Seeing Ourselves in Others:
Consumer Compliance with Recommendations Made by Ambiguous Agents
David A. Norton
Department of Marketing
The Moore School of Business
The University of South Carolina
1705 College Street
Columbia, SC 29208
E-mail: David_norton@moore.sc.edu
Phone: (803) 777-7305
Rebecca W. Naylor
Department of Marketing
The Moore School of Business
The University of South Carolina
1705 College Street
Columbia, SC 29208
E-mail: rnaylor@moore.sc.edu
Phone: (803) 777-7305
This research investigates how compliance with a recommendation made by another
consumer is affected by the amount of information provided about the reviewer and the
perceived similarity of the reviewer. In Study 1 we demonstrate that, given no information about
a reviewer (i.e., the reviewer’s characteristics are ambiguous), consumers project their own
characteristics onto the reviewer, resulting in compliance likelihoods that are similar to
compliance with a recommendation made by a reviewer presented as similar (vs. dissimilar) to
the target consumer. In Study 2 we demonstrate that the false consensus effect is responsible for
these results.
Consumers seeking recommendations on products and services have typically obtained
this information from two primary sources, professional agents who are paid to make such
recommendations (e.g., movie critics, interior decorators) (Solomon 1986; West 1996) or friends
and family members who provide this information through word-of-mouth communication
(Brown and Reingen 1987). However, Internet technology now offers consumers a new source
for product information: consumer-to-consumer recommendations in which no personal
information is provided about the reviewer (e.g., online user reviews that are posted with no
identifying information about the reviewer other than a user name).
The primary research question we seek to address is: how do consumers respond to
recommendations made by an ambiguous reviewer (whom we define as a reviewer about whom
no personal information is available)? Previous research on agent recommendations suggests that
giving consumers more information about a product recommender will increase compliance with
the recommendation if that recommender is demographically similar to the consumer (Feick and
Higie 1992). Gershoff and Johar (2006) demonstrate, for example, that consumers overestimate
an agent’s knowledge based on over-reliance on similarity cues in a close relationship. Hence,
past research suggests that reviewer ambiguity would decrease compliance with reviewer
recommendations since consumers will be unsure whether the reviewer is similar (which would
increase compliance) or dissimilar (which would decrease compliance) to themselves. However,
Norton, Frost, and Ariely (2007) argue that ambiguity can be beneficial in some contexts (e.g.,
potential dating partners are liked more when people have less information about them because
more information increases the likelihood of discovering dissimilarity between oneself and a
potential partner).
We propose that, when provided with no identifying information about a reviewer,
consumers project their own characteristics onto that reviewer in a manner consistent with the
false consensus effect (in which consumers tend to overestimate the prevalence of their own
attitudes in the general population; Ross, Greene, and House 1977), such that consumers respond
as positively to recommendations made by ambiguous reviewers as they do to recommendations
by reviewers that they know are similar to themselves. Essentially, we propose that when
information about reviewers’ tastes is left ambiguous, participants assume these reviewers share
their own tastes and attitudes.
To test our hypotheses, in Study 1 each participant read a restaurant review ostensibly
written by another consumer and then indicated their intentions to visit the restaurant they read
about. The review was designed to be slightly positive in order to make the restaurant attractive
but to avoid ceiling effects.The text of the review was held constant across all conditions, and
amount and type of reviewer information was manipulated such that participants saw either (1)
no information about the reviewer (ambiguous reviewer condition), (2) information about a
reviewer who was similar to the average student at their university (similar reviewer condition),
or (3) information about a reviewer who was different from the average student at their
university (dissimilar reviewer condition). In the similar and dissimilar reviewer conditions, the
participants saw nine units of demographic information (e.g. age, gender, university attended,
etc.) that were either similar or dissimilar to themselves (gender) or to the average student at
their university (age, university attended). After reading the review, participants rated, on a scale
of one (“not at all likely”) to nine (“extremely likely”), how likely they would be to try the
restaurant. Participants also rated how similar they thought the reviewer was to themselves in
terms of similarity in restaurant taste.
In order to analyze the results of Study 1, we used orthogonal contrast codes comparing
(1) the similar reviewer condition to the ambiguous reviewer condition and (2) the ambiguous
and similar reviewer conditions to the dissimilar reviewer condition. We found, as predicted, that
participants in the similar reviewer (M = 5.99) condition showed no difference in likelihood of
trying the recommended restaurant compared to participants in the ambiguous reviewer
condition (M = 6.00), F(1, 216) = 0.00, p > .95. However, compliance with the recommendation
was significantly lower for the dissimilar reviewer condition (M = 5.37) compared to the similar
and ambiguous reviewer conditions, F(1, 216) = 11.01, p < .01. In order to demonstrate that the
differences we find in restaurant trial are indeed caused by differing perceptions of similarity, we
compared participants’ ratings of similarity in the similar versus ambiguous reviewer conditions.
As expected, participants perceived the reviewer in the ambiguous condition to be equally
similar to themselves (M = 5.42) compared to the reviewer in the similar reviewer condition (M
= 5.59), supporting our false consensus explanation. Further, the difference in perceived
similarity between the dissimilar reviewer condition (M = 4.46) and the other two conditions
mediated the relationship between the dissimilar versus similar and ambiguous reviewer
conditions and likelihood of trying the restaurant, z = 3.93, p < .01.
The goal of Study 2 was to provide additional evidence that these results are driven by the
false consensus effect. We hypothesize that alerting participants to the potential bias of the effect
(i.e., warning participants about the tendency to see oneself in ambiguous others) decreases
compliance with an ambiguous reviewer’s recommendation. Hence, Study 2 used the same
restaurant review context and three-level (ambiguous reviewer vs. similar reviewer vs. dissimilar
reviewer) variable and added a two-level warning variable (such that participants were either
warned about the false consensus effect or not warned).
The results of Study 2 support our predictions; we found a significant interaction between
the warning variable and the similar reviewer versus ambiguous reviewer contrast, F(1, 286) =
2.90, p < .01. With no warning about the false consensus effect, the results of Study 1 were
replicated such that there was no difference in compliance with the recommendation between the
similar and ambiguous reviewer information condition (M = 6.17 for both conditions). However,
when participants received a warning about the potential impact of the false consensus effect,
they were significantly less likely to try the restaurant in the ambiguous reviewer condition (M =
5.18) than in the similar reviewer (M = 6.35) condition, thus providing additional evidence that
the false consensus effect drives our results.
References
Brown, Jacqueline Johnson and Peter H. Reingen (1987), “Social Ties and Word-of-Mouth
Referral Behavior,” Journal of Consumer Research, 14 (December), 350–362.
Feick, Lawrence and Robin A. Higie (1992), "The Effects of Preference Heterogeneity and
Source Characteristics on Ad Processing and Judgments about Endorsers." Journal of
Advertising, 21 (Summer), 9-24.
Gershoff, Andrew D. and Gita Venkataramani Johar (2006), “Do You Know Me? Consumer
Calibration of Friends’ Knowledge,” Journal of Consumer Research, 32 (March), 496503.
Norton, Michael I., Jeana H. Frost, and Dan Ariely, “Less Is More: The Lure of Ambiguity, or
Why Familiarity Breeds Contempt.” Journal of Personality and Social Psychology, Vol
92(1), Jan 2007. pp. 97-105.
Ross, Lee, David Greene, and Pamela House (1977), “The False Consensus Effect: An
Egocentric Bias in Social Perception and Attribution Processes,” Journal of Experimental
Social Psychology, 13 (May), 279–301.
Solomon, Michael R. (1986), “The Missing Link: Surrogate Consumers in the Marketing Chain,”
Journal of Marketing, 50 (October), 208–218.
West, Patricia M. (1996), “Predicting Preferences: An Examination of Agent Learning,” Journal
of Consumer Research, 23 (June), 68–80.