The distinction between statistics and the individual is among the most far-reaching concepts of the modern world. We can view everything from science and its philosophy to politics, ethics, and economics in the light of that distinction.
Recall that statistics is about averages, trends, and patterns that emerge from large numbers (effects of scale), whereas the individual, though it may appear in a statistic as a data point, escapes most typical methods of quantification.
This has sweeping consequences on how we should make judgments and decisions in the world, for example:
- When a study shows that diet A or drug A has better health outcomes than diet B or drug B, this says little about what diet I should follow or what drug I should use. If I have a relevant genetic abnormality, path B may well be better for me individually.
- When a study finds significant differences between two groups of people (say, in physical or cognitive ability), this doesn’t allow me to judge individuals of either group based on those differences. If I did, I would be judging with a systematic error, a bias, at least as long as the group differences are gradual rather than categorical.
- When a politician develops an agenda on, say, migration policy (or when a citizen votes on it), he must strictly divorce demographic and economic considerations from sentiments towards individuals. Neither a xenophobic nor a xenophilic attitude can rationally inform a decision about large-scale collective change.
- When a recommender system tells people what to buy, watch, listen to, or click on next, it may not control any one individual, who still retains her freedom to choose, but it controls groups at a larger scale by nudging actions and thus manipulating collective behavior.
These examples show that significant features and effects that are highly relevant on an individual level may vanish or even reverse on the statistical level and that, conversely, certain consequences that are absent from any one individual may emerge for a statistic.
Hence, it can be perfectly rational to make personal decisions that go against scientific evidence and to make collective judgments that contradict individual-level sentiments and comprehensions.