Finding Statistics

Why are Statistics Important?

Statistics are important because they help people make informed decisions. Governments, organizations, and businesses all collect statistics to help them track progress, measure performance, analyze problems, and prioritize. For example, the U.S. Census Bureau collects information from people about where they live and their age. This information can help cities decide where they should build a new hospital if they find that there is a high elderly population in an area or a new school, if they find there are many families with young children.

On a personal level, statistics can be a great way to enhance your argument in a research paper or presentation. They show that there is evidence to back up your claim and can add credibility to your work. Statistics often create an emotional response in your audience. Think about how you feel when someone can back up their argument with statistics? Don't the statistics make you feel more strongly to the argument?

The below video by Ms. Emma Stevenson will help explain how statistics can help you in a research paper or project:

Misleading Statistics

Statistics are an excellent way to enhance an argument and persuade your audience; however, there are some considerations to keep in mind. Statistics can be misleading, because they are often taken out of context. Sometimes, important information is left out about how the statistic was collected in order to make it seem more dramatic, proving big ideas or generalizations that it wouldn't if the rest of the information was included. 

For example, let's say you found a statistic that said 5 out of 5 dentists recommend a certain brand of toothpaste. That sounds like this is a great brand of toothpaste that everyone should use. However, what if you found out that the dentists were all asked if they would recommend that brand of toothpaste or not brushing your teeth at all? Of course all of the dentists are going to pick the brand of toothpaste. This makes the 5 out of 5 recommendation basically meaningless. You might assume when you see this statistic that dentists were ranking this toothpaste brand over other toothpaste brands, instead of against not brushing your teeth at all; this makes the statistic misleading.

Another way statistics can be misleading is in the sample size that the data was collected in. For example, let's say you found a statistic that says 4 out of 5 women prefer wearing high heels over flats to work. However, when you start looking closer at the source the statistic came from, you find that this statistic came from someone asking 5 women they work with in a corporate law firm if they liked wearing heels or flats to work. This is a problem for several reasons.

First, the information was collected from a very small sample size (5 women who all work at the same place). These 5 women cannot represent all women and their opinions on high heels. Second, this sample is very biased, because all of the women work in the same corporate law firm. These women's opinions are not going to reflect all women's opinions, regardless of the number of women sampled, because the women are too similar to one another. If all women in all industries were surveyed for this question, the statistic would look very different. Because of this, it's always important to know the context of any statistic before you use it in your argument. Similarly, you want to be wary of statistics you find that don't have context or can't be tracked back to an original source.

Just like evaluating the credibility of your sources, you will want to do the same for when you want to use statistics in your research. Ask yourself the following questions:

• Can you find the original source that this statistic was published in? This will help you understand the context of the statistics.
• Who published the original source and where was it published?
• Who collected the information for the statistics? Do they have any kind of agenda/stake in the statistics?
• When was the information collected? Could it be out of date?
• How big was the sample size/how much data was collected? What were the demographics of the sample size? This will help you figure out if the statistics are representative of a certain group or area. 

Here is an article that goes deeper into how statistics can be misleading and ways to determine whether your statistics are misleading or not.