Data and Statistics

Searching for data or statistics is different from searching for articles or other sources. Rather than thinking about what keywords to use, you'll have to think about who would have collected the data you are looking for, what they wanted to know, and what topics that information fits under. You may have to look around on a few websites before you find what you are looking for. 

Once you locate a government or research organization website that has data related to your question, you'll probably need to navigate through that website's topics to locate the specific information you need. Remember to keep looking if what you need is not in the first place you check. You might have to use related data to discuss you question if your perfect statistic is not available. 

Many research articles will include and interpret data, and most will orient their introduction using statistics that contextualize the issues they discuss. Start by searching for your topic with keywords, thinking about what type of article would discuss the information you need. 

For example: if you want to know how many people own cats compared to how many people own dogs, you are looking for statistics on the pet ownership rate or pet ownership data. Searching for cat ownership rate and dog ownership rate may also yield useful results. When looking at your results, consider that although an article may not be about pet ownership rates, a related article might still include the information you need. 

There is lots of data available on the internet, but like most web information, you will need to verify that the source is credible before assuming that the information is accurate. Below are some reliable data sources to start with. 

Are the data or statistics you found useful and credible? Here are some tips to keep in mind when looking for data and statistics:

• The source must be clearly stated and dated. If you can't tell who did the research or when, it's not a good statistic.
• Use sources whose primary purpose is to report data, not to influence opinion.
  • Examples include: scholarly and peer-reviewed sources, government sources, research organizations that share their funding sources, fact-based news sources
• Check the sample that is being tested. If the sample group is too narrow, it's not safe to assume that the statistic scales to a larger group.
• Is the information you found relevant to your topic? You may need to note differences between your topic and the data you found and include the specifics of the statistics you are referencing, rather than assume it is close enough.

When evaluating data found on the web, consider these questions as well: 

• Who collected this data?
• Is the organization biased, and in what ways might that skew the data?
• Where do they get their funding?

Check the website's "About" page to find this information. Most data has shortcomings, and a sign of good research practice to acknowledge and address those issues. If funding, methods, and goals of the organization are not shared, the data they provide may not be reliable. 

For more ways to evaluate credibility on the web, check out our Website Evaluation Checklist.

Data in this format has often not been analyzed, and can be difficult to use without proper tools and training. However, if you have taken a class that covered data analysis or if you have experience doing data analysis, these resources are a great place to start looking to datasets. 

Keep the following tips in mind when using statistics:

• Include visuals when available.
• Present the information clearly and completely. The context is important.
• If available, show data sets as they progress over time.
• Percentages and raw numbers from the same data set can paint different pictures. Use both if using only one is potentially misleading.
• Don't confuse correlation or association with causation.
• If you don't understand a statistic, don't use it.

In your text, name the source and explain the statistic.

Bad: Armenia leads the world in pirated software, with 93%.

Good: According to the Fifth Annual BSA and IDC Global Software Piracy Study, in 2007, Armenia had the largest software piracy rate with 93% (total number of units of pirated software deployed divided by the total units of software installed).

Make sure all tables and figures are appropriately labeled.

Consider your audience. Scientific papers rely more centrally on data and statistics, while argument papers use statistics as one of many persuasive techniques.

For more help with using data and evidence in your writing and assignments, visit the Tutoring & Writing Center.