The Fallacy of Univariate Solutions to Complex Systems Problems
Complex biological systems, by definition, are composed of multiple components that interact non-linearly. The human brain constitutes, arguably, the most complex biological system known. Yet most investigation of the brain and its function is carried out using assumptions appropriate for simple systems—univariate design and linear statistical approaches. This heuristic must change before we can hope to discover and test interventions to improve the lives of individuals with complex disorders of brain development and function. Indeed, a movement away from simplistic models of biological systems will benefit essentially all domains of biology and medicine. The present brief essay lays the foundation for this argument.
The Univariate Fallacy is when someone argues that, because there is no single quality that separates two categories, the two categories do not exist and are actually just one category.
So for example, there’s no one single quality that separates Windows from Mac iOS, therefore Windows and iOS are the same.
There are multiple differences between Windows and iOS. There are also many commonalities. Yet there’s no one indicator that all Macs have that Windows OSs do not, though. Or vice versa. Because there isn’t one, concluding that Windows and iOS are the same would be the Univariate Fallacy.
Another example: There’s no single brain structure that separates left-handedness from right-handedness, therefore left or right handedness does not exist.
Here’s an example from Tw****r:
Another example I like is accent recognition: it’s a lot easier to say “She has a British accent” rather than individually describing all the phoneme-level features that your brain is using to make that judgement
The Univariate Fallacy can probably be thought of as a type of statistical fallacy, since this sort of thing seems to always happen in discussions with laypeople about differing statistical populations.
While I’m on the subject of statistics, there’s another statistics fail I see happen pretty regularly. Someone has named it the “Everest Regression”.
The Everest Regression is what happens when you “control” for a fundamental variable when comparing two populations. You might even think of it as the opposite side of, or similar lane to, the Univariate Fallacy. Maybe a sort of multivariate fallacy? I defer to the creator of the Everest Regression.
Basically, “controlling for height, Mount Everest is room temperature”.
Another: Controlling for number of electrons, helium and carbon have the same freezing point.
Controlling for distance from the equator, Alaska and Italy are the same climate.
Controlling for AU, Mars and Earth can support complex life.
You get the point. It’s assuming multivariate differences between univariate phenomena. This is understandable if you’re dealing with new phenomena, but is pointless and frankly sophistry to apply to concepts and categories that we already know are different along one or few axes in order to prove an ideological point.