MTM Writing Clinic
MLK 252 
Phone: 713-313-7981
Email: owl@tsu.edu


Statistical Evidence

Statistics are one of the most difficult forms of evidence to evaluate and effectively incorporate into your writing. While "hard numbers" are a great form of evidence, be wary of trusting data presented from any source.

 

 

Credibility

Like any form of evidence, statistics must be credible. In addition to the rigors all evidence should undergo, consider the following when evaluating the credibility of statistics:

 

  • Where's the bias?

Research that results in statistical evidence is done for a reason. Someone formulated the questions, selected the persons who would be asked, and often, such data is often used to persuade future decisions. By evaluating the survey itself, and not just the results, it is often possible to tease out the bias involved in the data and determine whether the bias interferes with the validity of the data set.

  • Where's the data?

Generally speaking, data does not come fully reported. Authors commonly present data that they have analyzed and report their interpretation of the results. If you cannot access the entire data set, and therefore cannot verify the result, should you trust it? The safest course is to find similar analyses on the same topic, and through a variety of sources verify that the data is correct.

  • Is it correct?

Even if you've verified that an author's data is good, it does not mean that the author has correctly interpreted that data. A common error for authors to make is confusing correlation with causation. This means that that change noted does not necessarily indicate that it is the cause.

The simplest way to explain this error is with an example. Let's say a study finds that children ages 7-9 who eat less candy weigh more. Do we need to start feeding our kids more candy? Or should we consider other possibilities? 

There is a high probability that another cause is responsible for this result, and finding additional studies on weight in children may help find the cause.

Using Statistics as Evidence

Once you have determined a statistic is credible, it is ready to use as evidence. However, it is not enough to merely provide your statistic. You must also assert its credibility and relevance in your own writing. 

For example, if you were discussing poverty in your region, you might look up data presented by the U.S. Census Bureau. You would discover that Texas has a poverty rate of 17.4%. But what does that mean? Is Texas poverty high or low? Searching further, you'd discover that the nationwide poverty rate is 14.9%.

In presenting this information, you would want to indicate where you got your information (credibility) as well as how the numbers support your claim. 

Between 2008-2012, the U.S. Census Bureau reports that 17.4% of Texas residents lived below the poverty line, while nationwide, the poverty rate was lower at 14.9%. Although median income for Texas families is deemed high, the poverty rate indicates that there is large discrepancy in the earnings potential of the richest and poorest in Texas.