Statistically Improbable

Statistically Improbable

Having (as I haven’t) never studied statistics, or probability theory, I’m going out on a limb to assume the two are naturally related, (which I am). I’m going to go further and attempt some definitions: statistics are limited collections of examples of states of affairs derived for the purposes of generalising over the characteristics of that state of affairs as they could be expected to develop in separate instances. Probability theory is the study of how an overwhelming percentage of occurrences of behavioural activity may be convincing for the prediction of the natural progression of behaviour under similar circumstances. Now I’m going to tiptoe to the end of the limb and state that neither of these disciplines can make any claim to contribute towards the attainment of objective knowledge. One hundred per cent of examples achieves nothing like certainty; there is nothing deductive about probability; a universe of stars, however suggestive, implies nothing about the stellar constitution of a multiverse. Statistics are ‘there’, I’ll give them that, epistemologically speaking. They’re informative within their own ‘inertial frame of reference’, so to speak. To state a universal law on the basis of a set of statistics is to assume what is within that frame automatically accounts for all and any examples which lie outside. It’s not that the data is not indicative, it is simply that it is never, ever, ever implicit; and therefore one cannot state a law.

Do statistics produce objective knowledge in psychology? If you believe under a set of circumstances, with a 99.999% similarity of physical, historical and social resemblance, there is a 100% likelihood you would act the same way, make precisely the same decision, as another human being with respect to the situation you were both faced with, then it is an inescapable consequence that there is no such thing as free will. The 0.001% chance of an exception is what breaks the rule; and there will always be a fractional discrepancy. 100% of examples does not constitute certainty, because they are examples - they are by definition not exhaustive descriptively. That 0.001% chance is the rest of the multiverse. The one-out-of a-hundred people who diverged from the flow of the majority is the one-out-of-one who defies the universal application of the conclusion. Or else we sacrifice the veracity of a solitary truth-bearer in a crowd of liars.

But surely, surely, retorts the psychologist, even if one out of ten can occasionally be the faithful portrait of truth, when you start to observe the same behaviour exhibited in 99 out of a hundred cases, you can start to discover REAL truth - you have evidence of objective knowledge - and can form judgements accordingly. I suppose it depends on the size of the conspiracy doesn’t it? (For example). In the contingency of real life there is always the possibility of a black swan causing a splash amidst an ‘overwhelming’ majority of white swans. I could go through a list of kinds of psychological experiment which fail to produce objective knowledge in spite of statistical evidence but, paradoxically, we can rely on just a few examples…

Let’s take an easy example first: (this is not a real experiment, just a hypothesis) 10 people claim (- personal claims - the first hurdle on the road to objectivity) to have a phobia of wasps. They are each encouraged, one by one, to confront their fear by being put in a room with a large number of said creatures, in order to accustom them to the object of their anxiety, with the intended outcome of allaying their phobia. 8 out of the 10 subjects express a decline in their fear of wasps as a result of the ordeal, while the remaining two claim an elevation of their phobia. The question is - does this count as evidence that confronting your phobias reduces their ferocity? I would argue the answer to this question is yes, but with a thousand qualifications:

1. Yes, but not necessarily.
2. Yes, but with the sufficient condition that it concerns wasp-haters in a room full of wasps.
3. Yes, but it is possible (although unlikely) your phobia will be exacerbated.
4. Yes, but with the sufficient condition that you satisfy the anthropological features of the subjects included in the experiment whose phobias were successfully abated.
5 - 1000. Yes, but probably not for 200 people out of a thousand other unexamined cases.

I think it is rational to conclude there is little scope for objective knowledge in the previous wasp experiment, given the stark abundance of caveats brought into consideration. Let’s take a different type of (hypothesised) experiment to evaluate: a million people are asked if they would prefer being stung by a hornet, or being stung by a bee, if it meant the creature died as a result of the attack. The idea being to discover to what extent people would suffer physical pain in order to preserve one or the other species. The majority of people (I am supposing) choose to save the life of the hornet, and let the bee sting them. Let’s say 999,541 people choose the bee sting. This result would be absolutely overwhelming in psychological terms. But what would it prove? That people overwhelmingly value their own pleasure over the mortality of a bee? Psychologically speaking, it could be taken to imply that people overwhelmingly prefer to save the life of a hornet over that of a bee! It would depend on how the experiment’s results were presented and interpreted. This discrepancy represents the chasm of difference in psychological opinion in view of the results of an experiment; in addition, there are the same range of contentions as in the first experiment. Although more convincing statistically, this experiment is as fallible in an attempt to produce objective knowledge, because human nature, or, more specifically, animal behaviour, always throws in a number of wild cards. Statistics, and by proxy, probability theory, in the sphere of human activity, always comes up against an unconquerable opponent: context. And this factor belies the key to understanding how the phenomenon of free will is manifested in our everyday lives.

Hence we can make a leap to the point that probability, insofar as it is unproductive in the endeavour to find objective knowledge in the arena of psychology, is similarly uninformative as to the typology inherent in the realm of economics - (the organisation of finance in human societies). Human behaviour is to be considered unfathomable to any scientific degree, however far the psychological/mathematical theory appears to support a model of logical progression. Without free will there is no intention; without intention there is no consciousness; without consciousness all that is left is mere physical determinacy; therefore we cannot determine what a human being will do with 100% accuracy without denying their claim to possess free will. Probability theory dictates what is theoretically probable, and statistics represent what happens to be the case; neither can aspire to be universally predictive in the sphere of human nature, be it social sciences or economics or any other domain of social studies. Statistics belong to theoretical analysis of the implications of surveys of types of behaviour for the understanding of how we might not will act.



Hover in the cells to see a range of categories.

  Choose One...

Epistemology/Metaphysics Logic Theology Politics/Ethics Psychology/The Mind Reviews Swat


...or swat the wasp for a chance discourse!