Prediction (Psychology)

Many psychologists have promoted the use of personality traits and
other dispositional constructs for the prediction of human behavior.
However, other psychologists have concluded that individuals do not
exhibit sufficient cross-situational consistency in their behavior to
warrant the use of dispositional measures as predictors of behavior
(e.g., Mischel, 1968; Nisbett, 1980). The present research was
designed to demonstrate that cross-situational consistency correlation
coefficients may be inadequate indicators of the utility of
dispositional constructs. In particular, it was proposed that a
consideration of the situational constraint present within a situation
and the degree of the opportunity to self-select into a situation are
important factors which aid in the specification of when personality
traits can and cannot be used successfully to predict human behavior.

The goal of this study was to exaoine the relatively reinforcing
properties of predictability (or the absence of it)
in human subjects' choice for control and no control presentations
of a mildly aversive stimulus (Time Out from positive
reinforcement). Subjects, overall, preferred predictability
with control significantly more than predictability without
control (Experinent I). There was less of a difference in
their preference for control in the absence of predictability
(Experiment II). Results demonstrated greater preference for
control among males than females and female variability was
large. The "motive to avoid success" on the part of female
subjects was purported to explain the variability and lesser
preference for the control conditions.