The Quiet Pull: Teaching the Psychology beneath Group Polarization

My mom, a French teacher, had brought a group of students to tour France. At 15, I got to go and I was trying to absorb everything. I even bought a French newspaper, Le Monde, because I knew it was widely read by French intellectuals. Carrying it into a cafe, an older man seating us asked me in French, “Are you going to read that?” I answered, also in French, “I’m going to try.” He launched into a sharp critique of Americans - how we’re uncouth, uneducated, uninterested in the world. I remember standing there, stunned. Even when presented with a counterexample - an American teenager speaking his language and curious about his culture - his views didn’t shift. How can people hold such strong entrenched views? That moment stays with me, and comes back to me, whenever I teach about group polarization – the tendency for us to become more extreme in our views when we get together with like-minded people.

“Group polarization” was first coined by Moscovici and Zavalloni (1969), who found when French students discussed a topic, their views shifted further in the direction they already leaned: more positive toward Charles de Gaulle and more negative toward Americans. Just by talking with people who shared their general view, participants became more certain - more polarized.

This dynamic isn’t limited to national identity. Psychology documents it across domains. People tend to become more risk-seeking or more cautious depending on their group’s initial leanings (Isenberg, 1986). Groups with similar political views become more ideologically consistent after discussion (Schkade & Sunstein, 2003). High-prejudice groups become more prejudiced; low-prejudice groups become more inclusive (Myers & Bishop, 1970). Juries tend to deliver verdicts matching their majority when they begin deliberating (Kalven & Zeisel, 1966). Even student identity shifts - like joining a fraternity or sorority – can deepen members’ shared beliefs years to come (Pascarella & Terenzini, 1991).

More recently, we study how these same psychological forces are amplified by social media. Algorithms prioritize engagement, and among the easiest ways to generate engagement is reinforcing what we already believe (Cinelli et al., 2021; Bail, 2021). But the engine beneath the algorithm is us - our own tendency to lean further when we’re surrounded by agreement.

And we don’t notice it happening. People believe their views are unaffected by group discussion despite data consistently showing measurable shifts (Keating et al., 2016). I teach group polarization not just as content, but as a mirror. Once students recognize the quiet tug, they can begin noticing when they’re leaning further than they meant to.

To help students experience group polarization before they learn about it, I run a three-part activity in the class session prior to introducing the concept. The design is simple but creates a powerful moment of awareness.

1. Pre-Survey

At the start of class, students complete a short, intentionally broad, survey: What are people like? Directions emphasize it’s about your personal views, right now - not what you’ve heard, not what you maybe used to think. There are no right answers. I collect them before we continue.

2. Group Discussion

Partway through class, at a natural break, students engage in a 7-minute small-group discussion. The prompt is open-ended: Can people change? Really change? I group students sitting near friends or naturally clustered in the room, because these small social cues often mirror similarity. Each group is assigned a number, and their handout includes spaces for names.

I have tried and fumbled with so many different variations of this activity, semester after semester. Among the hardest parts was choosing the right topic - something rich enough to spark honest conversation, but not so loaded it shut people down. I learned political topics didn’t work well in my setting; most of my students lean left, so we rarely saw shifts both ideological directions. And anything requiring specialized knowledge just left some students out. What I needed was a question where everyone had stories, everyone had doubts, and no one could predict what their group would say. The prompt Can people change? Really change? finally opened that space. It gave students a way to talk about themselves or people they love, and those conversations always go somewhere.

3. Post-Survey

At the end of class, students take the same questionnaire again, with the same instructions. The repetition is intentional - not to prime a change - but to highlight stability. I remind students: “Please answer based on your own view, right now.” This minimizes demand characteristics and helps ensure any shift is authentic.

Between classes, I analyze the data. Each group’s pre- and post-discussion responses are averaged, allowing us to explore how conversation shapes subtle shifts.

The questionnaire is carefully balanced (figure 1). Half the items are reverse-scored, and they fall into three subcategories:

• 6 neutral items about change (i.e., generically people can change or not)
• 6 positively valenced items (e.g., people can grow, improve)
• 6 negatively valenced items (e.g., people will always be selfish or deteriorate)

From this, I calculate two scores:

• “Can people change?” – based on neutral items
• “People are good.” – based on valenced items

In the next class, I show scatterplots with each group plotted by their initial average (x-axis) and post-discussion shift (y-axis). We include prior semesters, with our current class’s groups clearly labelled (figure 2)

The data tell a quiet but meaningful story:

• For “Can people change?”, initial average predicts change: r = .23, p = .17, N = 37. Not statistically significant, but suggestive.
• For “People are good”, the effect is stronger and significant: r = .52, p < .001, N = 37.
• A multiple regression model predicting the 175 individual post-discussion attitudes from their pre-survey score and group average (excluding the student) shows: pre-scores explain 50% of variance (p < .0005), while the group added 2% of variance (p < .007)

Regression model:
Post = 0.81 × Pre + 0.33 × GroupPre − 0.75

I frame the change as small but meaningful. This was just 7 minutes, with peers, on a topic they may not feel deeply about. Imagine what happens over weeks, months, or years - on issues tied more closely to identity, morality, belonging, or personal history.

What makes the moment powerful isn’t the effect size. It’s the realization: we shifted. Even a little. Even when we didn’t mean to. Because each group is numbered and labeled, our class discussion becomes more than just data. It becomes a space for group reflection. One group might share how their conversation centered on stories of growth, therapy, and redemption. Another might recall examples of broken trust, addiction, or people becoming hardened over time. What students begin to see is not just how their group shifted, but what shaped the tone of the shift itself. That’s the entry point into group polarization - not as an abstraction, but as something we now feel.

By now, many students are nodding. My students tend to lean left, so my character resonates.

I quietly return to center.

I pause, walk slowly back to the middle, and say:

My classroom goes still.Students recognize their own voice in one side. They hear the same passion and same language – on the other side.The shift is in our hearts, but also in our heads. Students see groups of like-minded people don’t just reinforce belief. They curate knowledge. We rarely bring counter-evidence to conversations where everyone already agrees.

That’s the persuasive arguments theory in action. Not abstract, but embodied. My performance isn’t about pushing students away from conviction, but inviting them into curiosity about how conviction is built.

Another mechanism behind group polarization is social comparison - our tendency to evaluate ourselves in relation to others, especially those within our own group. Once we know what our group values, we often adjust our attitudes or behaviors to match, or even exceed, those norms. Psychology calls this a normative influence: we don’t just want to be right, we want to be good (Lamm & Myers, 1978). But the standards of “goodness” aren’t fixed. They shift with our reference group.

I share a personal example because storytelling by instructors aids student memory, learning, and engagement (Grobman, 2015).

That’s what social comparison can do. It doesn’t just make us lean toward our group; it can make us question whether we belong at all.

When I started using this example in my teaching, students connected with it. But when I moved to California and began teaching at my current university, something shifted. Students didn’t resonate with the discomfort I felt at that potluck. In fact, many said things like, “I wish I were a vegetarian.” I was accidentally triggering admiration, not recognition. I realized in my first classrooms - teaching in the Deep South - students saw vegetarianism as an out-group trait, so how my experience of in-group not-good-enoughness made intuitive sense to them because they could contrast me with themselves. But in a different environment, I was being perceived as more virtuous than they were.

So I tried something new.

I started playing the country music video “Rednecker” by Hardy - a kind of tongue-in-cheek anthem of redneck one-upmanship:

My town's smaller than your town
And I got a bigger buck and bass on my wall
...
You might think that you're redneck
But I'm rednecker than you

Suddenly, it clicks. The same psychology applies whether your group values are progressive or traditional. We want to be seen as “good” by our group, whatever that means in our particular social context. For one group, maybe it’s sustainability, compassion, and progress. For another, it’s grit, toughness, and tradition. Regardless, the pull is the same: we adjust our self-presentation, and sometimes even our views, to fit in - or stand out - in the direction our group already leans.

That’s the social comparison theory in action. Not so much about vanity or conformity, but about belonging. It’s not always the group that moves us. It’s our sense of who we are within it.

Suddenly, it clicks. The same psychology applies whether your group values are progressive or traditional. We want to be seen as “good” by our group, whatever that means in our particular social context. For one group, maybe it’s sustainability, compassion, and progress. For another, it’s grit, toughness, and tradition. Regardless, the pull is the same: we adjust our self-presentation, and sometimes even our views, to fit in - or stand out - in the direction our group already leans.

That’s the social comparison theory in action. Not so much about vanity or conformity, but about belonging. It’s not always the group that moves us. It’s our sense of who we are within it.

If persuasive arguments show how knowledge accumulates, and social comparison shows how identity shapes belief, then among our real question is: how do we teach in a way making room for reflection - without asking students to surrender conviction?

It’s a delicate balance. We could mistakenly fall into “both-sides-ism” - the idea every viewpoint deserves equal time, or every perspective is equally grounded in truth or dignity. Some issues are asymmetric. Sometimes the extreme at one pole is way more extreme, harmful, and unwarranted than the opposite pole. At the same time, when polarization deepens, it can lead not merely conformity and a lack of critical thinking, but to real-world consequences - dehumanization, radicalization, even violence

I try not to teach group polarization as simply a spectacle (look how extreme people get!), but as a mirror. A way we can ask ourselves: when do I feel most certain? When do I silence myself and even change to fit in? If we can help our students notice when they’re leaning further, without demanding they lean less, we’re teaching psychology in a way letting them stay curious, even when it’s easier to shut down.

When designing your class coverage of group polarization, please feel free to use any of the materials I created.

• Classroom Demonstration Materials (PDF & above): contains pre- & post-discussion surveys with directions, as well as scoring information and a group discussion handout. Everything you need to replicate the activity.
• Dramatization Examples (above): Real-time classroom narratives embodying the persuasive arguments and social comparison accounts in ways students can feel.
• Breakout Discussion Prompts (above): A set of reflection questions designed to help students explore how polarization shows up in themselves and their conversations.
• Assignable Reading (PDF): A student-friendly 10 minute read providing an overview of the psychology beneath group polarization.
• Annotated Bibliography (PDF): Organized by theme to help faculty select studies for lecture or discussion.
• Conference Presentation Slides (PDF forthcoming): Shared at the 2025 Psychology One Conference, outlining activities and insights.
• Faculty Roundtable Handout (PDF): Companion to the breakout discussion session I facilitated at the same conference -designed to support faculty reflection.
• Abbreviated Teaching Guide (PDF draft awaiting review): An invited contribution to a faculty compendium summarizing my approach to teaching group polarization.

Bail, C. A. (2021). Breaking the social media prism: How to make our platforms less polarizing. Princeton University Press.

Bruneau, E., Kteily, N., & Laustsen, L. (2018). The unique effects of empathy and perspective-taking on intergroup attitudes. Journal of Experimental Social Psychology, 76, 393–403.

Broockman, D. E., & Kalla, J. L. (2016). Durably reducing transphobia: A field experiment on door-to-door canvassing. Science, 352(6282), 220–224.

Cinelli, M., Morales, G. D. F., Galeazzi, A., Quattrociocchi, W., & Starnini, M. (2021). The echo chamber effect on social media. Proceedings of the National Academy of Sciences, 118(9), e2023301118.

Grobman, K. H. (2015). Antsy students impatient to leave class and faculty captive in NPR driveway moments? Enhancing science classes with personal stories. In K.Brakke & J. Houska (Eds.) Telling stories: The art &science of storytelling as an instructional strategy, (pp.98-115), Washington, DC: American Psychological Association Society for the Teaching of Psychology, Retrieved from http://teachpsych.org/ebooks/index.php

Isenberg, D. J. (1986). Group polarization: A critical review and meta-analysis. Journal of Personality and Social Psychology, 50(6), 1141–1151.

Kalven, H., & Zeisel, H. (1966). The American jury. University of Chicago Press.

Lamm, H., & Myers, D. G. (1978). Group-induced polarization of attitudes and behavior. In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 11, pp. 145–195). Academic Press.

Moscovici, S., & Zavalloni, M. (1969). The group as a polarizer of attitudes. Journal of Personality and Social Psychology, 12(2), 125–135.

Myers, D. G., & Bishop, G. D. (1970). Discussion effects on racial attitudes. Science, 169(3947), 778–779.

Myers, D. G., & Lamm, H. (1976). The group polarization phenomenon. Psychological Bulletin, 83(4), 602–627.

Pascarella, E. T., & Terenzini, P. T. (1991). How college affects students: Findings and insights from twenty years of research. Jossey-Bass.

Keating, C. F., Van Boven, L., & Judd, C. M. (2016). Partisan underestimation of the influence of group discussion on political attitudes. Journal of Experimental Social Psychology, 65, 110–117.

Reicher, S., Haslam, S. A., & Rath, R. (2008). Making a virtue of evil: A five-step social identity model of the development of collective hate. Social and Personality Psychology Compass, 2(3), 1313–1344.

Schkade, D., & Sunstein, C. R. (2003). Deliberating about dollars: The severity of punitive damages and the role of group polarization. Columbia Law Review, 102(6), 1149–1216.

Tausch, N., & Becker, J. C. (2013). Emotional reactions to successful and unsuccessful collective action: Group-based anger and pride mediate mobilization via perceived efficacy. Journal of Personality and Social Psychology, 105(4), 641–660.

van Prooijen, J.-W., & Krouwel, A. P. M. (2019). Psychological features of extreme political ideologies. Current Directions in Psychological Science, 28(2), 159–163.

Vinokur, A., & Burnstein, E. (1974). Effects of partially shared persuasive arguments on group-induced shifts: A group-problem-solving approach. Journal of Personality and Social Psychology, 29(3), 305–315.

Voelkel, J. G., Brandt, M. J., & Willer, R. (2023). Interventions reducing partisan animosity: A meta-analysis of experimental research. Nature Human Behaviour, 7(1), 146–159.