Communication: Data & Statistics

Need some data, but don't know where to start?
Consider these questions:

1. Do you need data or statistics?
2. Have you seen the data you want published anywhere before?
   If Yes, refer back to that publication to find the data source.
   If No, continue to step 3.
3. Who may have collected this data?
   • Government organization
   • Academic researcher
   • Non-profit or NGO
   • Private company
   • Industry Association
4. Where can I access it?
   Is the data available freely online through the collector/producer's website?
   Is the data available through one of the Library's subscription databases?
   Is the data for sale, and can be purchased?
   Can the collector/producer of the data be contacted directly to get the data?

• What is Data?
  • Data are the raw ingredients of statistics. Data are un-interpreted pieces of information, the result of systematic observation or experimentation.
  • Data is most frequently used by graduate students and faculty who are doing secondary analysis. This analysis often requires the use of a statistical software package, such as SPSS, SAS, Stata, R, etc.
• What are Statistics? 
  • Statistics, or aggregated data, is data that has had a layer of analysis applied to it. This often takes the form of something like: 87% of all American adults are internet users, according to the Pew Research Center. 
  • Statistics are most frequently used by all researchers who are trying to support an argument.

The DREAM Lab supports a number of tools for acquiring and analyzing social media data. Feel free to schedule a consultation if these guides do not answer your questions.

• BrandWatch
  A browser-based tool for harvesting and analyzing Tweets, as well as other social media sources such as Reddit, WordPress.com blogs, and select public Facebook posts. Request access to BrandWatch
• Twarc
  A Free and Open Source Python script that uses Twitter's public API to harvest Tweets and their associated metadata
• NCapture
  A Chrome browser plug-in that will capture Tweets in a format usable in NVIVO qualitative data analysis software. Download NCatpure from the Chrome store
• Everything we know about Twitter
  An off-site guide to a compilation of our Twitter knowledge.
• Text Data Mining
  Social media data analysis can be considered a type of text data mining. The Library has a variety of other sources of texts that can be mined, and API's that can be used to compile your own full-text datasets.

There are also online sources of pre-existing datasets:

• Document the Now has a catalog of more than 100 datasets totaling more than 4 billion tweet ID's.
• Internet Archive
  • Nipsey Hussle death tweet archive
  • Obama White House Twitter Archives
    Separate files for @potus, @flotus, and @whitehouse
• The University of North Carolin has posted archived tweets from Confederate monument protests in 2017, 18, and 19.