Policy Project

Due Date:

Saturday, December 15, 2018

Goal of the Project:

The focus of these courses has been to help you strengthen your quantitative reasoning skills via exposing you to a wide variety of statistical analysis tools. Now is the time to put your enhanced quantitative reasoning skills to test. When you address real-life research questions through careful examination of a particular data set, you may encounter unexpected problems.Standard techniques you learned in class may or may not be directly applicable. The goal of the project is to further enhance your ability of quantitative reasoning by having you think about creative ways to extend and re-design standard methodologies, if necessary.

Picking Your Topic

The general guidelines are as follows:

Make sure you enjoy the topic
Have the project help you with your current job
If not, have it apply to your new job or new career
Analyze the topic with stats techniques we learned in class

Group Composition

Though I’m flexible with group size, each group should consist of 1-4 students. If you choose to form a group, try to form one with students who have diverse academic backgrounds and skill sets that can complement (not duplicate) yours.

Data

You are given a great amount of flexibility in choosing a data set that is most aligned with general interest of your group.Sources of data are at your job, on the web, or possibly from a publication.If you can’t find any data anywhere, I’ll be happy to give you some.

Proposal

It’s not necessary, but if you would like some feedback, prepare a proposal for your project.The proposal should include clearly specified research questions, a brief description of the data, and main statistical methodologies you plan to use to address research questions of your interest.If you would like, I’ll give feedback on your proposal.Or, you may want to try this to give structure for brainstorming your topic.

Presentation

Though it’s also not required, students find that a high-level presentation is effective in two ways:First, it helps define the scope of the project as the research is being conducted.Second, many students say that presenting this sort of “Executive Summary” to their colleagues, bosses and/or clients is quite effective for communicating advanced statistical methods.If you choose to prepare a presentation, you should focus on what questions you have been trying to answer and what you have learned through the whole process. A long discussion on details of your analysis should be avoided. Whenever possible, use graphical displays to highlight your key results rather than present complicated tables or computer outputs.The presentation should be 5-10 simple slides.

Final Report

If the final report looks like it will exceed 50 pages in total (at least 11pt and 1.5 line spacing), excluding references and appendices, you might want to ask me for guidance on how to trim it down. Figures or tables that carry substantial importance to your conclusion should be incorporated to the text. Others can be put into appendices along with a list of citations, if any are quoted.

Conclusion (There should be a section at the end with this title.)

Statistical Findings (There should be a short sub-section with this title.)

This is where you give your mention or echo your calculated statistic and corresponding p-value (if applicable).Alsomention if you accept/reject your null hypothesis.I also suggest mentioning the biz implications as well.

Scope of Inference (There should be a short sub-section with this title.)

Based on our discussions of random selection from the population and random assignment of units to groups, be very clear about the (appropriateness of) inference to the population and causal inference from your research.

Grading