STA 4164 - Statistical Method III Project 2– Guidelines

STA 4164 - Statistical Method III
Project 2 – Guidelines
Summer 2022
Please READ this document CAREFULLY before starting your projectGENERAL GUIDELINES:
1. This assignment is a group assignment. You may use notes from our class and online
resources, such as your preferred programming language documentation and help files. Youmay also discuss your approach towards the problem with your fellow students but makesureto prepare and write your report exclusively by your teammate. 2. The purpose of this study is to do a comprehensive and complete regression analysis fromwhat you have learned in this class. Therefore, the report should be detailed and conciseat thesame time. You should also consider your audience as an educated but non-technical personand do your best to explain your perspective as more uncomplicated as possible. 3. Your report’s body MUST be free of code. Only if there is a novelty in your code never
represented in the literature, it can be included it in the body. You should, however, appendyour code to the report as supplementary material or appendix. 4. You may use the software’s output to describe what you are discussing in the text. Theresult(plot, table, etc.) MUST be clearly explained by its caption and referred to in the body of your
report. Making your own tables to summarize the software’s output is highly recommended. 5. Page limit for your report is 10 pages at most (excluding reference and code in the appendix). 6. My intention in this assignment is to evaluate your coding and writing skills and howyoucanconvince the audience that you followed a professional methodology to answer a potential
question. You will also be assessed on your ability to simply and efficiently express your findingto the general audience.  7. Try to be innovative and push your limits as far as you can. This is what matters most. 8. Please put all your report into a single pdf file.
STA 4164 - Statistical Method III
Project 2– Guidelines
Summer 2022
REPORTS OUTLINE:
1. Abstract (You should write it at the beginning and summarize your research in a paragraph)
2. Objectives (Express your Main question(s) doing this study)
3. Source of your data (Where did you find your data, how is it collected, dimension, etc.)
4. Environment of your study (OS, programming language, software, version, …)
5. Data exploration
a. Frequency exploration (mean, range, etc. of your variables)
b. General graphs (primal explanatory scatter plots, histograms, bar charts, etc.)
c. Exploring general assumptions (are all regression assumptions satisfied)
d. Collinearity evaluation (Possible colinear variables and your remedial actions)
6. Data preparation
a. Dealing with missing values (size of missed values, any imputation, discarding, …)
b. Treating Outliers (distinguished observations, how did you deal with them, …)
c. Possible transformations (Reasoning behind and type of transformation)
d. Possible definition of dummy variables (merging, dummy definition, etc.)
7. Modeling:
a. Univariate model exploration (Brief result of univariate analysis)
b. Interaction and confounding exploration (discovered interactions, reasoning, etc)
c. Comprehensive model (Result of your comprehensive model exploration)
d. Model selection (Approaches you followed and findings)
e. Diagnostics (Model diagnostic and redefining the model)
f. Final best model
8. Discussion:
a. Final model interpretation (What does your model mean? Clear and simple.)
b. Model performance (accuracy) (if you pursued testing your model)
9. Conclusion (Clearly answer your main question from part 2 above)
10. References
11. Appendix (code, etc.)