As I have been developing this data course, I have been trying to keep in mind that this is an economics course, and the focus is supposed to be (as the title of the course states) the Collection and Use of Data in Economics. From a data standpoint, that means exposing students to the types of data that economists use (which is not quite the same thing as 'economic data' but that's a topic for another post). From an analysis standpoint, that means that we have to talk about inference from observational data, issues with identifying causation vs. just correlation, multiple regression, etc.
But I think it also means that we need to talk about the larger issue of what types of empirical questions do economists use data for? I'm planning to start the semester with a discussion of what is, and what isn't, an empirical question, and get the students thinking about what kinds of questions can and can't be answered with data. In economics, we talk about this in terms of positive and normative questions; most economists try to stay in the realm of positive analysis when we are doing data work but usually, our empirical analysis is still driven by normative questions (I usually refer to this as the 'why do we care' part of any research project). My plan is to have students first consider how values affect the trade-offs we are willing to make, and then we'll look at how data can inform those normative decisions. I think this will lead pretty naturally into a discussion of bias and how our normative values affect what data we are willing to believe, but I want students to think about how that goes both ways (i.e., prior beliefs may affect whether you accept some new piece of data but beliefs can also be affected by new data, and that's where a lot of social research comes in).
All of my students will have already taken two Principles courses in which they should have at least been exposed to the concepts of positive and normative analysis; however, most intro classes do little more than define positive as "what is" and normative as "what should be", which sort of drives me nuts because that seems so simplistic as to be pretty useless. I'm trying to find a simple paper or chapter that discusses these concepts more in the context of what economists do but haven't had any luck. If anyone out there knows of something that I could use, please let me know!
But I think it also means that we need to talk about the larger issue of what types of empirical questions do economists use data for? I'm planning to start the semester with a discussion of what is, and what isn't, an empirical question, and get the students thinking about what kinds of questions can and can't be answered with data. In economics, we talk about this in terms of positive and normative questions; most economists try to stay in the realm of positive analysis when we are doing data work but usually, our empirical analysis is still driven by normative questions (I usually refer to this as the 'why do we care' part of any research project). My plan is to have students first consider how values affect the trade-offs we are willing to make, and then we'll look at how data can inform those normative decisions. I think this will lead pretty naturally into a discussion of bias and how our normative values affect what data we are willing to believe, but I want students to think about how that goes both ways (i.e., prior beliefs may affect whether you accept some new piece of data but beliefs can also be affected by new data, and that's where a lot of social research comes in).
All of my students will have already taken two Principles courses in which they should have at least been exposed to the concepts of positive and normative analysis; however, most intro classes do little more than define positive as "what is" and normative as "what should be", which sort of drives me nuts because that seems so simplistic as to be pretty useless. I'm trying to find a simple paper or chapter that discusses these concepts more in the context of what economists do but haven't had any luck. If anyone out there knows of something that I could use, please let me know!
This comment has been removed by a blog administrator.
ReplyDelete