When Attitudes Predict Behaviour

People often expect attitudes to behave like promises.

If someone cares about the environment, they should always make the environmentally preferable choice. If they value exercise, they should train. If they disapprove of dishonesty, they should never lie. When behaviour diverges from a stated attitude, observers may conclude that the person did not really hold it.

Sometimes people do misrepresent what they think. But inconsistency alone cannot reveal its cause. It may reflect a mismatch between what was measured and what the person had to decide, competition from another behavioural process or the absence of a feasible opportunity.

Attitudes can predict behaviour. They do so under particular conditions, not by operating like instructions that the body must obey.

What an Attitude Is

Definitions vary across research traditions, but this Essay uses attitude to mean an evaluative organisation concerning an object, person, behaviour, issue or idea.

An attitude may be favourable or unfavourable. It may also contain competing positive and negative evaluations. Someone might appreciate the convenience of driving while disliking its environmental consequences.

Attitudes are related to beliefs without being identical to them. A person can believe that exercise reduces health risks and still dislike exercising. Beliefs about likely consequences can contribute to an attitude, but the evaluation remains a distinct form of organisation. This distinction is developed in beliefs and attitudes are not the same.

An attitude is also not a value, goal or intention. Someone may evaluate public transport positively without forming a goal to use it today. They may hold that goal without deciding when to travel. They may intend to take the bus and still fail to do so.

An evaluation therefore does not specify one inevitable action. A favourable attitude toward environmental protection could be expressed through transport, purchasing, energy use, voting or public advocacy. One attitude can support several behaviours, while one behaviour may reflect several attitudes and goals.

The relevant question is not merely whether someone has an attitude. It is whether the measured evaluation corresponds to the behaviour and becomes influential under the conditions in which that behaviour occurs.

Prediction Depends on What Is Being Compared

Suppose someone reports strong support for environmental protection and later drives on one particular journey. The measures overlap in topic but differ in their level and object of evaluation.

“Environmental protection is good” evaluates a broad issue. “Taking the bus for this journey tomorrow would be worthwhile” evaluates a particular action in a particular context and time. The second attitude corresponds more closely to the behaviour being predicted.

Attitude–behaviour prediction generally improves when the measures correspond in:

• the action being performed;
• the target of the action;
• the context in which it occurs;
• the relevant period.

The broad attitude may still contribute to the decision, but the journey also depends on cost, timing, destination, available routes and other goals. Weak prediction does not necessarily reveal a psychological contradiction. It may reveal poor correspondence between the evaluation measured and the action observed.

Specific attitudes toward performing a behaviour therefore tend to predict that behaviour better than broad attitudes toward the object involved. Greater correspondence improves prediction; it does not prove that the attitude directly caused the action.

Broad Attitudes Predict Patterns Better Than Single Acts

A broad attitude is often better matched to a broad behavioural pattern than to one isolated act.

Someone who supports environmental protection may drive when they need to transport equipment, then use public transport on another day, reduce household energy use and avoid unnecessary purchases. One journey provides limited evidence about the wider evaluation.

When several relevant behaviours are considered together, circumstances unique to one act exert less influence over the overall pattern. A broad environmental attitude may therefore predict a representative collection of environmental behaviours better than tomorrow’s transport choice.

That aggregate must be defined independently and genuinely represent the domain. Researchers could otherwise inflate consistency by selecting only behaviours that already correlate with the attitude. Aggregation can also conceal meaningful differences among situations. A stronger relationship with the overall pattern does not mean that the same attitude governed every component action.

Nor does a group-level relationship determine individuals. Even well-constructed behavioural patterns provide probabilistic prediction, not certainty about what one person will do next.

An Attitude Must Be Available in the Moment

An attitude is more likely to shape interpretation and choice when the relevant evaluation becomes accessible in the situation.

Accessibility concerns how readily an evaluation can be retrieved or activated. Researchers often use response speed as one indicator, although response latency is not identical to accessibility and is affected by the task used to measure it.

Repeatedly activated or recently used evaluations may become easier to retrieve. Direct experience can also strengthen the association between an object and its evaluation. When an attitude is accessible, it can influence how information is interpreted and which options appear favourable.

When it is less accessible, other features of the situation may exert greater influence. A visually prominent product, an immediate reward or a salient social cue may dominate the moment.

Accessibility is not importance. An attitude can matter deeply without being active during one decision. Another can be retrieved quickly because it was recently repeated without being personally consequential.

A significant attitude can therefore remain part of someone’s psychological organisation even when another concern shapes the behaviour observed.

Strong Attitudes Are More Than Extreme Opinions

Two people can choose the same extreme point on an attitude scale without holding equally strong attitudes in every relevant sense.

One evaluation may be confident, stable, accessible and grounded in extensive experience. Another may be recent, uncertain and highly sensitive to context. The identical rating does not establish identical behavioural influence.

Attitude strength can involve:

• confidence;
• stability;
• accessibility;
• personal importance;
• relevant knowledge;
• resistance to change;
• consistency among evaluative responses.

These properties can coexist, but they do not always move together. An attitude may be personally important yet unstable, or highly accessible but weakly informed.

Direct experience may strengthen prediction by increasing confidence, accessibility and knowledge about how the attitude relates to action. Someone who regularly uses public transport knows more about its practical consequences than someone evaluating it through general information alone.

Yet direct experience does not guarantee consistency. A familiar action can remain unavailable or conflict with another goal. Repeated indirect exposure can also increase accessibility and confidence.

Time creates further uncertainty. The longer the interval between measuring an attitude and observing behaviour, the more opportunity there is for new information, changed goals and altered circumstances. An earlier report may have been sincere even if it no longer describes the evaluation active at the moment of action.

The causal relationship can also run in both directions. Attitudes may contribute to behaviour, while behaviour and its consequences reshape later attitudes. Observed consistency does not prove that a fixed evaluation came first and independently produced every matching act.

A Favourable Attitude Is Not a Goal or Intention

Someone can like exercise without holding a current goal to exercise. They can hold that goal without intending to train today. They can form the intention without deciding when or how to act.

Attitudes toward a behaviour can contribute to intention formation, but so can perceived norms and perceived control. A favourable evaluation makes an intention more likely under some conditions; it does not constitute the intention.

Current goals help determine when an attitude becomes relevant. As explained in goals organise behaviour, goals change which information and possible actions matter now. A positive attitude toward exercise may become highly relevant when the active goal concerns health or athletic performance. It may have less influence when the immediate goal is recovering from illness or responding to an urgent family demand.

The relationships are reciprocal rather than strictly linear. Attitudes can influence which goals people endorse, goals can make particular attitudes action-relevant, and experience can alter both.

Even a strong intention does not execute behaviour. A person may need to represent a workable route by simulating possible action in advance. The intended action must then become accessible at the appropriate moment, which depends on remembering to act at the right time. The broader transition from commitment to execution is a separate problem: intention is not action.

Attitudes operate within this larger system. They do not independently complete the route to action.

Habits Can Preserve Behaviour After Evaluations Change

When behaviour is repeated under stable conditions, features of the context can become associated with the response. Entering a familiar situation may then make that behaviour easier to initiate without requiring a fresh evaluation of every alternative.

Habits can align with a current attitude or pull behaviour in another direction. Someone may develop a negative attitude toward checking their phone during focused work while continuing to reach for it in familiar situations. The changed evaluation does not immediately remove the learned relationship between context and response.

As examined in what makes behaviour habitual, habitual behaviour reflects learning rather than an inner agent taking control. How context triggers habitual action explains how familiar conditions reactivate established responses.

Habit is not irresistible. Strong intentions, changed environments and deliberate control can alter behaviour. Past behaviour is also not a pure measure of habit: repetition may reflect stable preferences, skills and opportunities.

Still, learned contextual control places an important boundary around attitude prediction. Behaviour can remain stable after evaluation changes, and continued performance of the old response does not prove continued endorsement.

Norms Change the Consequences of Expression

Attitudes are formed, measured and expressed in social environments.

Injunctive norms concern what people believe others approve or expect. Descriptive norms concern what they believe others commonly do. Both can influence intentions, make actions appear normal and alter the costs of behaving differently.

A person may privately approve of something while believing that public expression would be condemned. They may privately disapprove while complying with a role expectation. Behaviour then reflects anticipated consequences as well as personal evaluation.

Social environments can also help create and stabilise attitudes. Repeated exposure to what others approve, condemn or routinely perform can shape evaluation itself, not merely determine whether an existing attitude is expressed. The way norms make behaviour feel natural belongs to that broader normative organisation.

Norms also affect attitude reports. Someone may conceal an evaluation because disclosure feels unsafe or provide the response they believe is expected. But disagreement between report and behaviour does not automatically demonstrate deception. The measures may concern different evaluations, the context may have changed or the person may hold ambivalent attitudes that one response cannot capture.

Private evaluation, reported attitude and public behaviour are related forms of evidence. They are not interchangeable readings of one simple inner state.

Attitude-Consistent Action Requires a Feasible Opportunity

Someone may support public transport while living where no usable route exists. They may value healthy food while lacking affordable access, storage or time to prepare it. An employee may disapprove of an organisational practice but be unable to challenge it safely.

Perceived behavioural control concerns whether someone believes they can perform an action. That belief can influence intention and persistence. Actual control concerns whether conditions make the action feasible.

The two can diverge. Someone may underestimate an available opportunity or confidently intend an action that later becomes impossible.

Feasibility is rarely all or nothing. An action may be technically possible but prohibitively expensive, socially dangerous or available only at unsuitable times. Time, money, transport, health, disability, caregiving, permission, institutional rules and environmental safety all influence the available action set.

These conditions are not minor additions applied after motivation has done its work. They are part of the causal system through which attitudes can become behaviourally relevant. Environments contain behaviour: they make some responses accessible while rendering others costly or unavailable.

Opportunity is not motivation. But a particular attitude-consistent action requires sufficient opportunity to occur.

Explicit and Implicit Measures Do Not Reveal Two Simple Selves

Explicit attitude measures ask people to report evaluations directly. The response reflects how the question is interpreted, which evaluative information becomes accessible and what the person selects or is willing to express.

An explicit report can provide meaningful predictive information without being a complete, context-free reading of the person.

Indirect measures infer associations from performance on tasks such as the Implicit Association Test. Many produce relative scores: the result reflects associations among compared categories rather than an absolute evaluation of one category in isolation.

These measures can contain predictive information, but average relationships with behavioural outcomes are generally modest and vary across tasks, criteria and research designs. Reliability and task construction further constrain interpretation.

Disagreement between explicit and implicit measures does not reveal two simple selves—a public attitude and a hidden true one. The measures may capture different associations, use different comparisons or respond differently to context. An implicit association can exist without conscious endorsement, while a sincere explicit report need not capture every learned association.

Predictive association at the sample level does not make one score a reliable individual diagnosis. Neither measurement class determines what a person will do.

Neural activity associated with task performance does not solve this problem. A neural correlate is not the attitude’s content, and physiological evidence should not be silently classified as behaviour.

One Action Cannot Reveal the Whole Attitude

Taking a bus could reflect environmental concern, financial necessity, habit or lack of access to a car. Driving could reflect preference, time pressure, disability or the need to transport equipment.

The action provides evidence without announcing which process produced it.

Attitude-consistent behaviour may involve intention, habit, social reward or ease of opportunity. Inconsistency may involve poor measurement correspondence, changed evaluation, another goal, learned contextual control or restricted feasibility.

Consistency does not prove one causal pathway. Inconsistency does not independently establish insincerity or hypocrisy.

This does not make attitudes irrelevant. It explains why their predictive power depends on the wider system in which behaviour occurs.

Attitudes Influence Behaviour Without Commanding It

Attitudes predict behaviour most strongly when the evaluation is specific enough to correspond with the action, remains accessible and stable, and becomes relevant within a current goal and context. Prediction improves when habitual control does not preserve another response, social conditions permit expression and an appropriate action is feasible.

Even then, the relationship remains probabilistic.

An attitude changes how an object or action is evaluated. That evaluation can shape attention, intention and selection, but it operates alongside goals, learning, norms, capacity and opportunity.

One inconsistent act does not expose a person’s “real” attitude, just as one consistent act does not reveal the entire pathway that produced it.

Attitudes alter behavioural probability. They do not command action.