Differentiated assignments

Has anyone had experience with differentiated assignments? That is, where you specifically give students more than one option for an assignment and different options might be worth different grades (e.g., you have to do option A for a chance to earn an A or you can choose option B but then the highest grade you can get is a B)? I'm thinking about doing this because I have a couple of assignments where there is a possibility for some students to do something that is really a level above other students but I don't want to tell students they have to do that. For example, in my Economics for Teachers class, one of the projects is to develop a "mini-lesson" using an example from pop culture (music, movies, TV, etc.). Ideally, students would come up with their own example but there are also many sources that provide examples for them (some of which we will be specifically discussing in class). So I'm considering telling students that in order to earn an A, they must come up with their own example; they have the option of using one of the examples from a source that already exists but then the maximum grade they can receive is a B+.

Similarly, in the final project for the data course, students must do a simple empirical analysis. I don't actually want to require that students do multiple regression; I would rather students focus on thinking through what analytic methods would be appropriate to the empirical question and discussing the relevant issues, not get bogged down (or freaked out) trying to figure out how to run a regression. But there are certainly some students who are capable of doing regressions and fully understanding what they are doing, and I want to make sure those students are challenged too. Since I will be giving students a set of questions to choose from for their projects (sadly, with 135 students total, I simply cannot give them complete free rein over their topics), I'm considering telling students that they can choose a question from list A (which will be questions that require regression to answer well) if they want to shoot for earning an A, or they can choose a question from list B, which will be somewhat simpler projects but then the maximum grade they can earn is a B+.

On the one hand, I feel like differentiated assignments are sort of 'giving in' and lowering expectations for a lot of students. On the other hand, it seems more realistic, and more fair, to specifically tell students "this is somewhat easier so it's not worth as much but it's up to you to make the choice." What do you think?

One down...

I survived the first week - 14 more to go! A brief summary of the week:

The good: I was reminded many times this week that the vast majority of my students truly are good kids who just want to do well and learn something. Sadly, I have a tendency to forget that because I usually hear a lot more from students who are having problems (and are trying to avoid responsibility). Many students had questions about their clickers, or how the class is going to work, but every single one was polite and they all approached me with an attitude of "am I understanding this right?" rather than "why are you making us do this?". And even though the air conditioning does not seem to be working correctly in any of the classrooms, students were awake and engaged enough to speak up. In other words, so far, so good...

The bad: I completely stressed myself out about the creation of the teams in my two sections of the data course. I had already planned to create the teams myself (rather than create the teams in class, as many TBLer's do), simply because with the size of the classes, doing it in class seemed physically really difficult. I sorted the students to get a mix of gender, class and laptop availability on each team (each team needed to have at least two people willing and able to bring a laptop to class, which I polled them about on the first day) and then I did go through to check that the non-native English speakers were also distributed equally across teams, and that there were no teams that might have cliques (e.g., there are a few members of the same fraternity in the class and two of them ended up on the same team so I moved one). I explained all that to the students but I still worry a bit about the lack of transparency. But the thing that really stressed me out was that one of the classes has relatively few women. There are 13 women in a class of 60 so with 10 teams of 6 students each, I could either spread the women out and have several teams with only one woman, or make some teams with no women at all. After a shout out for input from several female friends (I really love Facebook!), I decided to go ahead and spread the women out. We'll see how that goes...

The ugly: I'm happy to say that there wasn't anything truly ugly about the week (other than the threats made to some of my colleagues by a very disturbed former student). I was bracing for some serious issues with crashers but although there were more students trying to crash the class than I could possibly take, it wasn't as bad as I had expected. In the past, I have tried to avoid choosing among students by simply using a lottery to select who gets in, but given the state of things in the CSU, I decided that this time, I would give priority to econ majors (for whom the data course is required) and also favor those who had registration times after the class filled up (since they never even had the opportunity to enroll). I did get a few pleading emails which make me feel terrible, but I stuck to my guns.

One thing I realized is that I'm likely going to be somewhat stressed all semester simply because I don't like uncertainty and everything about this data class is uncertain. But the reaction from the students so far is encouraging...