What to look for in a student?
A successful undergraduate student researcher requires a combination of general and specific skills. Mentoring students from underrepresented groups can be a great way to help increase diversity in the economics profession, which I’m a firm believer is important to do [I’m still gathering my thoughts on how to do that well for a future post].
First, to evaluate general skills the best place to start is in the classroom. If you have not had the student asking your colleagues in the department is a good place to start. The student’s GPA is also reflective of these general skills, of course you will need to make your own adjustment based on the average GPA at your university and the courses taken. You also need to know more generally if the student can write. In section 4, I discuss having the student write a 1 page abstract, you can use this as a screening device.
In terms of specific skills you will want a student who likely has taken at least an intermediate level of statistics or econometrics in a class that covers regression analysis. Having a relevant field course should also be a prerequisite if the course is available at the university. Finally, the student will need to be able to do the analysis in a statistical software program such as STATA, SPSS or R. Ideally they know the program you use in your own research. If not someone who knows one language can quickly pickup another. I would advise extreme caution in working with a student who has only worked in Excel. You want a student who can at minimum open a dataset, run descriptive statistics, and run basic regression commands. If the student does not have these skills suggest they should go and learn them before you agree to work with them. Online tutorials for programs should be easy to find.
You may need to recruit students if you are interested in mentoring the best undergraduate researchers. The ideal student has a high GPA student, statistical software skills and is a good writer. Unfortunately, for professors these students also have other great opportunities including paid summer internships so there is a high opportunity cost to working with you. In order to recruit I would suggest you start by visiting the intermediate statistics/econometrics class offered by your department. Ask for 5 minutes at the beginning of class from your colleague if you do not teach this or there are other sections of this class. Sell the opportunity of research, I would suggest reading the benefits to students in the previous section. If you have current students doing projects have them join you as part of the presentation, students are more likely to be interested in hearing from other students.
In many intermediate statistics/econometrics classes students are asked to do a project with their own data. If you teach this course you may consider assigning a project and using the lessons of this guide to build a framework for the assignment. If you are not the teacher of this class ask your colleagues to tell you about the best papers from the previous semester to identify good potential students. You can then recruit these students individually through e-mail
You might also consider if classes in related applied fields such as Political Science, Geography, Sociology and Public Policy might be worth looking for recruits. I have found some of my best students from a Political Science class focusing on development. The ideal class would be aimed at Juniors and have a reputation for attracting the best students. Honors college classes are a good bet.
Section II Professor’s questions:
- Does the student have the prerequisites for independent research (high GPA, knowledge of software, related course work)?
- What is the student’s reputation in the classroom?
- Where are the best places to recruit students stat/metrics classes, colleagues, other courses?