Influence
Overview
Influence (影響) refers to a relationship in which an entity, event, or action brings about changes in another entity's state, behavior, thinking, or values. It encompasses a broader scope than a simple cause-and-effect relationship and is a key subject of analysis in various fields such as sociology, psychology, economics, media studies, and network science. In modern times, as the speed and scope of influence propagation have expanded rapidly through social media and recommendation algorithms, its social implications have grown even larger.
Main Content
Concept and Etymology
'Influence' is a Sino-Korean word combining 影 (shadow) and 響 (echo, resonance). Shadows and echoes both originate from a causal object or sound, but are indirect results rather than direct entities in themselves, which illustrates the metaphorical nature of influence. In other words, influence implies changes that appear through indirect paths such as persuasion, imitation, diffusion, and learning, rather than physical coercion or direct action. For this reason, influence is distinguished from 'power' or 'force' and often presupposes voluntary acceptance by the other party.
Distinction between Causality and Influence
Causality emphasizes a necessary connection between cause and effect, whereas influence is a continuous concept in terms of direction and strength. Just because A influenced B does not mean B's change is entirely due to A; in most cases, multiple factors overlap. Therefore, when analyzing influence, methodology is required to distinguish correlation from causation and to control for confounding variables and selection bias. Representative tools for causal inference include randomized controlled trials (RCTs), natural experiments, instrumental variables, difference-in-differences (DID), and regression discontinuity design.
Social Influence
In social psychology, social influence refers to the process by which the presence or behavior of others changes an individual's judgment, attitude, or behavior. Representative examples include conformity, compliance, obedience, and persuasion. Solomon Asch's line experiment showed that group pressure can distort even judgments of obvious facts, and Stanley Milgram's obedience experiment revealed the powerful influence of authority on individual behavior. Robert Cialdini presented six principles of persuasion—reciprocity, consistency, social proof, liking, authority, and scarcity—organizing the practical mechanisms of influence.
Media and Opinion Leaders
Media effects research has shifted from strong-effects theory to limited-effects theory and then again to conditional strong-effects theory. The 'two-step flow of communication' theory advanced by Paul Lazarsfeld and others held that media messages diffuse to the general public through opinion leaders. Today, influencers and creators serve as opinion leaders in the digital age, and credibility and contextual fit, rather than follower count, have emerged as the core of influence.
Networks and Diffusion
Influence is amplified or blocked within network structures. Research suggests that weak ties are advantageous for the diffusion of new information, while strong ties are effective for the spread of behavioral change. In addition, scale-free network theory—in which a small number of hub nodes determine overall diffusion—threshold models, and complex-systems diffusion models explain the propagation of influence. Everett Rogers's diffusion of innovations theory distinguishes adopters into innovators, early adopters, early majority, late majority, and laggards, revealing the temporal structure of influence.
Measuring Influence
Influence is measured by citations, reach, engagement rate, conversion rate, and network centrality (degree centrality, betweenness centrality, closeness centrality, eigenvector centrality, etc.). In academia, citation indices and the h-index are representative; in social network analysis, algorithms such as PageRank are typical. However, measurement indicators can be manipulated, so attention must be paid to Goodhart's law, whereby the indicator itself becomes the goal.
Recent Trends
Entering 2024–2025, the creation and distribution structure of influence is being fundamentally reorganized by generative AI and algorithmic recommendation systems. First, as AI-generated content and virtual influencers participate in opinion formation, the question of 'who exercises influence' is becoming blurred. Second, as recommendation algorithms determine individual exposure, influence has become increasingly dependent on algorithmic affinity rather than follower count. Third, countries are introducing regulations requiring algorithmic transparency and labeling of false information, with the EU's Digital Services Act (DSA) and AI Act as representative examples. Fourth, in network science, amid objections to the minority superspreader theory, research is drawing attention showing that diffusion grows larger when many ordinary users participate repeatedly. Fifth, as the influencer economy—which converts influence into marketing metrics—matures, the verification industry that filters out metric distortion such as purchased followers and fake engagement is also growing. Amid these changes, influence is being redefined from an individual attribute into a systemic phenomenon combining platforms, algorithms, and collective dynamics.
Related Topics
- [[인과관계]] (Causality)
- [[사회적 영향]] (Social influence)
- [[여론]] (Public opinion)
- [[인플루언서]] (Influencer)
- [[네트워크 이론]] (Network theory)
- [[설득]] (Persuasion)
- [[확산 이론]] (Diffusion theory)