonable issues. You might want to have some discussions with the other people in your department. Sometimes the department will have a general philosophy about what is considered illegal collaboration and what is not. In our department we have a strong policy that there's to be no collaboration tolerated. If anybody gets caught then they flunk that class. A notice is written out. They aren't expelled or anything from school but a notice is sent all the way up to the Dean of Students. We do that because we want to see that if they're cheating in other classes too it gets heard. But all schools don't do that and sometimes when you send a cheating case up to the Dean of Students and then you go into a disciplinary committee hearing, you always lose the case because people in other disciplines don't understand what is cheating on an assignment. Talk to some other people in your department and at least try and get some consistency. It might be something that could be brought up in curriculum committee discussions so that there can be some consistency. As for reusing questions, students definitely will network and get ahold of previous semester tests and assignments. And when students are under a lot of pressure, which they are most of the time, it is very hard for them not to make use of that information. Well this isn't a question. This is an advertisement for a project that CRA has just started. We're working on putting together a collection of Master Teacher tapes which are to be used by first and second year assistant professors just getting started. There was an advertisement that went around to your department heads for the four Master Teacher tapes. This is a project that was started by Dave Patterson at Berkeley. I would highly recommend that you go and see if your department head has ordered them. They're $300 dollars a class and you get four to five tapes. They're not to be used in place of you in class. What they're to be used for is for you to take them home and see how a master teacher presents that material: the pace that is used; the style used. The four tapes are Manny Blum in algorithms, junior, senior level; John Ousterhout on operating systems; Dave Patterson in architecture; and Randy Katz in logic design, a sophomore level course. We're hoping to add to this set and we're hoping that you can use it to spend less time on preparation and still come out with nice teaching evaluations. I'd like some suggestions on how to teach courses in which there's a broad range of abilities between the students. How do you challenge the best students in the course and how much time and effort do you spend on the students at the bottom? I try to accommodate all of them. I have extra credit questions for the more able students and then I do help sessions for students who are struggling. I actually have make-up mid-terms to give people a second chance. If they have flubbed it, they come to a help session and then they get a chance to do it again. But, that's all extra time. Again the reward is internal. You don't get credit for it; it takes your time but I find it quite rewarding even to work with the slower students; to have them actually come up to speed and maybe turn out not to be a slow student after all. You have a fine line here too because some students are not adequately prepared to be in this field. You also need to have some awareness of what the standards are in your department as to what passing is so that you aren't always just doing a bell curve on whatever distribution you get in your class. Office hours can also be used effectively to help out students. It's relatively important to be there for all the students in the class so that the people who feel they do need extra help can avail themselves of basically one-on-one tutoring during your office hours. Another good teaching tip is to run review sessions before the hourlys and before the final. Not in class time but in extra time because you can't afford to give up the class time. You usually have too much to get into the lectures. But these things really help the students organize their studying and figure out a structure to outline what it is you've tried to give them. It also gives them some insight into the kind of problems you're thinking about and what might be on the test. This really helps those students whose skills at taking classes and absorbing the material are not as strong. I just finished my first term at teaching and I found it a rather grueling experience. I learned a lot and the next time I teach it will be much better, I'm sure. But a couple of points of advice for new teachers that I'd recommend is that they try to get a smaller size course with fewer students, and also try not to teach a required course so that the course is not filled with students who don't want to take that course. I also have a question that hasn't really been addressed. I taught two sections of the same course and one of those classes became quite antagonistic and obnoxious. There was a lot of background noise in the back of the room and it wasn't until later when individual students approached me and apologized for the class in general, that I found that it was really more against me than anything else. What I didn't learn is how to control a mass movement like that at an early enough stage to prevent things from getting out of hand. I think I could probably recognize it but what should I do about it? There's no easy answer to that -- different things work for different people and sometimes it seems you just can't connect with a class no matter what you do. I think, however, that the earlier you are able to recognize that things are going in a weird direction, the easier it is to take care of. As you said, the ability to recognize the situation comes with experience. One thing that you can try if this happens again is when you think someone or some group is talking too much in class you just stop and you say "Is there something you wanted to say to the class?" You don't have to be ugly about it but you can still call their attention, in a nonconfrontational way, to let them know: "I know you're doing that and we're going to stop if that's what you're going to do. I'm not just going to keep talking louder and louder to try and talk over you." In general, students will get upset when they aren't getting their money's worth. They might not say that to you and they may even be laughing at the one who's being disruptive, but they really don't appreciate that. When you are able to get your point across to the disruptive ones, again in a nonconfrontational way, then the others will often put on some peer pressure if it happens again. One more thing. As a teacher you try to be friendly with your students and show them that you're on their side, but at the same time you must be their judge and evaluator. Being too much on either side can have bad repercussions. Let me make a few closing remarks. We want to stress to you that you should make an effort to be considerate of your students and also true to yourself. Maintain a reasonable pace for yourself and your students. Recognize that you both have lives outside of school. Capitalize on your strengths and adjust for your weaknesses. Everybody has both and you can really find out a lot about what's working, as well as what needs to be worked on, by reading the student comments on your evaluations. And finally, treat your students with the respect that you want them to treat you with. If you don't then you will always have conflicts. But if you do -- it just goes a long way when you let them know that you think that they're important as well. It's a great profession, and I wish you all the best!