All over the internet and social media numbers and statistics are being thrown at us. Where did they come from? How were they collected? What do they really show? They wouldn’t be out there if they were not proving someone’s point. But can we trust them? Never accept this type of information blindly. Look up the study to see if it was done correctly. Also, see if the results and being relayed to you in the right way. Here are some things to look for.
It’s all in the numbers! How many people did they get to take their survey? The larger the number, the better, and the more the percentages will mean. Comparing a study with one hundred participants to a study with one thousand participants is not possible. The research that had the most take part in it will have results that represent more of the community’s beliefs. Beware of statistics that come from studies with low numbers.
Look to see who paid for the survey to happen, and who conducted it. If there was a study to find the best soda brand, and Coke was running it, would it be surprising if Coke came out the winner? A survey run by an independent company would be preferred. Being questioned by a company about them could lead to skewed results. It is easier, to tell the truth when one of the options is not sitting right in front of you. Telling them you don’t like their product as much as their competitor could be a little awkward. Always check the source of the survey.
Finding numbers to back up an argument is normal. It’s common to find those quoting part of a study that supports them while leaving out the parts that do not. Without seeing the complete results, you can’t completely trust it. When searching make sure the survey has to do with what they are speaking about. Relevance to the subject is critical. You can draw many conclusions from one set of results. If the survey did not cover your topic, specifically, you shouldn’t use those results to support it.
The National Science Foundation warns us of the wording of questions, “One can be reasonably sure a survey is not scientific if questions are biased—that is, designed to elicit the answers that the survey’s sponsors want to get.” Survey results will show the questions that were asked. Read through them to see if there is any bias. Anything that could influence the answer in any way, no matter how subtle, is a biased question. These questions should make you think twice about trusting the results.
Trust is something that should be earned when it comes to surveys. Do your homework before believing and sharing data. Know where it came from and who conducted the survey. Understand the questions being asked and what the results show.