The Polling Crisis: Incongruence

Our current way of surveying the political landscape is by looking at diverse congregations of people within a larger global environment. Or, of course, as we need to hone in on America in this exercise, our current way of surveying in America would be national.

“National” — (or global, if you “scale” up)

Our current way of debating politics is from an individualistic-type lens, I would argue? Or at least small group-ish. (Ha).


But, our community outreach / public health/voting / etc. infrastructure— those are all designed around a community type level (of course, as they should be). City/county. So I guess you could say our current way of … hmm, exercising (?) politics/government is on that city/county level.


These different “levels” of viewing society from an anthropologic perspective (so, very BIG picture) are all inherently at odds with each other. Hence… that is the crux of the political polling crisis in the United States. Think about it. We are using mass amounts of data to analyze individual people. This does not work, and this does “trickle-down,” as these are interlocking systems that rely on one another. We can see this now more than ever. Why am I so confident in this assessment?

Because science does matter, ABA’s science is primarily researched via an experimental design known as “single-case design, which is not often discussed. I will briefly explain below.

Single case designs allow the researcher to avoid researching large swaths of populations at once; instead, it will enable the researcher to test specific interventions on individual people (or a small group of people). Each person acts as their own control group. But, what does this mean (in practical farms)?

It’s something you do “to” yourself all the time. When you take aspirin for a headache (as one example), you are essentially acting as your own control group. The “intervention” was the aspirin; if your headache goes away, yay! If it doesn’t, that sucks. Either way, though, you now know whether or not it works, and you essentially acted as your control group.

That is the type of research methods and precise analysis that should be at the forefront of political commentary these days. If we want to be a nation that believes in science, fact, and the importance of diversity, we must stop basing politics on any polling information that goes broader than immediate community “ecosystems.” If we don’t, I fear we erase constituents’ diversity and, subsequently, the science that espouses (albeit inadvertently) to highlight that very diversity.

Think of it this way— it would be like using global warming data to determine if you wanted to bring an umbrella that day. Like, okay, yeah, you will probably be right, based on how things are going, and that definitely matters— but you could also be very wrong. (This is hyperbolic, but it’s an illustration).

Things are predictable until they aren’t.