Skip to Main Content

Finding Data and Statistics Sources: Evaluating Statistics

Overview

Evaluating statistical information is crucial.   There are two ways in which you must evaluate statistical information:

  • Reliability of the source of the information
  • Relevance to your research

You will need to evaluate every statistic that you use on these two points.

The Misuse of Statistics

Evaluating Statistics

How can you determine if a statistical source is reliable?

Evaluating the source of compiled statistics is crucial. Answer these questions to evaluate how reliable the data is for a particular source. Be sure to review all sources for:

Authenticity

  • Who is publishing the data?
  • Is the authorship easy to verify?
  • Who conducted and paid for the study?
  • What was the motivation for creating the data?

Reputability and Bias

  • Are the statistics made available by a reputable agency or organization?
  • Do you know anything about the agency/organization or can you find any information about it?
  • Established sources may produce biased data. For example, a cigarette company might produce biased data about the number of deaths caused by smoking. Be critical!
  • Remember, anyone can publish on the web—just because data is available on the Internet does not automatically guarantee reliability.

Validity

  • Is it possible that organizational or personal biases might impact the validity of the data?
  • Does the source note how the data was collected?
  • Does the methodology used appear to be appropriate?
  • Does the data seem plausible? Do the statistics seem to fit with the other data that you have found? If the data seems unlikely, are you able to verify it in another source?

Accuracy

  • Are there typos, poor grammar, or other clues that would lead you to question the accuracy of the data?
  • Is the material included current?
  • Is a date given for when the information was collected?
  • How frequently is the data updated?

Are there sources that are always reliable?

  • Government sources, educational institutions, and policy institutes are generally good sources to rely on.
  • Private parties or personal web sites, organizations that may not be objective, and unknown sources should be scrutinized carefully before using statistics from them.

How can you tell if a statistic is relevant to your research? 

When selecting statistics, be sure they meet your needs in the following three areas:

  1. Definition of terms: Do the statistics represent the specific topic and parameters you have chosen. Are you using and defining the same terms in the same way as the source? For example, if you are looking for the number of golf courses in the United States, and you want both public and private courses, does the data you have found include both public and private courses? 
  2. Time period: Is the time period you need represented by the data you are using? For example, data in a 2018 study might actually reflect 2017 statistics. Are you looking for "calendar" year information but the source is listing "fiscal" year information?
  3. Location: What location or area is covered by the data? For example, if you want information on the city of Chicago, does the source you are using include the "metropolitan" area or just the city?

Using Statistics or How do I read this table, anyway? 

Statistical data can be complex. There is often a lot of information contained in a statistical table, graph, or data set. There are some common elements you should look for when examining a statistical source. This information will be critical to your evaluation of the source, as well as how you use the data for your research.

Every piece of information in a table or graph is important to the understanding of the statistics presented. Statistical tables and graphs pack a great deal of information into very little space.   

  • The title and the note following the title (circled in blue) provide information about the content of the table. This information says what the table is about and provides an explanation of what is and is not included in the scope of the statistics.
  • Most tables will present information about a topic that is modified by another criteria such as time or place. The information will be displayed in columns and rows (a tabular format). It is, therefore, important to know what is communicated in the table itself by reading the headers for the columns and rows.  
  • Compiled statistics will sometimes have an original source for the data and the source of where the table is printed. Knowing the source of the statistics is important for several reasons: (1) it helps you determine credibility of the statistics, (2) you must cite the statistical source so others know the source is credible (and to avoid plagiarism), and (3) you might want to look at the original source to find other similar statistics. 

 

re-use/adapt note

The Information in this guide was re-used and adapted with permission of Angela Bonnell of Illinois State University, Milner Library.