Previous experiments have demonstrated but not explained people's
tendency to exaggerate the probability of conjunctive events. The
present study explores this tendency in several different contexts
designed to reveal how the overestimation process works; the events
which served as stimuli were either repetitive or unique with or
without causal links between them. The design was either a within-subject
or a between-subject design. The response mode was either percentages
or chances. The results suggest that people use different strategies
to assess conjunctions in different situations, all of which lead to
overestimation. These processes are discussed and some suggestions
are made about ways to overcome their negative effects.
