Statistics: Regression Analysis (Business-related analysis)
GSB 420: Business Analytics Tools Final Project Description Your final project, worth 25% of your grade, will be to perform your own regression analysis and present your findings in written form due on the last day of class (Nov. 26th at 11:55pm). Ultimately, what I would like to see in this paper is an estimation of the relationship between at least two variables. That is, to estimate how X influences Y. Don’t take this to mean you only have to find two variables. As we will see, omitting other variables can bias the estimated relationship between X and Y. You may use any data that you can collect or create, but the preferred course topics are business-related. Your project must test some theory which is clearly stated in your project. Ultimately the goal of applied business statistics is to support or refute hypotheses, thus your project will be graded on the clarity of your hypothesis and your use of analytical tools to test this hypothesis. Remember, there exist a number of possible outlets for a good research project—for instance there are a number of journals that specialize in student publications. Thus, think carefully about the data you choose to analyze. A quality research paper takes time and structure. The format that I look for in these papers is: 1. An introduction describing the question/issue you are addressing. 2. A discussion of the relevant literature that has been written on the ideas you are addressing. 3. A description of the theory you are testing as well as a description of how this theory fits in with the findings of the research described in part 2. 4. A description of your data (this includes a written description and a statistical-output description). This section should tie in with your theory; that is, you should answer why this data is appropriate to use for your given topic. 5. Your statistical findings. You should be sure to discuss your original findings as well as any corrections for violations of classical assumptions. 6. A conclusion that reviews your theory in light of your statistical findings. Good statistics writing is good writing in general. Thus be concerned with issues such as sentence structure and grammar. Further, most students writing their first statistics paper tend to focus on the technical aspects of the paper while paying short attention to the theories under consideration. Although this course is one that focuses on the technical aspects of estimating relationships, good statistics writing never stresses the technical aspect at the expense of the hypothesis being tested. Remembering who your audience is will always improve your paper. I expect that you include your regression findings in your paper and your original data and Excel files in your submission, as well. Your paper should be submitted electronically on D2L. You will be graded on the quality of your writing, the clarity of your hypothesis, the quality of your regressions, and your application of those regressions to the research question your paper addresses. Finally, some words of advice. Invariably I have students spend too much time looking for the “right” data set. Although the right data is important to long term research, finding a perfect data set rarely happens nor is it good policy to for students to spend precious time looking for data. Rather, I would choose your data by Week 5 and then apply what we learn throughout the course to your data. When it comes time to compile your results for your paper, you will find much of your work to be finished already.