Monday, 8 February 2010

Group Polarisation and UFOs

This article was a spin-off from a final year project I undertook when taking a psychology degree at Birkbeck College, University of London, which I completed in 1992, so it dates from around that time. I sent it to Kevin McClure's small magazine The Wild Places but it was not used. Kevin did take a related article, on 'UFO Witnesses and Fantasy Proneness', published in 1993. This one is dated now, and rather thin on deatail, but still worthy of preservation.


UFO sightings occur under a variety of conditions, often unfavourable to a definitive conclusion as to their nature (for example when seeing lights in the sky), so that in effect identification represents a probability judgment. It is difficult to estimate how many sightings can be attributed to natural phenomena, misperception, hallucination, fraud etc., and thus how many might have a veridical basis. The usual figure given is that 90-95% of sightings can be explained by psychological factors, aircraft, or meteorological/astronomical misidentifications.

There has long been an acknowledgement in the literature that issues of eyewitness testimony are of significance to the understanding of a UFO sighting. Thus witnesses' estimates of speed, distance and time can all be suspect, and reports can be 'cleaned up' retrospectively in order to fit in with prior belief structures. The investigator's role in possibly eliciting spurious detail has also been examined. There are other areas of psychology which have been brought to bear less frequently, but which could prove useful in interpreting reports of sightings.

It might be assumed that multiple witnesses are going to be more accurate than an individual, as they would have an opportunity to discuss what they had seen and thus pool all available information. It does seem likely that percipients are mutually influential. Zimmer found a correlation between a subject having had a sighting of a UFO, and knowing somebody else who had had a sighting, though not necessarily the same one. She puts forward two possibilities to account for this. The first is that there is a social network of witnesses who would tend to come together as a result of sightings; the second is that knowing someone who has had an experience increases one's own confidence that a stimulus really was a UFO. Networks certainly exist for those who claim to have had abduction experiences, but Zimmer's second hypothesis would seem to be more plausible for the bulk of low-definition sightings.

Randles & Warrington have noted that the incidence of UFO reports is in inverse proportion to a decrease in definition of the object sighted. In the same vein, Hilton & von Hippel argue that if a behaviour is ambiguous in the sense that it can fall into more than one category, the category label assigned to it would have a distorting effect on the observer's judgment; assigning the label exaggerates the difference between the behaviour in question and behaviours in other categories which might actually be very similar. By analogy, UFOs can be characterised as stereotypical of ambiguous aerial phenomena, so as a sighting increases in ambiguity, there should be an increased tendency to contrast it to the population of aerial phenomena, and label it a UFO rather than, say, an aeroplane.

A parallel to the examination of low-level sightings, with an emphasis on the difficulty in benchmarking the stimulus against a known frame of reference, can be seen in work carried out on the autokinetic effect. In this set-up, the subject looks at a point of light in a dark room. With no context within which to be judged, the light will appear to move about erratically. After a number of trials, the light will still appear to wander, but within narrowly confined limits. Subjects assessing the 'movement' of the light on their own will develop their own, idiosyncratic, description.

Sherif found that if a stooge made a firm declaration of what the limits were, the naive subject would tend to agree. This agreement would form the group norm. The norm is not necessarily a convergence towards the mean of a set of judgments – the consistency of one member can have a disproportionate influence. The emergence of a consensus thus represents a rational response to uncertain conditions. The autokinetic effect can be seen as similar to a low-definition UFO sighting in that both represent a probabilistic response, and that where more than one person is present, the decision made need not be some kind of average response, cancelling out extreme opinions so that a bland conclusion is reached.

'Group polarisation' is another concept which has become popular in social psychology in recent years. It has been defined by Hogg & Abrams as:

"the tendency for groups to make decisions which are more extreme than individuals in the direction initially favoured by the group."

The paradigm for examining its operation is analogous to that employed in examining the autokinetic effect. In both, an individual makes a judgment in isolation and then is exposed to group influence. The post-decision consensus is usually more extreme than the mean of individual pre-test responses. These shifts are essentially relativistic, concerning cultural norms, ie what is considered normative in a particular society.

Turner introduced a concept called "Referent Informational Influence", or RII, as a model to explain how polarisation occurs. It involves self-categorisation caused by group membership becoming psychologically salient, and which in turn produces stereotypical behaviour conforming to the perceived group norm. The in-group norm is the one which is best perceived to minimise intragroup yet maximise intergroup differences. This maximisation leads to a shift of the group mean away from the outgroup (which is why it is not merely the aggregate of the members' decisions).

Instead of an approach based on the individual characteristics of the participants, the RII emphasis is on the way in which an individual categorises her- or himself as a group member, with both social and physical reality constructed through a social consensus. Uncertainty arises as a result of disagreement with those one expects to be able to agree with, which is the case with other in-group members, those with whom one feels an affinity. No uncertainty will arise from disagreement with those whom expects to disagree. In an uncertain situation, the process of influence is able to take place, even in the absence of group pressure to conform. The corollary is that information is more influential when it has come from consistent in-group members than from other sources.

How does this apply to the interpretation of a stimulus as a UFO? Often more than one witness is present at a UFO sighting, and the decision reached could be influenced by the mechanism which causes group polarisation. A decision could of course be made either way, leading an initially sceptical person to conclude that a UFO had been sighted (risky shift), or a person who had thought that a UFO had been seen that there was a more ordinary explanation (cautious shift). Both are polarised decisions, but in opposite directions.

It could be argued that the group would be likelier to make a radical decision than an individual, and thus conclude that a UFO sighting had occurred. This is because of the finding that risky shifts tend to occur when the stakes are low, and a cautious shift when they are high (for example concerning somebody's health or happiness). With UFO sightings, apart from the possibility of ridicule, stakes would be low.

There would thus be a prediction that individuals in a group would be inclined to interpret an anomalous aerial object (defined as possessing a high degree of ambiguity) as a UFO in preference to a natural phenomenon. It should be noted that the prediction only covers sightings of stimuli which could be interpreted as UFOs, and does not cover abductions, which would be high-definition events, and for which other psychological mechanisms, such as fantasy proneness, might operate. Also, no conclusion as to the reality or otherwise of UFOs need be arrived at for an investigator to acknowledge that polarisation can occur.

The effect that the presence of another person and possibly group membership may have in the possible conclusion reached is dramatically shown by a case quoted by Randles & Warrington:

"With a friend (the witness) watched a strange object for half an hour. It changed no fewer than six times during this period. Each change followed a discussion between the witnesses when they said, "Oh, the experts will say 'That's a so-and-so'." The object promptly changed so that it was decidedly not reminiscent of whatever each particular 'so-and-so' was!"


Hilton, J.L. & von Hippel, W. (1990). The Role of Consistency in the Judgment of Stereotype-Relevant Behaviours. Personality and Social Psychology,16, 430-448.

Hogg, M.A. & Abrams, D. (1988). Social Identifications. Routledge.

Randles, J. & Warrington, P. (1979). UFOs: A British Viewpoint. Robert Hale Ltd.

Sherif, M. (1935). An Experimental Study of Stereotypes. Journal of Abnormal Social Psychology, 29, 371-375.

Turner, J.C., Wetherall, M.S., & Hogg, M.A. (1989). Referent Informational Influence and Group Polarisation. British Journal of Social Psychology, 28, 135-147.

Zimmer, T.A. (1984). Social Psychological Correlates of Possible UFO Sightings. Journal of Social Psychology, 123, 199-206.