Final Project Topics
Topic Selection Due by end of Monday, April 19
Email me with your top 3 (ranked) preferences for which of these projects you would like to undertake. Based as much as possible on your preferences, I will assign teams of 3-5 students to projects on Tuesday, April 20.
Each project is sketched below. Each has considerable latitude for variation depending on the interests of the students involved. More detail (and data, where appropriate) will be supplied to the groups by April 22 so that you can begin your analysis.
The outcome of your project should be a report describing your results (due at noon Monday, May 10) and a 30-minute presentation during the final-exam period on the afternoon of Thursday, May 13.
Project 1: Demand for automobiles in Ecuador
Use Ecuador's 2003 Census to examine the demand for automobilies by urban residents. The outcome of your study should help the government of Ecuador predict the number of cars on the road in 2015, based on your expected trends in population and income growth. The data for this project will require some "cleaning" and all documentation is in Spanish, so reading knowledge of Spanish will be helpful.
Project 2: Monte Carlo study of the DF-GLS and other unit-root tests
Stock and Watson claim that the power of the DF-GLS test is more powerful than the Dickey-Fuller and Phillips-Peron test. This project will use Monte Carlo simulations to assess that claim for varying values (less than one) of the autoregressive parameter and for varying sample lengths. No data are required for this project, just a good research design and a willingness to apply basic programming skills to constructing Stata do files.
Project 3: Assessing grade inflation for a sample of colleges
Gradeinflation.com tracks grade inflation at a sizable sample of U.S. colleges and universities, including Reed. We have obtained grade distribution data for their sample. This project would combine these grade inflation data with other data on college characteristics (selectivity, location, size, ranking, etc.) that are easily obtained from other sources to try to figure out what factors are associated with severe grade inflation.
Project 4: Choice of majors at Reed
It would be very useful for Reed academic departments and divisions to be able to anticipate the number of majors to be writing theses. This project uses Reed data to model the choice of major, based on admission characteristics such as SAT scores and on the courses taken and grades received in the first year. The data for this project are in Stata, but will require some manipulation to assemble into usable form.
Project 5: Returns to education for French and English speakers in Canada
This project uses the 2001 Canadian Census to examine whether the returns to education are different depending on native and learned languages, both in Quebec and in other provinces. The dataset is in a "text" format that can be easily input to Stata.
Project 6: Cointegration tests for purchasing-power parity
Cointegration analysis is a natural fit for analysis of long-run equilibria such as purchasing-power parity: the principle that exchange rates should change to offset differences in two countries' inflation rates. This project involves analysis for a group of countries (that you select) of whether the PPP relationship holds in the long run. The data are easily available from the International Financial Statistics that you used in Project #6.
Project 7: Which characteristics are most important in determining Portland house prices
This project uses a Stata dataset compiled by Noel Netusil to examine how various home and neighborhood characteristics affect home prices in Portland. There are many structural and environmental characteristics in the dataset that could be examined.
Project 8: Do it yourself!
If you have an idea for a project that you'd like to do, you can propose it and I'll do my best to try to make it doable.