The BDN Opinion section operates independently and does not set news policies or contribute to reporting or editing articles elsewhere in the newspaper or on bangordailynews.com
Kelsi G Hobbs is an assistant professor of economics at the University of Maine. She is a member of the Maine chapter of the national Scholars Strategy Network, which brings together scholars across the country to address public challenges and their policy implications. Members’ columns appear in the BDN every other week.
In his 1954 text, “How to Lie with Statistics,” Darrel Huff wrote: “The secret language of statistics, so appealing in a fact-minded culture, is employed to sensationalize, inflate, confuse, and oversimplify.” Although the text’s examples are now dated, the concepts remain useful. I want to discuss one of these concepts: the sample.
A sample is a subset of observations from a population that is used to estimate the characteristics of the whole population. A sample is useful, because it allows us to gain an understanding of the population without having to survey the entire population. However, the way a sample is chosen is important.
For example, according to a 2024 survey from the Pew Research Center, three in 10 Americans reported making at least one resolution last year. Did Pew ask every American whether they made a New Year’s resolution or not? No, of course not. I was never asked this question, and you probably were not asked either. Instead of asking every American about their New Year’s resolution, which would be logistically impossible, Pew used a sample. Their data came from a survey that was sent to 5,140 adults, who were chosen through a national random sample.
Although a “small” number of adults were surveyed at a certain point in time, which may contribute to the results, the individuals surveyed were selected from a random sample. Those of us who deal with statistics love random sampling, because it is often the best way to estimate the characteristics of the whole population.
As an economics professor, I have a lot of students that tell me that they “do not like math.” Can I assume that all students dislike math? Well, according to a 2018 survey commissioned by Texas Instruments (TI) and conducted by Research Now Group, Inc., 46 percent of students report that they like or love math. The data came from a survey that was sent out by an independent survey research firm to 1,007 students, who were between the ages of 13 and 18.
My initial conclusion was derived from a very specific sample: students taking my Principles of Microeconomics course at the University of Maine who are willing to tell me directly that they “do not like math.” I used a convenience sample, a group that was easy to access. There may be times when convenience sampling may be effective, but in this situation, my sampling technique was more likely than not flawed.
Texas Instruments did not run the survey themselves, instead they used an independent survey research firm. Why? Because TI has a reason for students to like math — they sell calculators! If the survey was done internally, we may be concerned that the results were biased. Do students really like math, or did TI just send out a survey to students who were likely to say they liked math? The fact that they did not run the survey themselves makes their results more credible.
However, unlike the Pew data, it is unclear if the Texas Instruments data came from students who were randomly sampled. This matters, because it is not clear who was surveyed. Maybe the research firm wanted to make sure their client, TI, got a positive result, so they only surveyed students who own TI calculators. Just as I should take my conclusion with a grain of salt, maybe we should also be wary of these results.
As we begin a new year, I want to encourage all of us to think critically about the statistics we are presented. Statistics are powerful, they help us to understand trends and make decisions. Numbers do not lie, but the people who use them may be trying to spin the results to their benefit. I challenge you to find the context of the numbers you read, hear, or use this year. It is important that we understand the context of the statistics we are presented, so that we are not fooled or misled by the data.
