Key points are not available for this paper at this time.
An attribution analysis viewed opinion change as the outcome of a series of stages in message recipients' information processing. This processing was initiated by information about the communicator's background that led recipients to form an expectancy concerning what position the communicator would take in the message. The degree to which this premessage expectancy was confirmed or disconfirmed by the position the communicator took in the message then affected the outcome at each step of recipients' postmessage processing. In the first postmessage step, disconfirmation of the expectancy led recipients to attribute the message primarily to the factual evidence associated with the issue, and confirmation led them to attribute it primarily to the communicator's background. Then, to the extent that recipients attributed the message to the factual evidence rather than to the communicator's background, they perceived the communicator as unbiased, a response, which, in turn, increased opinion change toward the message. In addition, perception of the communicator as unbiased led to poorer comprehension of message content when the communicator's position had disconfirmed recipients' expectancies. In general, poor comprehension of message content lessened opinion change toward the message. A measure of subjects' expectancies demonstrated the premessage formation of expectancies about the communicator's position, and a structural analysis documented the postmessage stages.
Wood et al. (Sun,) studied this question.