Subjects ( n = 200) received a detailed description of a product and were asked to rate their attitudes about this product. Presentation order, source credibility and message framing were manipulated in a 2× 2× 2 completely crossed factorial design. Subjects who received a positively framed message rated product attitudes significantly greater than those subjects who received a negatively framed message. Also, significant differences in message framing effects were found for those subjects who received the framed message first in the nonexpert condition (credibility) and those subjects who received the framed message last in the expert condition. Findings are then discussed.
Buda et al. (Sat,) studied this question.