Write data cleaning script and generate codebook(s)

[mollyow]

• Data should be kept locally and not posted to repository; keep the raw and cleaned data files in a local folder named data in the repository root folder, i.e., Misinformation-replication/data.
• Cleaning scripts should live in the Misinformation-replication/code folder and run relative to that location. All cleaning code should be committed to the repository. The idea is that another team member could pull the cleaning scripts and run the cleaning code locally, on their own copy of the data, and get the same results.
• Only variables used for analysis should be included in the codebook, and kept in the cleaned data. All other variables should be discarded.
• Discuss cleaning decisions with @mollyow; it's important to know if results would be meaningfully changed based on coding decisions.
• Use dataMaid for codebook generation.
  • See example codebook from Offer-Westort et al., 2020, here; feel free to search Harvard Dataverse for other codebook examples.

[mollyow]

(note that some of this already exists, we will continue to revise cleaning script as we go)

[Tony-Zong]

Variables & Hypotheses doc: https://docs.google.com/document/d/1q3H142REXYi0ZY9v99kF7DHPw2K3zS5NMnfMmToCsOY/edit?usp=sharing