The study was focused on suspense and the relationships that it may ha
ve with the traditional elements of a gambling situation. The framewor
k was Information Integration Theory. Two experiments were undertaken:
an exploratory experiment in which subjects were not placed in an act
ual gambling situation and a second experiment which did involve actua
l gambling conditions. In Experiment 1, the rule which, at group level
, accounted best for the integration of information on value and proba
bility was a rule where suspense depends primarily on the probabilitie
s and reaches its maximum when probability is 0.50. However, at indivi
dual level findings were considerably different. Overall, about half t
he subjects apparently equated suspense with attractiveness, expected
value, probability or value. The other half apparently acted on the ba
sis that the most suspenseful situations were those where the quantity
of information transmitted by the occurrence of one of the two outcom
es is maximal (case where probability is 10%), or when the average qua
ntity of information transmitted by the system formed by the two outco
mes is maximal (cases where probability is 50%) The results of Experim
ent 2 essentially confirmed those of Experiment 1.