Imagine being asked to defend a position you do not fully support. Perhaps a manager asks you to explain a new policy, a teacher assigns you the unpopular side of a debate, or a friend asks you to make the strongest case for a decision you would not choose. As you prepare, examples begin to surface. One argument connects to another. The position does not necessarily become right, but it may begin to feel more reasonable.

This is the central insight of biased scanning theory. The position we are preparing to defend can guide the information we retrieve, develop, and rehearse. Supporting considerations become mentally available, while competing considerations may receive less attention. The resulting imbalance can influence the person making the argument, not only the audience hearing it.

Biased scanning is sometimes described as another form of confirmation bias, but that description is incomplete. Confirmation bias usually begins with an existing belief or hypothesis. Biased scanning can begin with an assigned role or a temporary goal, including a position the person did not previously hold. The theory therefore belongs to a broader story about self-persuasion: what we say, write, and practice can help shape what we later find believable.

What Is Biased Scanning Theory

Biased scanning is a selective search through one’s available knowledge for material that supports a particular position. A task, incentive, expectation, or desired conclusion supplies the direction. Memory and imagination then provide examples, arguments, and anticipated consequences that make the position easier to defend (Petty & Cacioppo, 1986; Kunda, 1990).

The word biased does not imply conscious dishonesty. A person may sincerely experience the arguments that come to mind as the most relevant evidence. The imbalance lies in the search: compatible material receives more retrieval and rehearsal, while contrary material is less likely to be generated or developed.

The historical research used role-playing tasks, but the mechanism is not limited to acting. It may be engaged whenever people prepare to advocate, explain, justify, or sell a conclusion. The effect is also bounded. A person cannot necessarily persuade themselves of anything they choose; the shift depends on what reasons can be found, how credible those reasons seem, and how thoroughly the position is considered (Janis & King, 1954; King & Janis, 1956).

How Biased Scanning Shapes Belief

From an Assigned Position to Self Persuasion

The process begins when a direction is established. Someone may be asked to advocate a conclusion, may want a decision to appear justified, or may anticipate having to explain a position to others. That direction changes the question being asked. Instead of asking, What does all of the evidence suggest, the person begins asking, What supports this case?

Once a person begins building a case, supporting examples are more likely to come to mind. They may also create new reasons, put them into familiar language, and connect them with personal experience. When the issue comes up again, those recently practiced reasons are easier to remember than alternatives that received little attention.

Attitude change can follow because the mental evidence now available is uneven. The person is not necessarily ignoring a complete record placed before them. More often, the record itself has been reconstructed through selective retrieval.

Why Self-Generated Arguments Matter

Arguments we generate ourselves can feel more personally fitted than arguments we receive from someone else. They use our vocabulary, draw from our memories, and answer the objections we find most important. King and Janis proposed that improvisation may reduce resistance because an externally supplied message begins to feel partly like one’s own reasoning (King & Janis, 1956).

Later persuasion research placed this effect within the broader importance of elaboration. Petty and Cacioppo described role-playing as a persuasion situation that requires substantial issue-relevant thinking. They also reviewed evidence that self-generated messages can produce more persistent change than passively received messages, even when the initial degree of change is similar (Petty & Cacioppo, 1986).

This does not make self-generated arguments more accurate. It makes them more available, meaningful, and resistant to forgetting. Fluency can then be mistaken for evidentiary strength: the position feels well supported partly because the person has practiced supporting it.

Historical Role-Playing Studies and Self-Persuasion

The Original Role-Playing Study

Irving Janis and Bert King tested whether active advocacy could influence the advocate’s private opinion. Male college students first answered questions about several future events. About four weeks later, groups of three students encountered persuasive communications advocating estimates more extreme than any they had originally given. Each student presented one communication as a sincere advocate and was passively exposed to the other two (Janis & King, 1954).

For two of the three topics, active participants shifted significantly more than passive controls. For the third, the groups showed similar opinion change, although active participation increased confidence among some students whose estimates moved in the advocated direction. The authors concluded cautiously that overt verbalization could augment a persuasive communication. They did not claim that every act of role-playing changes belief.

The study also revealed an important clue. The stronger effects appeared where speakers improvised more and felt more satisfied with their performance. Because those observations were correlational, the authors treated improvisation and satisfaction as competing explanations requiring a more direct test.

The Improvisation Follow-Up

King and Janis then compared three conditions using a communication about military service. One group of 32 students improvised a talk after reading the message. A second group of 23 read the prepared script aloud without improvising. A passive group of 20 read the message silently. The oral-reading group felt more satisfied with its performance, but the improvisation group showed the greatest opinion change (King & Janis, 1956).

On the combined measure, 87 percent of the improvisation group changed on at least three of five opinion items in the advocated direction, compared with 65 percent of the non-improvisation group and 65 percent of passive controls. The non-improvisation group did not differ significantly from the passive group. Experimentally increasing or decreasing satisfaction within the improvisation condition also failed to alter opinion change.

These findings favored the inventive work of improvisation over mere repetition or satisfaction. Participants were not simply saying a message aloud. They were generating additions, illustrations, and appeals that helped tailor the case to their own cognitive world. Even so, the experiment involved small groups of male undergraduates, a consequential military-service message, and immediate attitude measures. Those conditions make the study historically important without making it a universal law.

Biased Scanning and Related Psychological Processes

Confirmation Bias

Confirmation bias is the broader tendency to seek, interpret, or remember evidence in ways that favor an existing belief, expectation, or hypothesis. It can be motivated by a valued conclusion, but it can also arise from ordinary cognitive habits, such as testing only one possibility at a time or finding positive instances easier to imagine than disconfirming ones (Nickerson, 1998).

Biased scanning overlaps with confirmation bias because both create an uneven evidence sample. The difference is the starting point. Confirmation bias commonly protects a belief already in place. Biased scanning can be directed by a role or assignment and can move thought toward a position the person did not initially endorse.

Motivated Reasoning

Motivated reasoning explains how goals influence the cognitive processes used to reach a conclusion. Ziva Kunda argued that directional goals affect which beliefs, memories, and inferential rules become accessible. People can maintain an illusion of objectivity because they experience the supporting reasons they retrieve, not the unselected alternatives that might have appeared under a different goal (Kunda, 1990).

This framework helps generalize biased scanning beyond formal role-play. Defending a purchase, explaining a relationship decision, or protecting a political identity can all create a direction for memory search. Kunda also supplied an essential boundary: desired conclusions remain constrained by the person’s ability to construct a justification that appears reasonable.

Cognitive Dissonance and Self-Perception

Cognitive dissonance offers a different account of attitude change. Advocating a position that conflicts with one’s private belief may create an uncomfortable inconsistency. Changing the private attitude is one way to restore coherence (Festinger, 1957). Self-perception theory offers another possibility: people may infer something about their own attitude by observing what they have said or done.

These explanations focus on different parts of the experience. Biased scanning asks which reasons the person generated. Dissonance asks whether inconsistency created pressure to change. Self-perception asks what the person inferred from seeing their own behavior. More than one process may be present, so a change in attitude does not by itself tell us which explanation is correct (Triandis, 1971; Vaidis et al., 2024).

What Modern Evidence Adds

A large 2024 multilab study offers a useful modern boundary. Nearly 4,900 participants from 39 laboratories in 19 countries completed an induced-compliance task involving an unpopular university policy. Some wrote a counterattitudinal essay after being reminded that they were free to refuse, some received lower-choice instructions, and a comparison group wrote about neutral policies (Vaidis et al., 2024).

The result is easier to state than to explain. People who wrote against their initial view shifted slightly toward the position they had argued, and they also reported more discomfort and conflict. Yet having more choice about writing the essay did not add the attitude change predicted by a classic version of dissonance theory. Generating the argument mattered; the reason it mattered remained uncertain (Vaidis et al., 2024).

The authors considered several explanations, including discomfort from inconsistency, inferences drawn from one’s own behavior, and self-persuasion through generating reasons. The study keeps biased scanning plausible, but it does not establish that biased scanning is always the mechanism whenever advocacy changes an attitude.

Biased Scanning in Everyday Life

The original experiments used speeches and assigned positions, but similar demands occur in ordinary life. A worker may repeatedly defend a policy adopted by the organization. A student may prepare one side of a debate. A social-media user may spend an evening composing replies for a political position. A person explaining a difficult choice to family members may become increasingly practiced at retrieving only its advantages.

These examples should be treated as applications of the mechanism, not diagnoses of the people involved. Advocacy does not prove insincerity, and repeated explanation does not guarantee attitude change. The important observation is that producing a case changes which considerations receive cognitive work. Over time, that work can alter what feels obvious, memorable, or personally convincing.

Biased scanning may also contribute to rationalization after a decision. Once a course has been chosen, generating reasons for it can stabilize action and reduce uncertainty. That stability may be useful when endless reconsideration would prevent effective action. It becomes costly when the same one-sided search protects a failing decision from relevant correction.

How Biased Scanning Can Broaden Perspective

Selective scanning is not inevitably a defense of the status quo. An assignment to advocate an unfamiliar position may make previously neglected evidence available. Triandis’s review of role-playing research described findings in which people expecting to defend their own position evaluated supporting material selectively, while those expecting to advocate the opposing position became more evenhanded in judging arguments from both sides (Triandis, 1971).

This possibility gives the mechanism a constructive use. Deliberately developing the strongest case for another perspective can interrupt the search habits attached to one’s current view. The exercise works best when it is a genuine attempt to understand the alternative rather than a performance designed to make it look foolish.

The deeper question is therefore not whether thought is influenced by a direction. Human reasoning nearly always begins somewhere. The practical question is whether the direction helps us encounter neglected evidence or merely rehearses what we already prefer to believe.

Limits and Common Misunderstandings

Biased scanning is not a claim that people believe everything they repeat. The early studies found different effects across messages, and later work shows that argument quality, prior knowledge, personal relevance, and the ease of generating reasons can matter. A person who cannot construct a plausible case may remain unconvinced.

The mechanism also does not establish that the resulting belief is false. A selective search can retrieve accurate evidence. The problem is that the available sample may not represent the full body of relevant information. Confidence may rise because supporting reasons are accessible, even when important counterevidence was never considered.

Nor should biased scanning become a label for every disagreement. People can reach different conclusions because they possess different information, assign different values to the same outcomes, or reasonably interpret uncertain evidence in different ways. The concept is most useful when it identifies a process that can be examined, not when it substitutes for engaging another person’s argument.

Finally, the classic studies reflect the methods and samples of their period. Their small, all-male college samples and artificial speaking tasks limit broad generalization. Modern evidence supports continued interest in self-generated advocacy while cautioning against overconfidence about why and when it changes attitudes.

How to Encourage a More Balanced Search

Awareness is helpful, but an instruction to be objective may not be enough. Kunda reviewed evidence that people often generate support for their first answer spontaneously, whereas confidence declines when they are required to produce reasons against it. Nickerson likewise concluded that reasoning can improve when people use practices associated with critical thinking: identifying alternative hypotheses early and asking what evidence would show the favored conclusion to be wrong (Kunda, 1990; Nickerson, 1998).

A balanced search can begin with a few structured questions. What is the strongest reason against the conclusion I am preparing to defend? What evidence would change my mind? Which facts became easy to recall only after I adopted this role? Am I comparing the quality of evidence on both sides, or merely counting the arguments I generated for one side?

It can also help to separate advocacy from evaluation. A team assigned to build a proposal might ask a different person or group to test its assumptions. An individual using a structured decision-making process can write the best case for each realistic alternative before announcing a decision. The aim is not perfect neutrality. It is to prevent the demands of making a case from silently becoming the whole process of deciding what is true.

A Few Words by Psychology Fanatic

We often experience our reasons as discoveries. An example comes to mind, another connects with it, and soon the conclusion seems to have emerged naturally from the evidence. Biased scanning reminds us that the role we occupy may have helped determine which evidence became available in the first place.

This does not require distrusting every conviction or refusing to advocate for anything. Speaking and acting with commitment are part of a meaningful life. The invitation is more modest: before treating the fluency of our reasons as proof, we can ask what another role, another question, or another honest search might have brought to mind.

Our beliefs are shaped not only by the information around us but by the arguments we practice producing. Understanding that process gives us a better chance to advocate with conviction while remaining open to correction.

Associated Concepts

• Confirmation Bias: The broader tendency to seek or interpret evidence in ways that favor an existing belief, expectation, or hypothesis.
• Motivated Reasoning: The influence of goals and desired conclusions on which memories, beliefs, and reasoning strategies become accessible.
• Cognitive Dissonance: The aversive pressure that can arise when psychologically important beliefs, behavior, or commitments do not fit comfortably together.
• Self-Perception Theory: The proposal that people sometimes infer their own attitudes by observing their behavior and the circumstances in which it occurred.
• Elaboration Likelihood Model: A persuasion model that explains how the amount and direction of issue-relevant thinking affect attitude change.
• Rationalization: A plausible explanation that can protect a choice or self-image by emphasizing supportive reasons while leaving threatening evidence less examined.

References

Festinger, Leon (1957). A Theory of Cognitive Dissonance. Stanford University Press. ISBN: 978-0-8047-0911-8.
(Return to Main Text)

Janis, Irving L.; King, Bert T. (1954). The Influence of Role Playing on Opinion Change. The Journal of Abnormal and Social Psychology, 49(2), 211–218. DOI: 10.1037/h0056957.
(Return to Main Text)

King, Bert T.; Janis, Irving L. (1956). Comparison of the Effectiveness of Improvised Versus Non-Improvised Role-Playing in Producing Opinion Changes. Human Relations, 9(2), 177–186. DOI: 10.1177/001872675600900202.
(Return to Main Text)

Kunda, Ziva (1990). The Case for Motivated Reasoning. Psychological Bulletin, 108(3), 480–498. DOI: 10.1037/0033-2909.108.3.480.
(Return to Main Text)

Nickerson, Raymond S. (1998). Confirmation Bias: A Ubiquitous Phenomenon in Many Guises. Review of General Psychology, 2(2), 175–220. DOI: 10.1037/1089-2680.2.2.175.
(Return to Main Text)

Petty, Richard E.; Cacioppo, John T. (1986). Communication and Persuasion: Central and Peripheral Routes to Attitude Change. Springer-Verlag. DOI: 10.1007/978-1-4612-4964-1.
(Return to Main Text)

Triandis, Harry C. (1971). Attitude and Attitude Change. John Wiley & Sons.
(Return to Main Text)

Vaidis, David C.; Sleegers, Willem W. A.; van Leeuwen, Florian; DeMarree, Kenneth G.; Sætrevik, Bjørn; Ross, Robert M.; Schmidt, Kathleen; Protzko, John; Morvinski, Coby; Ghasemi, Omid; Roberts, Andrew J.; Stone, Jeff; Bran, Alexandre; Gourdon-Kanhukamwe, Amélie; Gunsoy, Ceren; Moussaoui, Lisa S.; Smith, Andrew R.; Nugier, Armelle; Fayant, Marie-Pierre; … Priolo, Daniel (2024). A Multilab Replication of the Induced-Compliance Paradigm of Cognitive Dissonance. Advances in Methods and Practices in Psychological Science, 7(1), 1–26. DOI: 10.1177/25152459231213375.
(Return to Main Text)