The first cognitive bias we examined was selective perception and, more
specifically, selective perception of software quality. This bias acted on both stages
of the decision process depicted in our model. As expected, the software quality
assurance treatment group had a higher level of selective perception of software
quality than did the marketing treatment group. In turn, higher levels of selective
perception of software quality led to higher levels of software quality problem
recognition, which is in accordance with Hogarth’s (1987) suggestion that selective
perception acts in the early stages of the decision process. In addition to this
indirect effect of selective perception on escalation, we also found evidence of a
direct negative relationship between selective perception of software quality and
proclivity to launch as planned. One explanation for this is that the software quality
assurance group may have exercised greater caution, believing that additional tests
were needed before the product could be released