The Myth of the Enduring Personality - Part VII

Part I of the series here

F.A.E - The Fundamental Attribution Error

In rating people we have a propensity to commit the 'fundamental attribution error' which is where we overvalue the effect of human agency, and underestimate the effect of the present situation in explaining behaviour (Butt, 2004). To illustrate, when we have to slam on the breaks because of someone's driving, we may well blame the poor skill and idiotic driving of another. 

However, we might be overlooking the fact that they are lost and are unsure of where they are going, and had a moment of panic. In this sense, we are biased towards prioritising human agency, and underestimate the degree to which a situation determines behaviour.

If we apply this to Eysenck, he assumed widespread use of words by the general population pointed to the underlying objectivity of traits. Whilst we need not deny there are stable patterns in peoples behaviour, Eysenck's research was situated within this folk psychology paradigm and any results he derived were simply reinforcing his theories about the existence of traits, without an attempt to test the validity of them. 

In this sense, his entire research project begged the question, since he simply assumed that our categorisations pointed to real enduring qualities. In any case, any results he derived were simply a collection of people committing the fundamental attribution error, and this undermines his claim to objective validity.

The fundamental attribution error demonstrates our propensity to conform to the general populations assumptions about traits in everyday discourse, and assume these shape our agency. However, what we find is that Eyesenck simply reinforced these assumptions, by getting people to take part in his research program and not, even undertaking an independent verification of the existence of such traits. Moreover, his statistical measurement technology was flawed and we see that the case for objectivity here is very flimsy.

Check out Dustinland's wicked cartoons: http://dustinland.com/archives/archives626.html

This problem arises because Eyesenck simply drew a pattern from statistical inference of observations as justification for his theory. As Richards (2002) points out, the very psychometric survey measurements that we undertake can cause reification; where we falsely grant objective reality to the thing being measured.

To give an example, we could design a religious survey in 1700 say, that tried to measure devoutness to Christianity. We could make statements such as "Homosexuality is a sin and punishable by death", or "Witchcraft is rife in this village". 

We could collate a number of psychometric surveys and provide statistically derived empirical results. However, from these results all we have achieved is measuring a bunch of worthless religious assumptions. 

We have not proved that these assumptions are true, we have merely taken for granted their truth and produced a load of mathematical workings out to support our folly. 

Furthermore, these questions may have seemed valid in 1700 but were we to try and ask them now, they would sound quite ridiculous to any Christian. In this case, if we have supposedly measured devoutness to religion and the statements are irrelevant now, we can ask what exactly would we have measured back then? 

The simple answer is that we have reified the concept devoutness to religion and would have been engaged in sophistry.

What this tells us is that any psychometric studies are based on assumptions, and they produce a load of clever sounding mathematics that purport to back up these assumptions. All this is done without ever actually backing up these assumptions with evidence. 

What we really find is that this methodology is utterly worthless as a measure of truth, and in fact, this has been utilised as a tool by those of lowly scruples, to engage in the most wretched acts of deceit. 

It should be of little surprise that this is the tool of choice for charlatans in marketing, studies of 'public opinion', and your impotent government.

 This dangerous tool has been used to subvert the need to make claims based on evidence, and you should be suspicious of any statistical manipulations that are thrown at you.

 However, we can forgive Eyesenck as these facts were not known at the time, and he genuinely believed he was advancing science. It should be clear to you now though, that this questionnaire style methodology is fundamentally flawed and is still abused to this day by those who engage in the art of statistical manipulation.

Mischel's Criticisms

To cast further doubt on Eyesenck's research, Mischel (1968) started to assess the evidence for trait theories and found the key assumption that Eysenck had made was incorrect. It should be starting to become clear that Eyesenck had simply assumed stable personalities were based on a structural arrangement of traits. 

What Mischel discovered was that this assumption is not justified. Mischel cited extensive evidence that suggested the attributing of traits reflected the prejudice of the person doing the attributing, who had to draw on a culturally shared trait theory to frame their observations. Given that we use trait theory in everyday parlance, any observation is necessarily going to be couched in terms of traits. However, this does not necessarily mean our cultural categorisations are pointing to anything concrete.

Mischel cited a study that used Eyesenck's methodology of rating people from behind a screen, and using a psychometric survey to measure the responses. However, this time they observed strangers briefly in one condition, and in another observed people whom they knew personally. 

This study yielded a striking finding that went right to the heart of this kind of objective methodology. It was found that those doing the rating, produced the same stable trait structure in strangers that they observed very briefly, as in those that they knew well (Butt, 2012). 

This showed that under controlled conditions, subjects were prejudicial in their attribution of personality traits to strangers, and were willing to attribute them on very little evidence. It also showed that these attributions were based on the way they personally attributed traits to those they knew well (Butt, 2004).

This clearly demonstrated the error on which personality research was based. It is only natural, given the knowledge was situated in a paradigm that accepted folk psychological doctrine, that through seeing statistical patterns, practitioners could devise psychological categories for people in order to determine treatments. 

This kind of categorisation is now widely acknowledged to be unreliable and PCT actually denies this is useful, since it aims to understand the individual and acknowledges there are many individual differences, rather than narrow categorical dimensions.

Interestingly, PCT would embrace this subjective aspect of a person doing the personality construction, because when listening to how they construct someone's personality we can uncover some of the meanings this process has for them, and how they perceive the world. 

To this extent as a phenomenological exercise we can discover how people form their constructions, and perhaps some of the assumptions participants make. However, this endeavour fits in to the field of social psychology, and we would be moving beyond the scope of this essay.

Conclusion here