The Self-Regulation of Conformity: Mental Contrasting With Implementation Intentions (MCII)

von Weichs, Vivica; Krott, Nora Rebekka; Oettingen, Gabriele

doi:10.3389/fpsyg.2021.546178

ORIGINAL RESEARCH article

Front. Psychol., 02 June 2021

Sec. Personality and Social Psychology

Volume 12 - 2021 | https://doi.org/10.3389/fpsyg.2021.546178

The Self-Regulation of Conformity: Mental Contrasting With Implementation Intentions (MCII)

$\r\nVivica von Weichs$ Vivica von Weichs¹

Nora Rebekka Krott^2*

Gabriele Oettingen^3,4

¹Institute of Psychology, Helmut Schmidt University, Hamburg, Germany
²Institute for Interdisciplinary Research on Conflict & Violence (IKG), Bielefeld University, Bielefeld, Germany
³Department of Psychology, New York University, New York City, NY, United States
⁴University of Hamburg, Hamburg, Germany

The self-regulation of conformity has received little attention in previous research. This is surprising because group majorities can exert social strong pressure on people, leading them to overlook the pursuit of their own goals. We investigated if self-regulation by mental contrasting with implementation intentions (MCII) can reduce people’s tendency to conform and facilitate their own goal-pursuit despite deviant majority influence. In a computer-based logical reasoning task, we exposed participants to a conformity manipulation, where we presented bogus diagrams showing the supposedly correct answers of a majority ingroup. Compared to participants who were not given a self-regulation strategy (Studies 1, 2, and 4) or who were in an active control group (Study 3), MCII helped participants to self-regulate conforming behavior in trying to solve the task and to independently solve the logical reasoning task, as indicated by increases in correct answers in the task. The findings suggest that MCII is an effective strategy to regulate people’s tendency to conform and supports them to attain their goal despite deviant majority influence.

Introduction

Imagine it is quiz night and you and your friends reached the final round. You hear the question, ponder about how to solve the tricky problem, and have an answer in mind of which you think it is the right one. You are not a hundred percent sure, and all your friends insist on another answer. What will you do? Will you keep the answer you had in mind? Or will you conform and go with your friends choosing the other answer although you think it is wrong?

This example shows the conflict of a person who tries to perform correctly but also feels the urge to conform in order to find the right answer or to prevent exclusion from the group. Conformity has been defined as the tendency to change one’s attitudes or behaviors to match the responses of others (Cialdini and Goldstein, 2004; Kim and Hommel, 2015). People adapt their attitudes and behaviors to those of others when facing uncertain situations, thereby using other people’s judgments as their source of information about the “real” value of the object under consideration (informational influence; Deutsch and Gerard, 1955; Cialdini, 1993; see also Sherif, 1936; Festinger, 1954; Erb and Bohner, 2007). People also adapt their attitudes and behaviors to those of others in order to obtain social approval or to meet other people’s expectations, thereby avoiding sanctions or even punishment for being deviant (normative influence; Deutsch and Gerard, 1955; Cialdini and Goldstein, 2004; Baumeister et al., 2007). And they sometimes do so, even if they know that these behaviors are incorrect (Asch, 1955). Relatedly, people are more likely to conform to majority groups when they are viewed as an in-group (i.e., shared social identity; self-categorization theory; Turner, 1985, 1991) than when they are viewed as an out-group (i.e., no shared social identity; Turner, 1991; David and Turner, 1996, 2001; Crano, 2001; Erb and Bohner, 2007).

Consequences of Conformity

Conformity is an important force that keeps groups together and facilitates communication (e.g., Argyle, 1957; Hollander, 1958; LaTour and Manrai, 1989; Bond and Smith, 1996). However, the pressure to conform can also have detrimental effects for the individual. It can seduce people to concur with false information or to engage in risky behaviors (Asch, 1951, 1956; Santor et al., 2000; Lynn and Snyder, 2002; Gardner and Steinberg, 2005; Zhou et al., 2009). Conformity pressure may not only arise in the physical presence of others. Even in computer-mediated contexts people may feel the urge to follow the behavior of others (e.g., McKenna and Bargh, 1998; Postmes et al., 1999; Postmes and Spears, 2002; Riva, 2002; Lee, 2006; Cinnirella and Green, 2007; Lee et al., 2008; Schlosser, 2009; Laporte et al., 2010; Park and Feinberg, 2010; Bak and Keßler, 2012; Rosander and Eriksson, 2012; Anderson et al., 2014).

Conformity with the majority group can also cause people to disregard the successful pursuit of their own individual wishes or goals (e.g., Deutsch and Gerard, 1955; Asch, 1956). However, little research so far has focused on strategies aimed at reducing conforming behavior that may impede one’s own goal attainment. In the present studies, we tested whether the self-regulation strategy of mental contrasting with implementation intentions (MCII) would help people to regulate their tendency to conform, and thus support them in attaining their goals despite deviant majority influence.

Regulation of Conformity

Research on how to attenuate conformity points out that prompts protecting people’s self-esteem are effective in reducing susceptibility to other people’s opinions and perceptions (e.g., writing about a positive, stable inner trait leads to less conforming behavior, Arndt et al., 2002; see also Imhoff and Erb, 2009). While strategies affirming people’s self-esteem may be successful in reducing conforming behavior, they require an external prompt, such as feedback from others about one’s current behavior. But how can people protect themselves from submitting to majority opinions when those external prompts (e.g., by the researcher, educator, or any other authority) are not available or are not desired? Such regulating conformity is particularly difficult because people often conform without consciously realizing it and may even deny having been influenced by others (e.g., Chartrand and Bargh, 1999; Hornsey and Jetten, 2004, 2005; Bargh et al., 2012). Accordingly, people often fail to identify their tendency to conform to the majority opinion as an obstacle that holds them back from attaining their personal goals.

We introduce MCII to help people reduce their tendency to act in a conform way as MCII has been found to help people attain their goals even if they were not aware of the obstacle once it occurs. That is, MCII is a conscious imagery technique with non-conscious cognitive and motivational processes that enables people to surmount their obstacles and achieve behavior change outside of awareness (Kappes et al., 2012, 2013; Kappes and Oettingen, 2014). Applied to the present context, MCII should support people to strive for their individual goals in the face of a deviant majority even if they might not be aware of being tempted to give the majority’s answer once the majority’s answers are presented. Further, in contrast to previous studies, where participants wrote about a stable inner trait within a specific situation, MCII is – once it is learned by the individual – applicable to every situation in which people may be biased by the social influence of other people.

Mental Contrasting With Implementation Intentions (MCII)

Mental Contrasting

Fantasy realization theory (Oettingen, 2000, 2012, 2014) identifies mental contrasting as a self-regulation strategy that fosters effort and performance toward a desired and feasible future. During mental contrasting, people first positively fantasize about a desired future (e.g., finding the correct solution for a tricky problem) and then imagine the obstacle in the present reality (e.g., the urge to give in to your friends insisting on a different solution than yours). Imagining reaching the desired future followed by imagining the obstacle reveals that achieving one’s desired future requires acting on the obstacle of present reality (e.g., be daring, speak up). When the obstacle is surmountable, mental contrasting strengthens effort toward the desired future (e.g., to independently solve the tricky problem). Mental contrasting has been shown to foster effort and performance across different life domains, ages, and cultures, for short-term as well as long-term goals, and for different indicators of goal attainment (e.g., cognitive, emotional, and behavior; Oettingen et al., 2001; review by Oettingen, 2012; Oettingen and Sevincer, 2018).

With respect to non-conscious cognitive processes, mental contrasting helps people to detect obstacles of present reality standing in the way of attaining the desired future. It also changes the implicit meaning of reality as an obstacle that needs to be overcome. Further, mental contrasting strengthens the associative links between the desired future and the obstacle of reality as well as between the obstacle and the instrumental means to overcome the obstacle. Those cognitive processes mediate mental contrasting effects on effort and performance (Kappes et al., 2012, 2013; Kappes and Oettingen, 2014).

Fantasy realization theory identifies another mode of thought about the desired future: Reverse contrasting. When people engage in reverse contrasting, they elaborate on the same content as in mental contrasting, but in reverse order (i.e., the present reality followed by the desired future). Accordingly, the relational construct of reality standing in the way of the desired future is not activated and the reality is not interpreted as an obstacle (Kappes et al., 2013). Therefore, reverse contrasting fails to instigate behavior change (review by Oettingen, 2012).

Implementation Intentions

Mental contrasting has been combined with implementation intentions (MCII; Adriaanse et al., 2010; Kirk et al., 2013; Oettingen et al., 2015; review by Oettingen, 2012). Implementation intentions (Gollwitzer, 1993, 1999, 2014) are plans that prepare for goal pursuit by linking a predetermined action to a certain situation (“If situation X occurs, then I will perform behavior Y.”), thereby explicitly specifying when, where, and how one wants to act toward realizing one’s goal. They are differentiated from sheer goal intentions that come in the form of: “I want to perform behavior Y!”.

Implementation intentions (vs. sheer goal intentions) foster goal pursuit by making the mental representation of the situation specified in the if-part highly accessible (Gollwitzer, 1999) and by strengthening the associative links between this situation and the instrumental behavior specified in the then-part (Webb and Sheeran, 2007, 2008; review by Gollwitzer, 2014). Those associative links ensure that mental representations of the specified behavior are activated whenever the specified situation is encountered, leading to features of automaticity in goal directed behavior (i.e., immediacy, efficiency, no need of conscious intention, and autonomy; strategic automaticity, Gollwitzer and Brandstätter, 1997; Gollwitzer, 1999).

Implementation intentions (vs. goal intentions) have been shown to be effective across various life domains (reviews by Gollwitzer, 1999, 2014; Gollwitzer and Sheeran, 2006) but only when the conditions of goal pursuit are met, that is, when people are fully determined to reach their goal, and when the critical situation (if-part) as well as the goal-directed action (then-part) are clearly specified. Mental contrasting meets all three requirements. Specifically, it creates a strong determination or goal commitment, it highlights the critical situation (i.e., the obstacle of present reality), and it strengthens the implicit associative links between the obstacle and the behavior to overcome the obstacle (Kappes et al., 2012). Indeed, mental contrasting with implementation intentions (MCII) has been found to be more effective than either of the strategies alone (Adriaanse et al., 2010; Kirk et al., 2013).

The Present Research: Self-Regulation of Conformity Using MCII

Mental contrasting with implementation intentions captures three significant aspects relating to conforming behavior. First, by promoting the identification of the relevant obstacle (Kappes et al., 2013), MCII should allow people to anticipate situations where they might be giving in to the majority’s influence to then automatically resist the influence once it occurs. Second, by strengthening the associative link between the obstacle and the instrumental means to overcome the obstacle (Kappes et al., 2012), MCII should help people to carry out the pre-specified behaviors that help them overcome the obstacle (e.g., ignoring other people’s behavior), and third, by forming implementation intentions people should build particularly strong associative links between the obstacle and the behavior to overcome the obstacle. Thus, people should tightly stick to the goal-directed behavior they had previously decided upon once the obstacle occurs.

Pilot Study: Establishing the Paradigm

Because in the present research we recruited participants on Amazon Mechanical Turk (MTurk), we were able to introduce a respective shared social identity and an in-group context (i.e., the majority of MTurk participants). Specifically, we made social identity salient by introducing the community of MTurk participants as an in-group, whose results were allegedly compared to those of an out-group (i.e., bankers). Despite that identity of an MTurk worker is most likely not the central identity of our participants, making the in-group context of being an MTurk worker salient to our participants may additionally increase their tendency to conform to the behaviors of the group of MTurk participants.

In all studies, we report all measures, manipulations and exclusions. All sample sizes were determined before any data analysis. We first developed an on-line paradigm to induce conformity following a study design by Rosander and Eriksson (2012). We asked participants to answer logical reasoning items, which were taken from the Standard Progressive Matrices (Raven, 1965). In a pre-test with 50 participants, we assessed the difficulty level of eleven Raven matrices. Consulting the percentage of correct answers for each item, we chose eight matrices of varying difficulty, from the easiest item (96% of the participants answered correctly) to the most difficult item (40% of the participants answered correctly).

In this Pilot Study, we randomly assigned participants to a conformity condition or a control condition. While solving the Raven matrices, all participants were presented with diagrams showing how other MTurk participants had supposedly answered so far. Participants in the control group saw equally distributed answers for each item (Figure 1), participants in the conformity group saw that a majority choosing one particular incorrect answer (Figure 2)¹. In line with the results of Rosander and Eriksson (2012), participants in the conformity group (vs. control group) chose more often those incorrect answers that were indicated by the supposed majority (see the Supplementary Materials for methods and results; see also Riess, 2017).

FIGURE 1

Figure 1. Example of logical reasoning item (Item 4) with the fabricated diagram shown to the control condition. The Correct answer for the logical reasoning item (Item 4) is no 1.

FIGURE 2

Figure 2. Example of logical reasoning item (Item 4) with the fabricated diagram shown to the conformity condition.

Study 1: MCII vs. No Self-Regulation

In Study 1, we aimed to replicate findings from the Pilot Study, i.e., the induction of conformity within our computer-based paradigm. Specifically, we expected a priori contrasts to show that participants in the conformity group would more often choose those incorrect answers that were indicated by the supposed majority compared to participants in the control group. Second, and most importantly, we aimed to investigate whether MCII is an effective strategy to support people in regulating their tendency to conform and to help them reach their goal of independently solving the logical reasoning task. Specifically, we hypothesized that participants in the conformity condition using the strategy of MCII (vs. no self-regulation strategy) would give fewer answers complying with the incorrect answers of the supposed majority and give a greater number of correct answers.

Method Study 1

Participants and Power Analysis

Based on previous research regarding the induction of conformity in the context of CMC, we assumed a potential study drop out of up to 10% of the participants. We recruited 157 participants. A total of 20 participants were excluded because they indicated suspicion about the conformity manipulation (n = 15, for example, one participant wrote “I think that the other Mturk worker responses weren’t real.”) or did not follow the instruction of the self-regulation strategies (n = 5; for example, one participant wrote “Jumping up and down” when asked to elaborate on the outcome and obstacle)².

All participants were recruited on MTurk and received $2.00 for their participation. The final sample consisted of 137 participants, of which 70.1% were women. The participants’ ages ranged between 19 and 75 years with a mean of 39.46 years (SD = 13.89). All participants were randomly assigned to either a conformity or a control condition. Additionally, participants either received MCII or no self-regulation strategy, resulting in four conditions (see Table 1). A sensitivity power-analysis indicated that with 66 participants (within the two conformity conditions) and 71 participants (within the two control conditions) we would have 90% power to find, in each case, a large effect (d = 0.81 and d = 0.79).

TABLE 1

Table 1. Number of participants per condition.

Procedure and Materials

All participants completed the study online. Prior to their participation, they were informed about the procedure of the study and completed the consent form. After reading the cover story, participants were asked about their expectations, incentive and commitment to independently solve the upcoming logical reasoning task. They were then randomly assigned to the experimental conditions. Thereafter, participants completed the logical reasoning task and were finally asked for some demographic data. After completion of the study, all participants were thanked, debriefed and received credit for their participation in the survey.

Social identification

We presented participants the same cover story as in the pilot study, i.e., explaining that the survey was designed to compare cognitive abilities of people who occasionally deal with social psychology experiments (e.g., MTurk participants) and people who deal with economic problems only (e.g., bankers). We also asked participants to what extent they identified with the group of MTurk workers (see Supplementary Materials for more details on social identification).

Expectations, incentive, and commitment

After the cover story we presented participants an example of the logical-reasoning task. We asked them to think about independently solving the upcoming logical reasoning task: “Think about how nice it would be if you independently solved all of the following tasks successfully and could say to yourself: “Yes! I did it right!” Thereafter, we assessed participants’ incentive value (i.e., “How important is it to you that you will independently solve all of the following tasks successfully?”), expectations of success (i.e., “How likely do you think it is that you will independently solve all of the following tasks successfully?”) and commitment (i.e., “How disappointed would you feel if you did not independently solve all of the following tasks successfully?”). The Likert scales reached from 1 (not at all) to 7 (very), with high scores indicating high expectations, incentive, and commitment.

Manipulation of the Self-Regulation Strategy

Participants were randomly assigned to one of two experimental conditions. In the MCII condition, they were asked to write down one positive aspect they associate with independently solving the upcoming logical reasoning task. Participants read: “What would be the best thing if you independently solved all of the following tasks successfully and could say to yourself: “Yes! I did it right!”? What would be the most wonderful thing about it?” Participants wrote, for example, that they would feel good about themselves having the ability to independently complete the tasks successfully and that they would feel smart because they did the tasks by themselves. For example, one participant named “I would feel accomplished.” Then, participants were asked to imagine their positive aspect. Participants read: “Now take a moment and imagine your best outcome. Imagine things fully. Please write thoughts and images down.” For example, the same participant wrote “I would be proud of myself for being able to answer all of the questions on my own. It would be an accomplishment to be proud of.” Next, we introduced participants to the obstacle that might hinder them from independently succeeding in solving the logical reasoning task and asked them to imagine it. Participants read:

Sometimes things don’t work out as we would like them to. People tend to follow other people’s behavior when they are unsure of how to act. This can often lead to mistakes. Imagine yourself following the behavior of other people when solving the cognitive tasks. Imagine things fully. Please write your thoughts and images down.

Here, participants named, for example, that, especially when being insecure about how to behave, they would look at what other people are doing and emulate their actions even though they would be disappointed at themselves for not solving the test by themselves. For example, the participant cited above named “I would observe how others acted and try to mimic their behavior if they appeared to be answering the questions successfully.” Finally, participants were shown a plan, which we asked them to repeat, write down one more time, and remember whenever they would feel that they followed other people’s behavior during the task: “If I feel that I follow other people’s behavior, then I will tell myself: Ignore them!”

In the no self-regulation strategy condition, participants were immediately directed to the logical reasoning task after they had indicated their expectations, incentive and commitment to independently solve the logical reasoning task.

Results and Discussion Study 1

Expectations, Incentive, and Commitment

Mean values for expectations [M = 5.12, SD = 1.49; t(136) = 8.77, p < 0.001], incentive value [M = 6.07, SD = 1.10; t(136) = 5.86, p < 0.001], and for commitment [M = 4.84, SD = 1.68; t(136) = 22.01, p < 0.001] were above the midpoints of the scales. There were no significant differences between the conditions regarding the items (all p’s > 0.5, see Supplementary Materials).

Conformity on the Logical Reasoning Task

Identical to the Pilot Study, participants in the conformity condition were able to conform zero to five times. Again, we quantified conformity as the difference between the number of answers that agreed with the supposed majority answers for participants in the conformity condition and the number of the same incorrect answers for participants in the control condition.

We conducted a two-way ANOVA with conformity (conformity vs. control) and condition (MCII vs. no self-regulation) as between-subject factors and the number of conform answers on the task as dependent variable. There was a significant main effect of conformity, F(1,133) = 16.34, p < 0.001, d = 0.69, indicating that participants who were exposed to false majority answers more often chose the respective incorrect answers compared to participants who were not exposed to false majority answers. There also was a significant main effect of condition, F(1,133) = 5.04, p = 0.026, d = 0.38, with participants adopting MCII choosing the incorrect answer indicated by the majority less often compared to participants adopting no self-regulation strategy. The interaction between conformity and condition did not reach significance, F(1,133) = 2.75, p = 0.100, d = 0.20.

However, we exclusively aimed to investigate our first a priori hypothesis, i.e., the induction of conformity within the paradigm. We conducted planned contrasts (according to Furr and Rosenthal, 2003) for both conditions adopting no self-regulation strategy (control no self-regulation vs. conformity no self-regulation), thereby assigning a weight of 0 to both conditions using a self-regulation strategy. In line with our hypothesis, participants in the conformity no self-regulation condition more often chose the incorrect answer indicated by the supposed majority (M = 1.89, SD = 1.51) compared to participants in the control no self-regulation condition [M = 0.78, SD = 0.85), t(51.73) = 3.84, p < 0.001, d = 0.93]. Thus, we successfully induced conformity within the computer-based paradigm.

To test our second hypothesis, i.e., participants adopting MCII chose less often the incorrect answers indicated by the supposed majority, we conducted planned contrasts between both conformity conditions (conformity MCII vs. conformity no self-regulation), assigning a weight of 0 to both control conditions. In line with our hypothesis, participants in the conformity MCII condition chose the incorrect answer indicated by the supposed majority less often (M = 1.13, SD = 1.23) than participants in the conformity NSR condition (M = 1.89, SD = 1.51), t(63.78) = 2.22, p = 0.030, d = 0.55 (Figure 3). Specifically, participants in the conformity no self-regulation condition went along with the false majority 37.8% of the time; this effect was reduced to 22.6% in participants in the conformity MCII condition³.

FIGURE 3

Figure 3. Mean number of conform answers (dependent variable) for all conditions. Standard errors are represented by the error bars attached to each column.

Self-Regulation of Correct Behavior

We performed our analysis of correct behavior on the five critical items. Thus, participants could obtain a maximum of five correct answers on the task. We conducted a two-way ANOVA with conformity (conformity vs. control) and condition (MCII vs. no self-regulation) as between-subject factors and the number of correct answers on the task as dependent variable. There was no significant main effect of conformity, F(1,133) = 2.11, p = 0.149, d = 0.26. There was, however, a significant main effect of condition, F(1,133) = 7.27, p = 0.008, d = 0.46, indicating that participants who used MCII answered more items correctly compared with participants who did not use a self-regulation strategy. The interaction between conformity and condition was not significant, F(1,133) = 0.03, p = 0.856, d = 0.00.

However, to investigate our a priori hypothesis, i.e., participants in the conformity conditions would increase the number of correct answers on the task, we conducted planned contrasts for both conformity conditions (conformity MCII vs. conformity no self-regulation) on the number of correctly solved items, assigning a weight of 0 to both control conditions. There was a marginally significant difference between the two conditions; participants in the conformity MCII condition tended to give more correct answers on the task (M = 3.06, SD = 1.73) compared to those in the conformity no self-regulation condition (M = 2.34, SD = 1.57), t(133) = 1.76, p = 0.082, d = 0.30 (Figure 4).

FIGURE 4

Figure 4. Meta-analysis of the four studies. Forest plot, random effects model were calculated for the number of correct answers on the task as dependent variable, including condition as independent variable.

To conclude, even though the interaction effects did not reach significance, our main hypotheses focusing on comparing the effect of MCII (vs. not self-regulation) within the two conformity conditions were tentatively supported. Participants showed conformity in a computer mediated context and the findings speak for the observation that participants regulated their tendency to conform when engaging in MCII compared to participants engaging in no self-regulation strategy. Further, participants in the conformity condition tended to give more correct answers on the logical reasoning task when engaging in MCII compared to participants engaging in no self-regulation strategy. This marginally significant difference might indicate that MCII may have supported participants in attaining their goal of independently solving the task. The greater number of correct answers may even suggest that participants did not just blindly follow the instruction of ignoring other people’s behavior by choosing any other option than the majority. Rather, participants seemed to have more successfully completed the task on their own, though we need to interpret this result with caution as the interaction effect between conformity and the self-regulation condition was not significant.

In Study 2, we aimed to replicate our findings using a slightly modified version of the logical reasoning task. While we used logical reasoning items with increasing difficulty in the first study, we now used items with a medium level of difficulty (i.e., those items for which approximately 60–70% of the participants found the correct answer when working uninfluenced on the items), since they appeared most appropriate for the needs of our study. These items were easy enough for participants to answer them correctly when being focused, but still included some complexities so that participants would potentially turn over to the behavior of the majority to seek guidance for their choices.

Study 2: MCII vs. No Self-Regulation (Direct Replication)

We aimed to directly replicate the findings from Study 1.