by Elan / Jonathan / Shmuel / Daniel
Having reviewed some of the research pertinent to
the experiment we proposed, it became clear that there are several possible
directions in which the experiment can go. The primary research regarding
Hot-streaks has been reviewed in a book entitled "How we know what isn't
so" by Thomas Gilovich. Gilovich discusses the persistent belief held by
basketball players and fans in the "hot hand" despite the fact that statistical
analyses of shooting patterns actually finds that there is no such phenomenon.
People demonstrate an almost unbelievable immutability in maintaining their
belief in the hot hand. Red Auerbach , the brains behind the Boston Celtics
responded to the statistics presented by Gilovich saying " There are so
many variables in shooting the basketball that a paper like this doesn't
really mean anything". Gilovich posits that people are generally inclined
to allow their preconceived ideas about player's confidence and their corresponding
shooting skill to influence their perception of a pattern in the shooting,
seeing one where there is in fact none. Sequences of successive hits or
misses seem to be remembered more readily than sequences of alternating
hits and misses. A second explanation offered by Gilovich is that people
are generally inaccurate in determining what randomness really "should"
look like. People expect random series to contain few if any streaks
at all but rather an alternating pattern of hits and misses. As a result
truly random sequences, which do contain certain clumps of streaks will
not be perceived as being random. Gilovich refers to this as the "clustering
illusion". It is clear from this research that people sometimes have
a tendency to misconstrue random events, seeing a "pattern" where there
is, in fact, only chance at work. It would seem that there are certain
cognitive factors that allow people to make misperceptions of this nature.
We would like to posit that the type of phenomenon described in Gilovich's
experiment is more and less prevalent in different situations and contexts.
Circumstances where an individual is directly involved in an activity,
rather than an external observer, might lead him or her to more of the
type of "clustering illusion" described by Gilovich. In fact, Gilovich
concludes that those closest to the game of basketball exhibited a stronger
belief in the hot hand than those not as intimately involved in the game
of basketball. Although not exactly the type of relationship we are describing,
it may be extremely relevant in understanding the factors at work in such
type of decision-making processes.
Another experiment published in the journal of applied social psychology
investigated people's ability to assess coin tosses and determine whether
they were random or not. Consistent with Gilovich's results, this experiment
found that people were very poor at determining whether a series
was random or not. This is described by the authors as a reflection of
people's inability to recognize the "independence of events". This principal
applied in this situation translates into people's believing that a previous
coin toss could effect the outcome of a subsequent one. Thus, they would
tend to perceive what they believed to be non-random streaks in the patterns
of tosses. This finding buttresses the idea that people have a certain
notion of what randomness is supposed to look like and are inclined to
see patterns where none exist when their ideal concept of randomness is
not met.
The second part of this experiment tested
for the effect of motivational factors on this inaccurate perception of
random events. In fact, the experimenters found no relationship between
the motivational character of an activity and the ability to judge randomness.
Oh well.
A third experiment we reviewed, also published
in the Journal of Applied Social Psychology studied people's perception
about the risk level as a function of their level of control over the environment.
Subjects were asked to estimate the chance of a vehicular accident occurritical
probability involved. The tendency to associate less risk-under certain
circumstances- to situations where the participant, as opposed to a peer,
is involved are consistent with the idea that people might perceive there
to be less randomness when directly involved in an activity. However, the
fact that people don't always demonstrate this tendency (the "future" events)
indicated that there are other factors that influence this kind of bias
and it is important to determine under what circumstances it occurs.
- Gilovich, Thomas (1991). How We Know What Isn't So. New York: The Free Press
- Greenberg, Leilani; Chandler, Carla C. Journal of Applied Social Psychology. Vol 27(9) 760-780, May 1197.
- Klar, Yechiel; Medding, Aviva; Sarel, Dan. Nonunique invulnerability: Singular versus distributional probabilities and unrealistic optimism in comparative risk judgments. Orginizational Behavior & Human Decision Processes. Vol 67(2) 229-245, Aug 1996.
- Ladoceur, Robert; Paquet,Claude; Dube, Dominique. Journal of Applied Social Psychology. Vol 26(24) 2157-2166, Dec 1996.