Intuitive Conservatives and Analytical Liberals: Ideology and Dual-Processing Theory
University of Nebraska-Lincoln
Sample size: 583
Field period: 10/30/2012-1/30/2013
Are ideological inclinations affected by information processing styles? Dual-processing theory posits that people use two interrelated information processing systems. The first is fast and intuitive. The second is more deliberative and analytical, acting as an override of System 1-style processing. Recently Gervais and Norenzayan (2012) found that overriding intuitive thinking with System 2 processing led to less religious beliefs. Since political ideology is related to religiosity and traditionalism, as well as lower cognitive complexity (e.g. Brint and Abrutyn 2010; Eidelman et al. 2012), we predict attitudes will become more liberal when analytical thinking is primed and more conservative when intuitive thinking is primed. We propose to conduct an experiment using a modified verbal fluency task (Gervais and Norenzayan 2012; Uhlmann et al. 2011) to manipulate processing style. The primary dependent variables are political ideology and issue attitudes. A pilot study (n=189) demonstrated that more correct answers on the Cognitive Reflection Task (Frederick 2005)—indicating analytical thinking was overriding intuitive thinking—was correlated with liberal policy positions. By manipulating processing style we propose to test the causal order of the relationship between ideology and processing style, and investigate whether processing style can move attitudes in either liberal or conservative directions.
1. Subjects induced to think intuitively with a priming task will report more conservative political attitudes.
2. Subjects induced to think analytically with a priming task will report more liberal political attitudes.
3. Subjects induced to think intuitively with a priming task will report more conservative attitudes towards social issues but not economic ones.
4. Subjects induced to think analytically with a priming task will report more liberal attitudes towards social issues but not economic ones.
The project used a three condition, between group design with subjects randomly assigned to either the analytical, intuitive, or control condition. Each condition used a sentence completion task, which has previously been used to induce the target thought process within the treatment conditions (Gervais and Norenzayan 2012; Uhlmann et al. 2011). Subjects were given 10 groups of words, each consisting of five randomly ordered words. Subjects were instructed that they needed to drop one of the five words and type out a phrase using the remaining four words. All words in the control condition were neutral, with no relation to intuitive or analytical thinking. In the analytical condition half of the sentence sets included the words reason, ponder, think, rational, and analyze that function as the analytical prime. The intuitive condition received hunch, feels, instinct, intuition, and emotions in five of the ten word sets, priming intuitive thinking.
The focus of the project was to examine political attitudes following induction into analytical or intuitive thinking. We measured attitudes in two ways. First, subjects were asked to self-report their ideological stance on a 101-point scale ranging from very liberal to very conservative. Second, we measured political attitudes towards nine different policy issues. Three items concerned economic policies (welfare spendings, universal healthcare, and business regulation), three concerned moral issues (school prayer, abortion, and gay marriage), and three dealt with punishment issues (defense spending, torture of terrorist subjects, and the death penalty).
In addition to the main outcome variables, participants were asked about the extent to which they believed in a higher power to replicate the findings of Gervais and Norenzayan (2012). Finally, all subjects answered the three Cognitive Reflection Task (CRT; Frederick 2005), which is a widely used measure of the propensity to use analytical thinking that consists of three items. Each question has an easy and intuitive but incorrect answer, and a correct answer that involves analytical processing to reach. The CRT was used to test individual differences in the propensity to use analytical thinking in the control condition and as a manipulation check for the treatment conditions.
The main analysis used between-group ANOVAs to test whether the primes resulted in mean differences on political attitudes among subjects in the different conditions. There were no ideological mean differences using the self-reported measure. In addition, there were no mean differences across the groups for policy preferences, both for the aggregate and sub-issue item scales. Looking at the manipulation check, using the sum of the number of correct CRT answers provided by the subjects, the priming task did not sufficiently manipulate the targeted thought processes; there were no mean CRT score differences across the conditions.
The TESS study became a part of a larger project in which we looked at the relation between political attitudes and reflective versus intuitive thinking using four different studies. In all four studies, there was no evidence that priming reflective or intuitive thinking led to differences in political attitudes. In addition, the primes used in all four studies demonstrated an inability to induce the target treatment based on examination of mean differences in CRT scores across the conditions.
However, there was evidence across the four studies that there are individual differences in the extent to which someone is likely to be reflected and corresponding political attitudes. The overall pattern showed that higher CRT scores are related to liberal attitudes towards social issues. Across the four studies there were no consistent pattern regarding economic issues. However, looking at the TESS sample in particular, looking strictly at the control condition, there is a negative relationship between CRT scores and conservative social issues and no relationship with economic issues. Using all subjects within the sample regardless of treatment, there is a positive correlation between conservative economic issues and CRT scores while the negative relationship remains for social issues (Baron 2015).
Deppe, Kristen D., Gonzalez, Frank J., Neiman, Jayme L., Jacobs, Carly, Pahlke, Jackson, Smith, Kevin B., and Hibbing, John R. (2015). Reflective liberals and intuitive conservatives: A look at the Cognitive Reflection Test and ideology. Journal of Judgment and Decision Making, 10(4), 314-31.