Greetings STARFLEET,
Today’s topic for IDIC is statistics. These days, statistics are everywhere and used in all mediums of communication. Unfortunately, they’re also incredibly easy to manipulate into saying whatever the writer wants to without looking biased.
A statistic is a fact or figure derived from a numerical data set. When it’s a statistic about people, that data often takes the form of surveys. When it’s about the natural world, the data come from observation. For example, 23% of Americans surveyed report eating less meat in 2019 (Gallup).
With statistics about people or behavior, it’s challenging to gather truly unbiased data for statistics. For starters, the people who conduct surveys and studies have unconscious biases. Scientific studies do their best to mitigate these biases using double-blind studies, large sample sizes, and other measures.
Non-scientific studies rarely employ these measures. That should make them a little suspect from the start and generally require verification. Unfortunately, verifying the validity of a survey is rarely a priority for most people.
On the other hand, practices like selective polling and vague questions make it easy to manipulate the data behind the statistics. Selective polling means choosing the people who participate in the survey. It’s things like only surveying people in a city on rural issues.
As a non-contentious example, let’s look at apples. Say our survey question was “What is the best food?” and the options are apple, apple pie, caramel apple, or applesauce. You’ll get different data if you ask this to a group of kindergarteners than you would asking a group of mid-career adults.
Then there’s how data is phrased. For example, we could take our apple question and state it as 10,000 people say apples are the best. We could also say that most people agree that apple pie is the healthiest pie. With a bit of messaging manipulation, the same data can say many things.
Unfortunately, while humans want to treat numbers as objective measurements, they are not. It takes extensive measures to gain impartial data. Even then, the individual reader or watcher must verify that these measures are taken. So let’s make sure we’re only sharing well-researched and unbiased statistics.
Let’s be the change, STARFLEET.
Well put.
As someone who has had to write survey questions and review responses, this is quite true.
When you are filling out a questionnaire, every question has a purpose. Some may be to measure the demographics of participants, many are to specifically gain data in order to qualify and/or quantify already predetermined courses of action or to establish what may or may not resonate within a target audience.
If a question is tremendously vague, you can lay a substantial bet with a massive return that their is an agenda afoot