Abstract: From the birther movement to the push of “alternative facts” from the White House, recent events have highlighted the prominence of misinformation in the U.S. This study seeks to broaden our understanding of under what conditions factual misperceptions may be effectively corrected. Specifically, I use Social Identity Theory to argue that ingroup members, specifically co-partisans and peers, are perceived to be more credible, and in turn are more effective correctors, than outgroup members (out-partisans and elites), contingent on identity strength. I also argue that peers should be effective correctors among those with low levels of institutional trust. To test my expectations, this study employs a 2 x 2 experimental design with a control group to determine how successful various source cues are at changing factual beliefs about a hotly debated topic in the U.S.— immigration. Overall I find preliminary support for my expectations.
Discussion
The experiments in this paper help us understand how factual beliefs about politics
can be changed by manipulating the source of the corrections. I find that responses to
corrections from peers and co-partisans differ significantly according to subjects’ group
identity and trust in institutions. As a result, the corrections about immigration are most
successful among those that strongly identify with the group the source is a member of,
whether a co-partisan or peer.
My findings contribute to the literature on correcting misperceptions in several respects.
First, while prior corrections research has exclusively used a between-subjects design to
circumvent the possibility that people feel grounded in their responses before receiving
correction to after (Nyhan & Reifler, 2010), this study employed a within-subjects design
in order to establish how subjects actually change their factual beliefs. Indeed, my findings
demonstrate that substantial movement in reported factual beliefs does happen following
corrections and that these beliefs typically hold over time.
Second, the results from this study corroborate previous work demonstrating that the
impact of source cues on opinion varies systematically by individual identification with the
source (Hartman and Weber 2009). Among those who strongly identify with their party,
inparty corrections are successful in moving factual beliefs in the accurate direction while
outparty corrections lead to a “backfire” effect, pushing individuals in the opposite direction. On the other hand, weak partisans are receptive to corrective information from outparty
members. Therefore, while the literature on the impact of ideological source cues on correcting misperceptions has offered inconsistent findings (Nyhan and Reifler 2013; Berinsky
2015), the results in this paper demonstrate that not all Republicans and Democrats respond
uniformly to corrections—strength of group identification matters. While I theorize that
perceived credibility of ingroup and outgroup members is the causal mechanism at work, it
would be valuable to directly study how individuals evaluate various sources on key characteristics (trustworthiness, knowledge, etc.,). This would allow social scientists to gain a
better understanding of specifically why certain sources are successful and others are not.
While research on source cues in political science has largely focused on partisan cues,
these findings also contribute to our understanding of how voters respond to information
from elites and peers. In accordance with my expectations, I find some evidence that peers
are more successful correctors than elites, especially among those who strongly identify
with their peer group and among those have weak trust in institutions. These findings add
to existing work (Attwell and Freeman 2015) seeking to understand how social groups can
promote accuracy in cases when experts or elites appear ineffective. Future work should
further explore how voters evaluate the credibility of elites and peers differently, and what
other individual-level factors might explain why certain people are more receptive to political information from peers than elites. It would also be valuable to replicate these findings
on peer groups other than university students or use real peers that are not fabricated. Lastly,
while peer cues in this experiment did not have the substantial impact on factual beliefs that
were expected across all four statements, they should not be concluded as irrelevant. Walsh
(2004) illustrates the importance of face-to-face interactions among small peer groups in
political thinking. It is possible that factual corrections in the context of these sort of peer
interactions are effective, and scholars should aim to understand the significance of such
interactions in the real world.
Of course, I am mindful of the inherent limitations of the evidence presented here. The
most serious limitation to this experiment was the scale used to measure the dependent
variable, which confounds confidence and acceptance/rejection of false statements. For this
reason it is difficult to untangle the differences between changes in acceptance or rejection
of beliefs and actual confidence in beliefs. For example, it could be possible that a peer
correction could just make subjects more confident in their already (correct) beliefs, and not
actually encourage a switch from acceptance to rejection of a false statement. Subsequent
studies should unpack these distinctions.
There is also the question over whether individuals’ reported factual beliefs are in fact
sincere and not just expressive partisan cheerleading (Bullock, Gerber and Seth 2015). I
argue that sincerity is largely inconsequential here. If subjects are willing to report strong
confidence in falsehoods in a survey, this cannot be completely irrelevant to the way they
perceive the political world. Lastly, if misperceptions are successfully corrected, it is no
guarantee that subjects’ respective political attitudes are actually moved in a certain direction. As Gaines, Kuklinski, Quirk, Peyton and Verkuilen (2007) note, those who “update
their beliefs accordingly need not imply they update their opinions accordingly [emphasis
added]” (p. 971). However, because factual beliefs and attitudes are so conflated, measuring
the two together in a questionnaire might have discouraged individuals from accurately updating because they are reminded that certain beliefs are inconsistent with their worldview.
The purpose of this study is only to examine the conditions under which strong beliefs in
falsehoods (not issue positions) can be effectively challenged, but future work should explore how relevant factual beliefs shape opinions on immigration. While prior work has
found no evidence that correcting factual beliefs about immigrant population sizes leads
to attitude change (Lawrence and Sides 2014; Hopkins, Sides and Citrin 2016), these are
likely not the only factual beliefs that inform voters’ respective attitudes.
Posts
No comments:
Post a Comment