Descriptive Transcript

When visual content is read or explained in audio, it is not reproduced in the visual column below.

Audio Visual
As you identify and gather statistics or data for your project, it's important to think about the credibility and authority of the materials you've selected. How do I evaluate data and Statistics? title slide
To address the issues of credibility and authority ask yourself three key questions: Who, why and how. Who collected the information you'll use in your research? Is the collector impartial or nonpartisan? Has the information been gathered by an individual or organization with a particular viewpoint or agenda?  
Visit the organization or advocacy group's web page. Read its mission statement to find out who serves on the group's board of directors or other governing body. What affiliations do these members have? Are they associated with known entities like universities or government agencies? Screenshot of Brookings website. About us in the menu is highlighted
Library resources may also help. For example, the PolicyFile database indexes, studies and reports from think tanks and other policy groups. Select PolicyFile from the A-Z database list on the library home page. Databases & Journals (A-Z) highlighted on library website homepage.
As shown in this illustration, PolicyFile allows you to filter reports by broad subject area and by organization political leaning. PolicyFiles limiters for Organization type and Organization political leaning are shown. Type includes development, Environmental, Finance & Economics, International, etc. Political leaning includes center-left, center-right, centrist, conservative and progressive.
How was the information collected? It's important to know about the methods used to conduct the survey, sample size, etc., These are details that will help you assess the project. Can you detect any biases in the selection of research participants? Was the survey conducted by phone? With the increasing prominence of mobile devices are landline surveys still representative?  
Information about survey methodology is frequently included in the notes associated with charts, tables and infographics. Sample bar chart with methodology information below graph highlighted.
Scholarly articles and research reports provide detailed descriptions of survey or study methodology and reference data sets or statistics used in the publication. In this article, the authors describe the measures they constructed to replicate interaction among senators. Sample article title and excerpt showing study methods.
Some sources may answer more than one question about your data or statistical resource. This example shows a study from the Inter University Consortium for Political and Social Research. ICPSR. In addition to discussing methodology, the text provides information about the Paul Simon Public Policy Institute. The institute's mission is described and we learned that the institute is also the primary source of project funding. Sample project description including title, summary and funding source.
Why was the information collected? Do the data collectors have a vested interest in a particular outcome? Are they trying to advocate for a particular policy? Are the data gathered as part of the process of governing, whether at the federal, state or local level? The data may have been gathered for administrative purposes. For example, federal government agencies collect census data, unemployment numbers, corporate tax returns or airline on-time flight data on an ongoing basis. Decorative icons and the logos for two government agencies that collect data: the US Census Bureau and the Bureau of Labor statistics.
When working with data or statistics, it's important to be aware of common pitfalls. Statistics and data presented out of context are confusing and subject to misinterpretation. This infographic on the changing racial and ethnic profile of the U.S. uses a map to represent a concept that has nothing to do with geography. Changing face of America Infographic. The west coast of the United States map is labeled 1960. The middle of the country is labeled 2010, and the east coast is labeled 2060. Bands of color show percentage of the population based on racial categories where white decreased from 85% to 43% and Hispanic increased from less than 16% to 31%.
Another common pitfall is the relationship between correlation and causation. For example, if a school buys tablets for all students and standardized test scores rise, it's tempting to think that tablets were the critical element. This may not be the case. Correlation does not equal causation
This tutorial is focused on tips for assessing the credibility and quality of the data and statistics you've gathered. If you need further assistance, please use one of our Ask a librarian services. Closing Slide: Ask a Librarian library.northeastern.edu/ask