Research & Data Management Services: Collect & Create

Walker Library currently subscribes to two resources which can help you with your text analysis research:

Primary data is that data which has been originated by researchers conducting original research. Not every research project will be able to use primary data, for a variety of reasons such as lack of funding or insufficent recruitment of participants. 

So when you cannot gather primary data yourself, or choose not to conduct original research, where can you locate data? The solution is to use a secondary data source.

Secondary data is that data which has been originated by a previous research project but is being repurposed for a new research purpose. Many researchers and organizations, including state and federal government agencies, have made their primary data sets freely available for others to use. If you use someone else's data set, it is important to properly cite the data (see the tab for Reuse at the top of this LibGuide).

In most cases, you can search the following sources by going to their Search fields and entering your discipline or area of interest; for example "literacy" or "criminal justice".

Here are sources of secondary data sets (which are multidisciplinary sources):

• data.gov
• The Dataverse Project
• Dryad
• Figshare
• Harvard Dataverse

Here are sources of secondary data sets for Computer Science (which a focus on AI/ML):

• Hugging Face
• Kaggle

Here are sources of secondary data sets for the Humanities:

• Project Gutenberg (for text analysis)
• Text Encoding Initiative (for text analysis)
• Not a source of secondary data, however useful as a possible place to publish; also a place to learn about text analysis in the humanities: Journal of Open Humanities Data

Here are sources of secondary data sets for the Sciences:

• Astrophysics data system
• data.NASA.gov
• National Center for Science and Engineering Statistics
• National Centers for Environmental Information
• National Institute of Standards and Technology
• National Library of Medicine
• Zenodo

Here are sources of secondary data sets for the Social Sciences:

• Consortium of European Social Science Data Archives
• Inter-university Consortium for Political and Social Research

With permission from the University of Northern Iowa, here is a link to their extensive list of Secondary Data Sets, including the following data sets by subject area:

• Demographic and Social Statistics
• Economics and Finance
• Health, Human Services, and Criminal Justice
• Political Science, Public Policy, and Geography

Here are resources for learning how to conduct an evidence synthesis:

Getting Started:

• Cornell University Libraries has an excellent LibGuide, including a step-by-step process to conducting a systematic review: Cornell University
• Cornell University also has a very handy Decision Tree to help you decide which type of review is best for you.
• Information about conducting a scoping review can be found here: PRISMA for Scoping Reviews

Trainings:

• An extensive and well-regarded free online training for both evidence synthesis as well as meta-analysis is through Carnegie Mellon University: Carnegie Mellon Systematic Reviews and Meta-Analysis
• The Evidence Synthesis Institute is a training geared towards librarians, but you may benefit from reading their online materials: Evidence Synthesis Pressbook