Data Literacy Part 2 – Using data Ethically

As the caretaker of institutional data, the university has an obligation to protect the integrity and quality of institutional data, the privacy of data providers, and the security of institutional data while also maximizing its effective and efficient use. This guide brings together best practices, policies, and recommendations for the appropriate use and stewardship of institutional data resources throughout the data lifecycle to help data users find, interpret, evaluate, and effectively use data.

Data, Information, and Knowledge

The distinction between data, information, and knowledge is a critical component of the Data Use stage. Carol Tenopir, as quoted in Zins, 2007 defines theses concepts as:

• Data are facts that are the result of observation or measurement.
• Information is meaningful data. Or data arranged or interpreted in a way to provide meaning.
• Knowledge is internalized or understood information that can be used to make decisions.

Using data to create information or make decisions involves thinking critically, applying appropriate transformations and analytical techniques, considering bias, and conveying your interpretation of the data as information.

Ask the right questions

Not all questions can be answered with data. It’s important to form good questions when using data to make data-informed decisions. According to Barton Poulson in “Data Fluency: Exploring and Describing Data (2019),” using data involves taking a big picture view to ask:

• What are the most common values in the data (averages)?
• How much do things vary (ranges and outliers)?
• Is there a cause and effect relationship (causation vs. correlation)?
• What causes can you control (experimental controls)?

Get organized

Tips for staying organized when collecting data:

• Label your data with descriptive and consistent file names that are easy to understand, even if you open your data several years from now.  See examples of file naming schemas from UW–Madison Libraries Research Data Services
• Organize your files in well defined folders. Avoid over-nesting. Shallow folder structures, only two or three levels deep are easier to browse.
• Use tools to track versions of your data to make it easier to revert back to a previous version (tips on using version control). Google docs, Microsoft 365, and Box all have built-in version control to help track the changes made to each version of a file.
• Document all changes made in your processing and handling of the data.

Five Ways to Document Your Data

How understandable will your data be to other users, including yourself in the future? This guide considers a few best practice for documenting data, from just getting started to more complex tools..

1. Cite your data source
2. Define your data using a data dictionary
3. Create a metadata package describing the entire dataset (e.g., data specification or readme file)
4. Track the lineage of your data using data models, source-to-target mappings, and all transformations
5. Capture the reproducible environment of all changes made to your data

Learn the five ways to document your data

Data analysis and computing tools

There are literally hundreds of ways to analyze data. Statistical analysis, computational analysis, reporting of aggregate or clustered responses, and exploratory visualization each provide ways to derive insights from your data. Many books, encyclopedias, and handbooks offer discipline-specific training on research methods, for example see Sage Research Methods available online through the UW Madison Library. One of the overarching goals of data analysis is to apply methods and/or use tools in ways that are well documented and reproducible.

Services available to UW–Madison users:

• Center for High Throughput Computing (CHTC) – Access to big data computing resources are provided to all UW-Madison researchers, and most base services are free of charge.
• Research Cyberinfrastructure – DoIT’s Research Cyberinfrastructure group supports data science platforms for research and several cloud computing services.
• Social Science Computing Cooperative (SSCC) – Supports researchers at UW-Madison who use statistical analysis in their work.
• UW–Madison Design Lab – Offers software, training, assistance, and inspiration for visualizing your data.

Tip: Download popular licensed software from the UW–Madison Campus Software Library.