Shaping Reality: How Attitudes are Built and Deconstructed
Attitudes form the content of our mental models, defining our ideology, values, and aspirations. Since attitudes predict behavior, they are integral to our identities and actions. One of the most basic mechanisms of attitude formation is the mere exposure effect, demonstrated by Robert Zajonc in 1968. Zajonc exposed participants to nonsense characters for different durations and found that people tended to like the characters that had been presented for longer. This robust effect suggests we like things we are familiar with because familiarity translates to predictability, supporting the social mind’s need to build a stable model of the world.
Year Zajonc demonstrated the mere exposure effect
Attitudes can also be formed through associative learning, a primitive mechanism shared with animals. In one experiment, researchers paired nationality labels like “Dutch” or “Swedish” with words carrying positive or negative connotations; participants subsequently felt more positive or negative toward the associated nationality. Beyond these passive learning mechanisms, attitudes can be adopted consciously to fulfill functions, such as the utilitarian function (expressing attitudes to facilitate relationships and fit in with groups) or the knowledge function (predicting the world). Attitudes can also serve an ego-defensive function, such as blaming immigrants for unemployment to protect one’s self-esteem.
The Paradox of Attitudes and Actions
Despite attitudes being the “basic building blocks” of our mental models, early research raised serious concerns about their predictive utility. In 1934, Richard LaPierre traveled across the US with a Chinese couple during a time of high prejudice toward East Asians. Only one out of 250 establishments refused to serve them. Yet, months later, when LaPierre surveyed these same establishments by telephone, 90% stated they would refuse to serve a Chinese couple. This shocking discrepancy suggested that attitudes might not predict behavior.
of establishments claimed they would refuse service to Chinese patrons in LaPierre's survey
Subsequent work revealed that attitudes do predict behavior, but only under specific circumstances. Factors influencing the link include the level of specificity (LaPierre asked about specific people versus Chinese people in general) and the role of social desirability (public politeness overriding private prejudice). The time elapsed between measuring the attitude and the behavior is also critical, as the social mind is a dynamic system whose attitudes constantly change. This research led to the theory of planned behaviour, which posits that attitudes, subjective norms (what significant others think one should do), and perceived control combine to predict behavioral intention, which then predicts behavior.
When Behavior Changes Belief
The relationship between attitudes and actions is sometimes reversed: behavior can cause attitude change. Self-perception theory, developed by Daryl Bem, suggests we infer our attitudes by observing our own past behavior, functioning as a type of attribution. For example, if someone notices they consistently recycle, they may infer, “I must be environmentally friendly”. This primarily occurs when strong attitudes are not already present. This process is so powerful that simply being asked to hold a pen in one’s teeth (forcing a smile) leads people to rate cartoons more positively, misattributing the cause of the forced expression to the content.
In contrast, when a person performs behavior counter to a strongly held attitude, they experience cognitive dissonance, a psychological discomfort developed by Leon Festinger in 1957. To resolve this uncomfortable feeling, since the past behavior cannot be undone, the person changes their attitude to align with the action. Festinger and Carlsmith demonstrated this in 1959 by having participants perform a boring task, then asking them to lie to the next participant, claiming the task was enjoyable. Those who lied without sufficient external justification (a small or no monetary reward) subsequently reported actually enjoying the task, changing their attitude to justify their prior deception.
Payment that led to attitude change in Festinger's cognitive dissonance experiment
The Dual Routes of Persuasion and Group Norms
Persuasion, defined as attitude change resulting from intended external influence, operates via the elaboration-likelihood model (ELM). This dual-process model suggests people are persuaded through either the central route, involving systematic and analytic processing (the “naïve scientist”), or the peripheral route, relying on quick and easy cues (the “cognitive miser”), such as the attractiveness of the source. Which route is taken depends on motivation (e.g., issue involvement) and available time. Importantly, attitudes formed via the peripheral route are weaker and less resistant to counter-argument.
Beyond targeted persuasion, attitudes are profoundly affected by implicit social influence. Musaf Sherif’s 1935 study on norm development, utilizing the autokinetic effect (a stationary light dot appearing to move in a dark room), demonstrated this. When participants made estimates alone, their answers varied, but when doing the task with others, their estimates converged on a common social norm, often without the participants realizing the influence exerted. This is informational influence, where others’ attitudes are used as a rule of thumb when uncertainty is high, leading to changes in both public and private attitudes (conversion).
However, when the answer is obvious, normative influence—the desire simply to fit in—takes over, changing only public attitudes (compliance). Solomon Asch’s 1951 conformity study proved this. Participants were asked to match lines, an objective task that was easy when performed alone (99% accuracy). When placed with confederates who gave a clearly incorrect answer, participants conformed on 37% of the trials. This compulsion to avoid standing out is so strong that brain scans show the physical pain center is activated upon social exclusion. Conformity dramatically decreases when even a single person provides social support, breaking the group consensus. Conversely, consistent minority influence can stimulate the majority to engage in greater systematic thought (elaboration), sometimes leading to conversion and high-quality decisions.
of participants conformed to incorrect answers in Asch's line judgment experiment