Storytelling in Stats

Statistics is a language that connects the data to the people. Most of the time, the public eye is used to seeing them as a means of revealing the general opinion, such as in political polls. The collection of data is one of the points of contact between the inner and outer statistical communities. This point is a critical one, because questions ad answers must shimmy through a tight, winding tube constructed of wording and preconceptions. The questions, "Do you think Bernie Sanders ought to win the election?" and "Do you think the country would be a better place if Bernie Sanders won?" could bring very different results. Although these are one-sentence inquiries, they are stories pollsters implicitly share with the public. To illustrate what an impact this can have, I will share two very short stories. They are theoretically the same story, but I hope that the difference between the two is palpable.

 

 

Congrats. You have to take this poll call.

"...the first question involves your views on government spending. In your opinion, should the United States cut government spending?"

The initial question is short and to the point. The caller introduces no strong bias, although the use of the word "cut" does imply the negative. With no clear outline of what "government spending" is, the average person feels no guilt agreeing to the cuts. In fact, nearly 100% of people agree that government spending cuts are desirable. Now let's look at another way this call could be worded:

"... This first question involves the use of government resources. Currently, our taxes are being used to fund our road construction, the creation and sustaining of national parks, our military, social welfare, subsidize food expenses, and to assist the elderly. Many Americans are calling for cuts in government spending, which could impact some or all of these programs. Do you agree with them?"

This question is the same fundamentally, but the proposal of the question is bubble-wrapped. Most people would support at least some of the programs listed above, and I avoided mentioning anything that could be highly controversial. By the phrase "many Americans" I implied that they were outside the group, which would aid them in disagreeing with the idea of government cuts. Wording is very powerful in creating trends.