First impression bias refers to a limitation of human information processin
g in which people are strongly influenced by the first piece of information
that they are exposed to, and that they are biased in evaluating subsequen
t information in the direction of the initial influence. The psychology lit
erature has portrayed first impression bias as a virtually "inherent" human
bias. Drawing from multimedia literature, this study identifies several ch
aracteristics of multimedia presentations that have the potential to allevi
ate first impression bias. Based on this literature, a set of predictions w
as generated and tested through a laboratory experiment using a simulated m
ultimedia intranet.
Half of the 80 subjects were provided with a biased cue. Subjects were rand
omly assigned to four groups: (1) text with first impression bias cue, (2)
multimedia with first impression bias cue, (3) text without biased cue, and
(4) multimedia without biased cue. The experimental task involved conducti
ng a five-year performance appraisal of a department head. The first impres
sion bias cue was designed to provide incomplete and unfavorable informatio
n about the department head, but the information provided subsequently was
intended to be favorable of his performance.
Results show that the appraisal score of the text with biased cue group was
significantly lower than the text only (without biased cue) group. On the
other hand, the appraisal score of the multimedia with biased cue group was
not significantly different from the multimedia only (without biased cue)
group. As a whole, the results suggest that multimedia presentations, but n
ot tart-based presentations, reduce the influence of first impression bias.