Overview
This project is primarily an opportunity to apply and show what you've learned about data analysis, and statistics throughout the course. The project represents a large component of your grade.
The main goal is to analyze your own dataset. The analysis should include a descriptive summary of the data using appropriate charts and visual displays, a complete regression analysis and a final report.
Assessment Criteria
To aid in your planning, I share with you up front the criteria that I'll use to evaluate your final project report:
ANALYSIS: Are the chosen analyses appropriate for the variables/relationships under investigation, and are the assumptions underlying these analyses met? Are the analyses carried out correctly? Is there an effective mix of graphical, numerical, and inferential analyses? Did the student make appropriate conclusions from the analyses, and are these conclusions justified?
WRITING: How effectively does the report communicate the goals, procedures, and results of the study? Are the claims adequately supported? How well is the report structured and organized? Are text and analyses effectively interwoven? Does the writing style enhance what the student is trying to communicate? How well edited is the report?
1. Descriptive analysis
The analysis should include a summary of the data using appropriate charts and visual displays as well as descriptive statistics. Specifically, you should address the statistical differences among your variables. Critique your findings.
2. Regression Analysis
Play around with building various explanatory models for your dependent variable. Be sure to consult the correlation matrix when making a choice of variables to use, interpret all slopes/intercepts, evaluate the fit of your models, check whether model assumptions are verified, and, of course, use your model to make forecasts and give some idea of your confidence in your forecasts.
3. Overall Analysis
Critique the methods used in this project and offer suggestions for improvement or changes. Would you consider using those methods as a business person faced with uncertainty and needing to make decisions?
4. Report
First of all, you will submit a single report, no more than 10 pages long, and aimed at a general business audience. You may assume that your audience needs to make some decisions and is seeking a statistical perspective relatively free of technical jargon.
The document should be written in clear, concise, correct English. Just as with any formal writing assignment, mechanical mistakes and bad stylistic habits distract the reader from the points you're trying to make.
How much output should you include in your report? Where should it go? Good questions. Here's my best general advice: Focus on your own discussion and interpretation, and use the software's plots and calculations primarily to back up your own claims and analysis. So at the very most, include output only if you discuss and analyze it in your text. Avoid abusing important jargon which has very precise technical meanings.
Where should you put the software's results? Optimally you should import the relevant output into the appropriate spot of your document, just before or just after the discussion. Alternatively you may choose to put results in an Appendix. Regardless of where you include output, be sure to trim away all irrelevant detail, and include only what your reader really needs to see. Please remember to label all aspects of the plots appropriately.