Project

Due Date: March 16, 2018 4:30 pm

Answer a research question using the statistical and computational methods used in this course. It can be a descriptive, predictive, or causal question. What is important for this reports is that the question and stated clearly, the methods and data used to answer the question are described clearly, and a discussion of why the methods and data used to answer the question are appropriate, and a discusison of th assumptions made by those methods and the limits of what we can learn from this research.

Submission

Open a new issue in your project repo, UW-POLS501/username, titled Final Project Submission. Assign it to @jrnold.

Format

The format of the project is as a reproducible R markdown documentin your course github repository, e.g. UW-POLS501/username. All data and code necessary to reproduce the analyses presented in the paper must be provided. The data and analyses should be well documented and understandable to a reader. In particular, it should be clear how the code relates to the analyses presented in the paper. The analysis code must run and reproduce the analyses as presented in the paper.

You should have a file named README.md in the root of project with the following format:

# Title of your project

A few lines about what is in the repository.

# Dependencies

Describe what steps someone would need to take in order to run your code.
This includes listing the packages you use, preferrably with the necessary code:
\`\`\`
packages_used <- c("tidyverse")
install.packages(packages_used)
\`\`\`

Does the user need to download specific files not included in the repository in order to be able to replicate your analysis?


# Instructions

After installing any dependencies, what should a user do to replicate your code.

- What R scripts must be run in what order? 
- What R Markdown files should be knit? 

This may be as simple as saying open up `somefile.Rmd` and knit.


# Organization

Describe the organization of your project. 

- What is in each directory? 
- What are the important files? 

You should also delete any extraneous or unneeded files.

Your R Markdown file MUST:

knit successfully
have chunk describing all libraries used near the top of the document
the output of sessionInfo at the end of the file
no extraneous error messages, warnings, or overly verbose output (e.g. printing 1,000 rows of a data frame).

Content

Your project should effectively combine prose, code, and results to answer a research question.

If a descriptive question: How do your methods measure something that wasn’t obvious before?
If a predictive question:
- How do you measure out-of-sample errors? Is is consistent with the predictions you would like to make?
- What is your out-of-sample error? How does your model compare with alternatives?
If a causal inference question:
- What research design are you using? What are the assumptions of this design?
- You are encouraged to use confidence intervals to present results, and discouraged from using p-values.
- You must discuss the magnitude (substantive signficance) of your results.
What are the properties of your sample? How generalizable is it?
What are the assumptions of your research design? How plausible are they? What would be needed to improve on your research design?
Use visualizations where possible to describe your results