Introductory Information

Dataset description:
Database of four Democratic presidential primary candidates' tweets vying for the 2020 US elections. The candidates include Joe Biden, Pete Buttigieg, Bernie Sanders, and Elizabeth Warren.  The timeframe was August 1, 2019 to when each candidate dropped out. The start date represented data collection six months prior to the first-in-the-nation Iowa caucuses held on February 3, 2020, thus including the communication most likely on voters’ minds as they formed their primary vote choices. The end dates were: March 1 for Buttigieg, March 5 for Warren, and April 8 for Sanders and Biden because Sanders dropped out that day making Biden the presumptive Democratic nominee.

File format: 
Excel (.xlsx)
Comma-separated values (.csv)

Principle investigator:
Dr. Lindsey Meeks
Department of Communication
University of Oklahoma
Burton Hall
610 Elm Ave.
Norman, OK 73069
lmeeks@ou.edu 

Dates of data collection:
2020-01-04
2020-04-09 

Geographic location of data collection:
United States

Date file was created:
2020-04-09 

Language information:
Tweets in English or Spanish. All Spanish tweets are immediately preceded or followed by an English translation offered by the respective presidential campaign.

Methodological Information

Code used for data collection: 
https://github.com/rainersigwald/twitter_archiver

Data Specific Information

Explanation of each column with column header title included in the parenthesis: 
- Screen name of the user (screen_name)
- Unique Twitter-created ID (id_str)
- Direct link to the tweet (link)
- Content of the tweet (text)
- Date and time of the tweet in UTC (date)
- Whether the tweet includes one of the 31 keywords used to narrow the dataset to those tweets that are related to environmental issues and climate change, with true indicating they are related and false indicating they don't (contains keyword).
- The 31 keywords included stemming to capture variants, e.g., plurality: anthropogenic, biodiverse, biofuel, carbon, clean, climate, earth, ecosystem, emission, environment, fossil, glacier, green, ice loss, ice sheet, methane, mitigation, nature, ocean, ozone, Paris (e.g., Paris Climate Agreement), planet, pollution, renewable, sea ice, sea level, solar, sustainability, temperature, warming, and weather. 
- Whether the tweets that passed the keyword filter are deemed relevant to environmental issues and climate change, e.g., when they mention "green," it is about "green jobs" and not "Greenwood District" as a campaign stop, with 1 indicating the tweet is relevant and included in the final sample, and 0 indicating they are not relevant and omitted from the final sample (relevancy review)

Units of measurement:
Each row represents an individual tweet

Definition for codes or symbols used to record missing data:
No missing data. Note that tweets marked as "deleted" by the original tweet creator are not surfaced in this dataset as per Twitter's developer API agreement. 

Specialized formats use:
None

Sharing and Access Information
Links to publications that cite or use the data:
https://doi.org/10.1080/19331681.2022.2069182 

Recommended citation for the data:
Meeks, L. (2022). 2020 US Democratic presidential primary candidate tweets. [Data set].

Funding: 
No funding provided for data collection or analysis 

Additional Information

Size of Excel file: 385KB

Software used for data processing: Microsoft Excel

Where the data falls in the research process: Analyzed data