Two-Step Flow Theory
The message travels. People give it weight.
Explore how media meet personal networks. Learn what opinion leaders contribute, why roles change with the topic, and how to distinguish a shared message from a changed judgment.
In this chapter · Open the contents
Explain the route—and the role of the relationship.
- Describe the model’s research foundations and classic two-step route.
- Explain topic-specific opinion leadership without equating it with fame.
- Distinguish information flow, interpretation, reinforcement and behaviour.
- Compare direct, two-step and multi-step communication.
- Propose evidence and alternative explanations for a claim of influence.
A message reaches you. A conversation helps you decide.
You encounter a review of a new documentary. Later, a friend who follows cinema explains why it matters and challenges one of the reviewer’s claims. You reconsider your first impression. The review has travelled, but something else has happened: a relationship has helped shape its meaning and weight.
Two-Step Flow Theory proposes that media messages often reach people through opinion leaders who attend to relevant media and influence others in their personal networks. The classic model connects mass communication with interpersonal communication instead of treating audiences as isolated recipients.
The crucial move is to place media inside social life. Family discussions, friendships, professional contacts and informal groups can shape which messages people notice, how they interpret them and what they do next.
Two steps name a simplified route
The familiar route is media → opinion leader → other people. The two steps refer to communication relationships. They do not mean two apps, two screens, two days or exactly two contacts in every real situation.
A person can also encounter media directly, speak with several advisers, question a recommendation or influence the adviser in return. The model highlights a route worth investigating; it does not describe every possible flow.
Who helps whom interpret a message or make a judgment, on which topic, and with what evidence of influence? Trace the route and the difference the exchange may have made.
Foundations: Lazarsfeld, Berelson and Gaudet (1944); Katz and Lazarsfeld (1955); Katz (1957). All contemporary examples in this chapter are fictional teaching scenarios, not reports of measured effects.
From an election study to the study of personal influence.
The model grew from empirical research into media, decisions and social relationships. Keep the study dates and publication dates distinct.
| Study / work | What it contributed | Remember |
|---|---|---|
| The 1940 Erie County election study | Paul F. Lazarsfeld, Bernard Berelson and Hazel Gaudet examined opinion formation during the US presidential campaign in Erie County, Ohio. Repeated interviews helped track changes over time. | The research setting was a particular campaign and community, not every media situation. |
| The People’s Choice · 1944 | The first edition reported the election research and helped direct attention to personal contacts and social groups in political opinion formation. | 1940 is the study year; 1944 is the first publication year. Later editions followed. |
| Personal Influence · 1955 | Elihu Katz and Paul F. Lazarsfeld developed the argument through research in Decatur, Illinois, including everyday decisions about marketing, fashion, public affairs and movie-going. | This was a separate research setting, extending the discussion beyond the election study. |
| Katz’s review · 1957 | Katz reviewed the two-step-flow hypothesis and its development as a research problem. | A model becomes more useful when its assumptions and limits remain open to investigation. |
What changed in the understanding of media effects?
The research challenged a picture of media acting on socially isolated individuals. It drew attention to prior commitments, group relationships and interpersonal discussion. It also helped establish the importance of studying people repeatedly rather than relying only on a single interview.
This work is commonly placed within the limited-effects tradition. Here, “limited” should prompt questions about conditions, kinds of effect and competing influences. It should not be turned into a universal claim that media barely matter.
Credit Lazarsfeld, Berelson and Gaudet for The People’s Choice, and Katz and Lazarsfeld for Personal Influence. The 1955 book was not simply the publication of the 1940 election study.
The original 1944 edition is identified in Indiana University’s catalogue. The ICPSR record for the Erie County data describes the study and its variables.
Two relationships connect media with personal judgment.
Reports, programmes and other relevant material enter public circulation.
People attend to, select and interpret material within a topic of interest.
Others encounter ideas and advice through discussion in their networks.
Step 1: media become material for a conversation
A person interested in a topic follows relevant reports, compares accounts or notices a new development. This gives that person material to discuss. Media exposure alone, however, does not make someone an opinion leader: the relational question is whether others turn to that person or are influenced through the exchange.
Step 2: a personal relationship can reshape the message
The intermediary may summarise the information, explain its relevance, endorse it, criticise it or connect it with a local concern. The second step is therefore more than a delivery service. Conversation can alter emphasis and meaning.
But the person receiving advice can compare it with other sources, ask questions or reject it. The traditional label opinion follower describes a role in a particular relationship; it does not establish passivity or permanent dependence.
A message can pass through two people without changing anyone’s judgment. To claim interpersonal influence, examine what the exchange contributed beyond the fact that it occurred.
A trusted peer can matter more than a prominent name.
An opinion leader is someone whose informal influence matters to other people in a relevant area. The role is relational and often topic-specific. A person can guide a group’s choices about films while asking another member for advice about research software.
Opinion leadership does not require formal office, wealth, celebrity or a large online following. Those features may provide visibility or authority, but they do not establish the kind of interpersonal influence highlighted by the classic model.
| Feature | Why it may matter | What to examine |
|---|---|---|
| Interest and relevant media use | Someone who follows a topic can notice and compare information. | Which sources are used, how carefully, and on what topic? |
| Social contact | Conversation creates opportunities to give and receive advice. | Who actually discusses the issue with this person? |
| Perceived competence | Others may value practical knowledge or experience. | Do contacts regard the person as useful in this specific area? |
| Trust and accessibility | Advice may carry weight when a person is approachable and familiar. | What makes the relationship credible, and are claims still checked? |
| Recognition by others | Influence concerns a relationship, not only self-confidence. | Do other people identify this person as an adviser, and can they describe an example? |
Roles can switch
Imagine that Asha guides Kabir through a digital archive because she has used it extensively. Later, Kabir helps Asha evaluate a camera because he knows the equipment. Each can lead in one discussion and seek guidance in another.
Even within the same topic, a person may consult one contact for factual explanation, another for practical experience and a third for a recommendation. Treating a community as two fixed classes—permanent leaders and permanent followers—would flatten these relationships.
Is a popular creator automatically an opinion leader?
No. A creator’s public reach can be studied as mediated communication. Classic opinion leadership emphasises influence within personal networks. A creator may also act as an adviser in particular communities, but the relationship and effect need evidence.
Follower counts, views and likes measure aspects of attention or connection. They do not by themselves show that an audience trusts the person on a topic or changes a decision because of the advice.
Read Personal Influence and Katz’s 1957 review alongside later network research. The characteristics above organise a teaching inquiry; they are not a scale or a diagnostic instrument.
People interpret, compare, reassure and challenge.
Why might a conversation matter after a person has already seen a report? It can supply something the report did not: an explanation suited to the listener, an answer to a question or an account grounded in shared experience.
Interpretation
An adviser translates unfamiliar terms, separates options and explains how a general message relates to a particular decision.
Trust and experience
A known person can describe what happened in practice. That experience may carry weight, though it can still be mistaken or incomplete.
Group discussion
Members compare views, question claims and signal expectations. Agreement, disagreement and pressure can all enter the exchange.
Feedback
The listener can ask for clarification or offer new information. This can change the adviser’s interpretation as well.
Reinforcement can be an effect too
A discussion may confirm an existing preference rather than reverse it. It may make an uncertain view firmer, bring a neglected option into consideration or change the reasons a person gives for a choice. Define the outcome before deciding whether influence occurred.
Do not assume that personal influence always produces better decisions. Familiarity can help explanation, but it can also carry an error quickly through a trusted network. Trust in a person and accuracy of a claim remain separate questions.
“My friend sent me the review”, “my friend helped me understand it” and “my friend persuaded me to choose differently” describe different claims. Each needs evidence suited to it.
What does “limited effects” allow us to say?
Media influence can depend on relationships, prior views, the topic and the decision. A person may act on a direct announcement, consult an adviser about an unfamiliar choice, and ignore another message entirely.
The chapter therefore treats “interpersonal communication matters” as a proposition to examine in context. It does not infer that personal contacts are always more powerful than media or that every mediated effect is weak.
Follow the actual route, including the people who remain quiet.
The classic model draws attention to a neglected link. A complete account must also consider direct communication, multiple exchanges, reciprocal influence and the possibility that a person does not join the discussion.
A person encounters and responds to a message without an identified interpersonal intermediary.
An intermediary helps select, explain or evaluate the message for another person.
Several exchanges contribute to circulation and interpretation. Direct exposure can occur alongside them.
Three limits of the simple picture
- Direct access: people who seek advice may also consult the original media. A discussion does not prove that the adviser was their first or only source.
- Several steps: advisers can learn from other people, and recipients can become intermediaries themselves. Some initial information may come from observation or personal experience rather than media.
- Quiet participants and non-participants: some people read without discussing; others may not encounter the issue at all. Silence does not establish exposure, agreement or lack of influence.
A group can also contain disagreement between advisers. Asking who influences whom may reveal a shifting set of relationships rather than a single leader addressing a uniform audience.
Why two apps do not prove a two-step flow
Watching a video and then opening the linked article involves two platforms but may involve no interpersonal influence. Conversely, an adviser can read a newspaper and discuss it face to face: that fits the classic communication route without using two digital platforms.
Count meaningful relationships and exchanges, not interface changes.
A repost shows a route. Influence remains a further question.
Digital traces can help researchers observe circulation, but they do not make every aspect of influence visible. A share can mark endorsement, disagreement, humour or an invitation to check a claim. A private conversation may matter greatly while leaving no public trace.
Separate the originator of material, its intermediaries, its audiences and the outcomes studied. The person who creates a post need not be the person whose interpretation carries most weight in a group.
What research has added
Wu, Hofman, Mason and Watts (2011) examined the production and circulation of information on Twitter and found support for intermediary roles in that setting. Choi (2015), studying two Twitter-based political discussion groups in South Korea, also found explanatory value in the two-step-flow model.
These are findings about particular datasets and contexts. They do not establish a permanent law of every platform, nor do circulation patterns alone establish changes in belief or conduct.
Why the biggest account may not start the largest cascade
Watts and Dodds (2007) used computer simulations to examine the claim that a small group of highly influential people drives large-scale diffusion. Under many conditions they modelled, the wider network and its susceptibility mattered more than exceptional individuals.
This does not mean opinion leaders never matter. It cautions against assuming that identifying a prominent individual is enough to explain how a message spreads across a whole network. Simulation findings also depend on the assumptions built into the model.
Direct targeting can coexist with social mediation
Bennett and Manheim (2006) discuss a later one-step flow involving increasingly differentiated messages directed towards individuals. This is a specific argument about communication and social change, not simply another name for a crude theory that media affect everyone identically.
A sound contemporary study can look for direct targeting, interpersonal advice and multi-step circulation together. Ask which route matters for the topic and outcome, rather than declaring one model the winner before collecting evidence.
When a digital study reports a flow, ask what was observed: following, reposting, discussing, recommending or changing a judgment. Those measures answer different questions.
Choose the question before choosing the theory.
| Approach | Central question | Key distinction |
|---|---|---|
| Two-step flow | How does an intermediary’s personal influence connect media with another person’s judgment? | Focuses attention on interpersonal mediation. |
| Direct / one-step route | Can media reach and affect a person without the interpersonal route being studied? | Direct exposure can coexist with advice from others. |
| Multi-step flow | How do messages and influence move through several linked people or groups? | Expands the routes beyond one intermediary. |
| Diffusion of innovations | How does a new idea or practice spread and become adopted within a social system over time? | Opinion leadership can be relevant, but diffusion asks a broader adoption question. |
| Agenda-setting | How does media prominence relate to the importance audiences attach to issues? | Issue importance is a different outcome from following an adviser’s recommendation. |
| Framing | How is an issue organised and interpreted? | An intermediary can reframe a media account; the frame and the influence process require separate analysis. |
Use more than one approach when the questions justify it. But avoid treating every conversation as proof of two-step influence or every chain of sharing as a complete account of opinion formation.
For another angle on interpretation, compare News Framing. For a different approach to patterned media exposure, see Cultivation Theory.
Distinguish a route, a relationship and an effect.
Select a scenario. Write your diagnosis before opening the analysis. All cases are invented for learning; none reports an empirical finding.
Use Tab to reach the native radio group and arrow keys to change the case. Answers and selections are not submitted or saved.
Study who influences whom—and what difference follows.
A content count or a network diagram can show part of a process. Testing interpersonal influence requires evidence about exposure, exchanges, timing and outcomes. Keep those questions separate while examining how they connect.
Who encountered which media material, and when?
Who explained, discussed or recommended what to whom?
What, if anything, changed in understanding, judgment or action?
- Bound the question.
Choose a group, topic, media item and observation period. Define the outcome: awareness, understanding, preference, confidence, intention or behaviour. Do not treat them as interchangeable.
- Identify advice relationships.
Ask participants whom they consulted about that topic and why. Compare self-reported leadership with nominations and accounts from others. A person’s general confidence is not a sufficient measure of influence.
- Establish the sequence.
Use consented diaries, repeated interviews or a bounded survey to distinguish earlier media exposure, later discussion and subsequent judgments. Ask about direct use of the original source and other advisers.
- Examine rival explanations.
Friends may already share interests or preferences. This similarity—often called homophily—can resemble influence. Common exposure, shared circumstances and selection of like-minded contacts may also explain agreement.
- Match the conclusion to the design.
A retrospective interview can reveal reported influence but may contain recall errors. A sequence of observations improves the account but does not by itself eliminate every alternative cause. Strong causal claims need a design capable of testing them.
A small self-learning inquiry
With consent, ask a small group about a recent, low-stakes choice such as selecting a documentary or a study resource. Record the source they first encountered, whom they consulted, what advice was exchanged and what they say changed. Use anonymous role labels in your notes. Treat this as an exercise in asking better questions, not a representative survey.
Do not collect private chat logs, contact lists or identifiable network maps merely because they are available. Gather only what the exercise needs and respect each person’s decision not to participate.
Strengths and limits
| Issue | Contribution | Limit / response |
|---|---|---|
| Social relationships | Shows why media encounters are embedded in everyday networks. | Do not assume that all recipients rely on advice or that all advisers have the same role. |
| Active interpretation | Draws attention to explanation, criticism and personal relevance. | A conversation can reinforce an error as well as clarify a message. |
| Two-step simplicity | Offers a memorable route that generates research questions. | Include direct access, reciprocal influence, several steps and non-participation. |
| Opinion leadership | Helps examine unequal roles in advice and discussion. | Avoid identifying leadership only through fame, self-description or follower counts. |
| Network-level spread | Connects local exchanges with a wider diffusion question. | Local influence does not automatically explain a large cascade; network conditions also matter. |
| Evidence of effects | Encourages study of what conversations contribute. | Distinguish influence from similarity, common exposure and retrospective attribution. |
“Participants describe this adviser as helping them interpret the report” is different from “the adviser caused their final decision”. State which claim the data can support.
Turn the lecture into a guided study session.
Watch Dr. USP’s lecture on the Two-Step Flow Model, then return to the diagrams and cases. The chapter expands the lecture’s discussion of opinion leaders, social relationships and the limits of a simple two-step account.
| Start | Focus | Pause and ask |
|---|---|---|
| 00:00 | Limited effects and research foundations. | Which study and publication dates should I distinguish? |
| 00:56 | Informal opinion leaders and different audience roles. | What evidence would identify an advice relationship? |
| 01:50 | Opinion followers and the two-step route. | Why does receiving advice not establish passivity? |
| 02:46 | Media use, personal networks and leadership characteristics. | Is a person influential on every topic? |
| 03:42 | Groups, relationships and the beginning of the critique. | What about people who do not join a discussion? |
| 04:27 | Direct exposure, several steps and other sources. | Which alternative route fits a case I know? |
| 05:25 | Summary of the model’s contribution. | How would I test a claim of influence rather than assume it? |
Entry points are approximate guides derived from the supplied transcript, not official video chapters. Timestamp links may open in YouTube.
Trace it. Explain it. Test it.
First pass: draw the classic route from memory. Second pass: explain what the intermediary contributes. Third pass: add direct access, another adviser or feedback and identify the evidence needed to study the revised route.
Follow the route. Defend the claim of influence.
Choose an answer and explain it before revealing the feedback. These checks are for self-study; no responses are collected or automatically scored.
1. What is the central contribution of Two-Step Flow Theory?
- It links media with interpersonal influence inside social networks
- It proves that media never matter
- It requires every message to travel through exactly two apps
- It divides society into permanently active and passive people
Reveal answer 1
Answer: A. The classic route highlights the role of opinion leaders and personal contacts. It is a simplified communication relationship, not a universal numerical rule.
2. Which chronology is correct?
- Personal Influence appeared in 1940 and the election study began in 1955
- The Erie County research concerned the 1940 election; The People’s Choice appeared in 1944; Personal Influence appeared in 1955
- All three events occurred in 1955
- The 1955 book was written by Lazarsfeld alone
Reveal answer 2
Answer: B. Keep the study, the first report and the later Decatur-based work distinct. Berelson and Gaudet coauthored The People’s Choice; Katz and Lazarsfeld coauthored Personal Influence.
3. A classmate forwards a review. What does that establish by itself?
- A change in every reader’s attitude
- A permanent opinion-leader role
- A communication act; influence still needs evidence
- That the original media source had no effect
Reveal answer 3
Answer: C. Circulation does not by itself establish trust, interpretation or a changed judgment.
4. Which statement best describes opinion leadership?
- It belongs only to celebrities
- It is fixed across every topic
- It is identical to formal authority
- It can be informal, relational and topic-specific
Reveal answer 4
Answer: D. People can guide others in one area and seek guidance in another. Recognition by contacts and examples of influence are more informative than a label alone.
5. Does the term “opinion follower” mean a person cannot consult media directly?
- No; direct exposure and active evaluation can coexist with seeking advice
- Yes; followers never see media
- Yes; followers must accept advice
- Only if the adviser has a large audience
Reveal answer 5
Answer: A. The term identifies a role in a particular relationship, not a permanent inability to seek or evaluate information.
6. Which case most clearly requires a multi-step account?
- A person reads an announcement and acts without consulting anyone
- Several people reinterpret a report, pass revised advice onwards and receive questions back
- A person changes from a phone to a laptop
- An article contains two paragraphs
Reveal answer 6
Answer: B. Several interpersonal links and feedback extend the simple two-step route. A device change is not a communication step in this theoretical sense.
7. A discussion confirms an existing preference. What should the researcher conclude?
- No effect is possible because nobody reversed their choice
- The adviser necessarily caused all later behaviour
- Reinforcement may be an outcome, but its contribution needs evidence
- The media message was irrelevant
Reveal answer 7
Answer: C. Influence need not mean conversion. Define and investigate the outcome rather than assuming either change or no change.
8. Two friends agree after discussing a film. What alternative explanation should be considered?
- Their agreement proves a causal effect of the conversation
- Similarity is impossible in social networks
- Only the most popular friend can have influenced the other
- They may have shared preferences or encountered the same review before the discussion
Reveal answer 8
Answer: D. Homophily, common exposure and other experiences can resemble interpersonal influence.
9. What does a large follower count prove about classic opinion leadership?
- It identifies an aspect of reach, not necessarily topic-specific interpersonal influence
- It proves that followers accept every recommendation
- It proves a two-step effect on decisions
- It proves the account is accurate
Reveal answer 9
Answer: A. Reach, credibility, advice relationships and outcomes are different measures. They should not be substituted for one another.
10. Which finding most directly supports a bounded claim of personal influence?
- A post has a bright thumbnail
- A participant describes a specific adviser’s explanation, its timing and how it changed their interpretation, with supporting accounts
- Two people belong to the same group, with no record of exchange
- A message was available online
Reveal answer 10
Answer: B. This is relevant evidence of a reported process. The strength of a causal conclusion still depends on the design and competing explanations.
Applied assignment · 350–450 words
Choose a low-stakes media-related decision. Draw its possible communication route and identify a topic-specific adviser. Explain what the adviser may contribute. Distinguish exposure, interpretation and outcome; include direct access or another route. Propose evidence, one alternative explanation and a limitation of your conclusion.
| Criterion | Points | Check |
|---|---|---|
| Defined case and route | 2 | Are the topic, participants, channels and possible sequence clear? |
| Accurate concepts | 2 | Are opinion leadership, interpersonal influence and direct exposure distinguished? |
| Mechanism and evidence | 3 | Does the proposed evidence address what the adviser contributes and what follows? |
| Alternatives and feedback | 2 | Does the answer consider prior similarity, other sources or reciprocal influence? |
| Proportionate conclusion | 1 | Does the conclusion remain within what the evidence could show? |
Readiness check
Checkboxes are a temporary study aid. They do not save course progress.
Keep the route clear. Keep the evidence specific.
Key terms
- Two-step flow
- A model connecting media with other people through opinion leaders in personal networks.
- Opinion leader
- A person whose informal influence matters to others in a particular area or relationship.
- Opinion follower
- A person receiving guidance in a particular exchange; the role does not imply permanent passivity.
- Personal influence
- A contribution made through interpersonal contact to another person’s understanding, judgment or conduct.
- Interpersonal mediation
- The ordinary process through which people select, explain, evaluate or discuss a message for one another; not automatic proof of statistical causal mediation.
- Limited effects
- A tradition emphasising conditions and competing influences on media outcomes; not a claim that media have no effects.
- Reinforcement
- Strengthening or confirming an existing view rather than necessarily reversing it.
- Multi-step flow
- Communication involving several linked people or groups, potentially with direct exposure and feedback.
- Homophily
- Similarity among connected people; it can complicate an inference that agreement was caused by influence.
- Cascade
- A sequence of propagation through a network; local advice alone does not explain every large cascade.
Read the foundations, then test the extensions.
Begin with the lecture, Personal Influence and Katz’s review. Compare the later studies to see why routes, relationships and outcomes must be measured separately. Some publisher full texts may require institutional access.
- Pandey, U. S. (Dr. USP). Two-Step Flow Model of Communication [Video]. YouTube.
Open source / publication record ↗
Why read it? The lecture supplied for this chapter. Use the viewing guide and questions above. - Lazarsfeld, P. F., Berelson, B., & Gaudet, H. (1944). The people’s choice: How the voter makes up his mind in a presidential campaign. Duell, Sloan and Pearce.
Open source / publication record ↗
Why read it? The first report of the election research. Link: catalogue record for the original 1944 edition. - Katz, E., & Lazarsfeld, P. F. (1955). Personal influence: The part played by people in the flow of mass communications. Free Press.
Open source / publication record ↗
Why read it? The classic study of personal influence. Link: publisher’s anniversary-edition record, identifying the original 1955 publication. - Katz, E. (1957). The two-step flow of communication: An up-to-date report on an hypothesis. Public Opinion Quarterly, 21(1), 61–78.
Open source / publication record ↗
Why read it? Read the hypothesis and its development rather than relying only on the simplified diagram. - Bennett, W. L., & Manheim, J. B. (2006). The one-step flow of communication. The ANNALS of the American Academy of Political and Social Science, 608(1), 213–232.
Open source / publication record ↗
Why read it? A later argument about targeted communication, individualisation and the role of social mediation. - Watts, D. J., & Dodds, P. S. (2007). Influentials, networks, and public opinion formation. Journal of Consumer Research, 34(4), 441–458.
Open source / publication record ↗
Why read it? A simulation-based challenge to strong claims about exceptional influentials. Read the author-hosted paper. - Wu, S., Hofman, J. M., Mason, W. A., & Watts, D. J. (2011). Who says what to whom on Twitter. In Proceedings of the 20th International Conference on World Wide Web (pp. 705–714). ACM.
Open source / publication record ↗
Why read it? A study of information production, attention and intermediary roles. Read the university-hosted paper. - Choi, S. (2015). The two-step flow of communication in Twitter-based public forums. Social Science Computer Review, 33(6), 696–711.
Open source / publication record ↗
Why read it? A bounded empirical application in online political discussion. The issue year is 2015; the article first appeared online in 2014. - Lazarsfeld, P. F., Berelson, B. R., & Gaudet, H. (2006). Erie County Study, 1940 (ICPSR 7204, Version 1) [Data set]. Inter-university Consortium for Political and Social Research.
Open source / publication record ↗
Why read it? An advanced research resource describing the original data. 2006 is the repository version date, not the election year.