Interaction Adaptation Theory¶
A dyadic communication model relating requirements, expectations, and desires to a partner's conduct and the respondent's adjustment.
Core Idea¶
Interaction Adaptation Theory (IAT) is a model of how one person adjusts communication in relation to another person's behavior during an exchange. The model was proposed by Burgoon, Stern, and Dillman in Interpersonal Adaptation; the book title is not the model's original name. IAT treats a person's requirements, expectations, and desires (RED) as inputs to an interaction position (IP), a reference for assessing the partner's actual communicative behavior (A). The respondent may move toward a comparable pattern, counter it, or make little observable adjustment, depending on the interaction.[1][2][3]
The comparison is conditional rather than a fixed arithmetic command. A favorable or unfavorable difference between A and IP does not, on its own, guarantee literal matching or compensating. RED components can conflict, their relative weights may be unclear, and the meaning of reciprocal or compensatory response concerns communication function rather than copying the same gesture. Floyd and Burgoon's original experiment found desire particularly influential in its stranger-liking task, while leaving the weighting of incongruent expectation and desire open.[3][4]
The theory concerns a dyad even when a study observes only one partner's account. The evidence level must be stated: a direct experiment with a trained stranger and observed participant response is different from a counselor interview reporting what clients did and how the counselor adapted.[3][4]
Structural Signature¶
- Dyadic communicators. Two people participate in a related verbal or nonverbal exchange. The theory asks how one person's conduct responds to the other's, not simply whether an individual likes an interaction.[2]
- Requirements, expectations, and desires. The responding person brings constraints or needs, anticipations, and preferred conduct to the exchange. These are theoretical inputs even if a particular study measures only some of them.[3][4]
- Interaction position. RED informs a contextual reference for the respondent. It is not a universal numeric score with fixed weights; when expectation and desire conflict, the operative reference may need empirical clarification.[3]
- Partner act and comparison. The partner enacts communicative behavior A that the respondent interprets in relation to IP. Direct observation and retrospective report give different strengths of evidence for this role.[3][4]
- Adaptive response. The respondent's subsequent behavior can be reciprocal, compensatory, or minimally adjusted. Classifying the response requires its relation to the partner act and its communicative function, not only a satisfaction rating or shared physical motion.[2][3][4]
Removing a partner act, a RED-informed reference, or the response relation leaves an expectancy judgment or behavior observation but not the full named model. Face-to-face contact and video mediation are settings, not required roles.
What It Is Not¶
- Not generic interpersonal adaptation. People adjust to each other in many ways. IAT specifically relates RED/IP, partner conduct, and a communicative response; behavior change alone does not establish the model.[3]
- Not a deterministic discrepancy-sign rule. The original experimental article explicitly discusses uncertainty when expectations and desires are incongruent; its task found desire salient. A simple positive-gap→match and negative-gap→compensate formula erases that problem.[3]
- Not literal gesture mirroring. Reciprocity can occur in the same communicative function without duplicating a movement. Compensation may oppose or offset a partner's conduct; neither label can be assigned from raw motion similarity alone.[4]
- Not only expectancy violation. A surprise or evaluation of violated expectation lacks the complete model unless RED/IP, partner behavior, and the respondent's adaptation are traced.[3][4]
- Not proof of interaction success. A reported reciprocal or compensatory move is an adjustment. It does not itself show rapport, therapeutic benefit, or a preferred outcome.[4]
Scope of Application¶
IAT is used to analyze dyadic verbal and nonverbal communication. An original stranger-liking experiment manipulated expectation and desire, trained confederates to convey liking or disliking, and observed participant responses. A later telehealth study uses the model to interpret counselors' reports of requirements, expectations, desires, client conduct, and their own adjustments during video-mediated therapy challenges. These are literal communication settings with the same conceptual roles, although the second study's role evidence is one-sided and retrospective.[3][4]
The model can guide inquiry in other professional conversations, but transfer must be tied to measured or reported roles. A physician–patient study abstract reports both convergence and divergence in communication behaviors; the accessible abstract alone does not map the full RED/IP sequence and is corroboration of clinical relevance rather than a complete case here.[5]
Clarity¶
IAT distinguishes what the partner did from what the respondent expected, needed, or wanted and from how the respondent answered. The same silence might be anticipated by one person and troubling to another; classifying response solely by the partner's visible behavior misses the respondent's interaction position. Conversely, asking only whether someone was surprised misses the subsequent adjustment.[3][4]
It also clarifies what a study has established. Floyd and Burgoon experimentally varied expectation and desire and observed response, while requirements were part of the theory rather than a separately manipulated factor. Chatterjee and Ricks elicited counselor accounts of client conduct and counselor adaptation but did not directly observe both partners or measure client RED. These evidence levels should not be presented as interchangeable validations.[3][4]
Manages Complexity¶
Dyadic communication contains many gestures, turns, beliefs, preferences, and reactions. IAT groups them into a tractable chain: respondent RED → interaction position → partner actual behavior → respondent's functional adjustment. The chain makes room for both convergence and divergence without reducing communication to a count of mirrored motions.[2][3]
The model also prevents a single-word explanation such as “expectation” from hiding conflict between anticipated and desired conduct. In an incongruent condition, both E and D could point to different responses; the user of the model must investigate their relative force rather than silently choose one. In video-mediated counseling, channel constraints and limited cues can complicate the same comparison without creating a different theory.[3][4]
Abstract Reasoning¶
To apply IAT to a particular exchange, identify the responding person and distinguish their requirements, expectations, and desires as far as the evidence allows. Establish the partner's enacted verbal or nonverbal conduct, then compare the interpreted act with the respondent's IP. Finally, observe or obtain a qualified report of the respondent's subsequent behavior and classify its communicative function. If a component is unmeasured, name the gap rather than computing a fictional IP.[3][4]
The method produces conditional hypotheses. When E and D conflict, ask which component appears to predict the response in that setting. Floyd and Burgoon's task found desire dominated many participant responses despite differing expectations or confederate behavior; that result does not make desire universally dominant. If only a retrospective counselor interview is available, infer a reported application of the model, not directly measured bilateral synchrony or causal treatment effect.[3][4]
Knowledge Transfer¶
Within interpersonal communication, the role sequence transfers from face-to-face stranger encounters to video-mediated counselor–client encounters: identify RED, IP, partner act, and functional response each time. The measurements do not transfer unchanged. Controlled confederate behavior and observed participant reactions offer a different test from counselors' retrospective interpretation of clients' actions.[3][4]
The model may inform clinical communication research, as an abstract-level physician–patient report of both convergent and divergent behavior suggests, but the complete RED/IP relation needs its own evidence in that setting.[5] Outside human interpersonal exchange, “adapting to another agent” is an analogy unless the new setting independently supports the named model's requirements, expectations, desires, interpreted partner act, and communicative response.
Examples¶
Canonical: stranger-liking experiment¶
Floyd and Burgoon studied 96 adults in face-to-face encounters with same-sex stranger confederates trained to express liking or disliking. The experiment varied what participants expected and desired from the stranger and examined their nonverbal response to the confederate's conduct. In this task, desire to receive liking or disliking strongly shaped response, largely irrespective of expectation or actual partner behavior. The article does not experimentally isolate R or settle how E and D should be weighted whenever they conflict.[3]
Mapped back: participant and confederate are the dyadic communicators; participant requirements belong to the theoretical RED frame while E and D are the manipulated inputs; their combination supplies a candidate interaction position whose conflicting weights remain uncertain; trained confederate liking or disliking is the partner act; observed participant nonverbal conduct is the adaptive response. The case tests a bounded prediction, not a universal sign rule.
Applied: counselor-reported telehealth interactions¶
Chatterjee and Ricks interviewed 31 counselors about challenging video-mediated client interactions. They applied RED/IP to counselor needs for a workable, private and interpretable session, expectations about client conduct, and desired therapeutic rapport or progress. Counselors reported client acts such as silence or moving off camera and described responses including clarification, altered cue reading, turn-taking changes, or reciprocal silence. The authors interpret some responses as compensatory or reciprocal. These are counselor accounts, not direct coding of both participants; the study does not establish client RED or treatment efficacy.[4]
Mapped back: counselor and client are the dyadic communicators; reported counselor requirements, expectations, and desires provide RED; the authors' interpretation of those inputs provides counselor IP; reported client behavior is the partner act; reported counselor changes are the adaptive response. The study-level map is complete as an interpretation while its bilateral observational support remains limited.
Structural Tensions¶
T1: Anticipated conduct versus desired conduct when they diverge. Expectations help a respondent interpret what a partner is likely to do, while desires mark what the respondent would prefer to receive or elicit. If E and D point in different directions, treating each as decisive yields different IP and response predictions. Floyd and Burgoon explicitly leave the weighting question open and observed strong desire influence in their particular task. Favoring E alone could miss goal-directed response; favoring D alone could miss accommodation to a foreseeable social situation. Diagnostic: When expectation and desire conflict, which was measured, what did the partner actually do, and which input better aligns with the observed response?[3]
This is a tension in constructing and testing the model's interaction position, not an assertion that every dyad is internally conflicted. A respondent whose E and D align does not face the same weighting problem.
Structural–Framed Character¶
IAT is mixed, leaning framed: its RED/IP–partner-act–response sequence is a reusable analytic structure within interpersonal communication, but expectations and desires depend on the participants' social setting and interpretation. Evaluative weight is limited in the model's classification: reciprocal or compensatory adjustment is not automatically good, and a clinical benefit must be tested separately. Human-practice dependence is high because the roles are communicative expectations, preferences and interpreted acts; study instruments determine which are visible. Institutional origin is a named scholarly model, not a communication rule imposed by an institution, although counseling and other roles can shape RED. Vocabulary travel is limited: adaptation is a broad word, but IAT's RED/IP and dyadic functional response are specific to interpersonal communication. Import versus recognition: the face-to-face experiment and telehealth study are recognized as cases at their different evidence levels; applying the name to an algorithm's generic parameter update would import a metaphor without the social-communicative roles.[3][4]
Its character: a conditional, theory-framed map of dyadic adjustment whose recurring roles are useful across communication settings, while interpretation of position and response remains context-dependent and empirically qualified.
Structural Core vs. Domain Accent¶
What lifts: the comparison of a reference position and a partner act is a literal instance of the live Comparison Prime's co-framing, dimension selection, alignment, and relational read-off. It is an internal constituent of IAT, while IAT adds RED formation and functional response. A still broader “reference state followed by adjustment” skeleton is a future-prime question, not an asserted parent. The live Adaptation Prime names a persistent modification process, whereas IAT is a proposed explanatory model of communication; sharing a verb is insufficient for subsumption.
Domain-bound core: IAT's reference is formed from human requirements, expectations, and desires, the partner act is communicative conduct, and the response is interpreted by reciprocal or compensatory social function. Removing these to obtain a generic feedback relation loses the named theory's predictive content.[3][2]
Prime bar: the evidence supports unlike settings within interpersonal communication. It does not show that the full RED/IP model recurs literally in independent noncommunication substrates. Any broader portable relation would require a separate abstraction and evidence; this entry remains the named social-communication theory.
Instantiates / Related Primes¶
This entry is part of Comparison.
The current typed proposal records Comparison as a strict internal constituent: the respondent's RED-informed IP and the partner's actual conduct are brought into the same interaction frame, aligned on communicative dimensions, and read for fit or discrepancy before a response is classified. IAT is not itself a species of Comparison; it is a named model containing that operation. The live Adaptation Prime concerns an adaptive process, and the live Theory Prime has a broader multi-substrate transfer identity not established for this named communication model. Interpersonal Communication is a relevant domain context, while Uncertainty Reduction Theory and Expectancy Disconfirmation are neighboring models; no additional typed edge follows from topical overlap. The Comparison edge remains subject to the final prospective graph audit.[2][3]
Relationships to Other Abstractions¶
Current abstraction Interaction Adaptation Theory Domain-specific
Parents (1) — more general patterns this builds on
-
Interaction Adaptation Theory is part of Comparison Prime
Comparing interaction position with partner conduct is an identity-bearing operation within the IAT model.The IAT model necessarily places the respondent's RED-informed interaction position and the partner's actual conduct in the shared dyadic frame, aligns them along communicative magnitude and valence or function, and reads their fit or discrepancy before classifying a response. This is a full Comparison operation inside the larger RED/IP–act–adjustment model, not a claim that IAT is itself a kind of Comparison. Floyd and Burgoon's experiment operationalizes the relation with expected and desired liking, enacted confederate behavior and observed response; the counselor study interprets the same roles from one-sided reports. Remove the comparison and IAT's differentiating link between position and response disappears. Comparison also occurs independently in experiments, metrology and other settings without IAT.
Hierarchy path (1) — routes to 1 parentless root
- Interaction Adaptation Theory → Comparison
Neighborhood in Abstraction Space¶
Interaction Adaptation Theory sits in a sparse region of the domain-specific corpus (97th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.
Family — Persuasion & Interpersonal Communication (17 abstractions)
Nearest neighbors
- Interpersonal Communication — 0.78
- Hyperpersonal model — 0.77
- Markedness Model — 0.76
- Interpersonal gap — 0.75
- Coordinated management of meaning — 0.75
Computed from structural-signature embeddings · 2026-10-17
Not to Be Confused With¶
Generic interpersonal adaptation names adjustment behavior, rather than the RED/IP–act–response explanatory model. Expectancy violation can identify a departure from anticipation without measuring desire, position, or subsequent adjustment. Uncertainty Reduction Theory centers prediction and information about another person, not the whole IAT response relation. Expectancy Disconfirmation in the current catalog is a consumer-evaluation identity, not a synonym for any disappointing partner act. Literal mimicry can occur without a RED-informed comparison or functional communication adjustment.[3][4]
References¶
[1] Judee K. Burgoon, Lesa A. Stern and Leesa Dillman, Interpersonal Adaptation: Dyadic Interaction Patterns (Cambridge University Press, 1995), doi:10.1017/CBO9780511720314, https://www.cambridge.org/core/books/interpersonal-adaptation/89B02200054971EC522D7D22199B2B71. Publisher description and original frontmatter/contents inspected; chapter 11's principles, model, and RED-example pages are contents pointers only. The theory-chapter body was not inspected. registry ↩
[2] Judee K. Burgoon, “Interaction Adaptation Theory,” The International Encyclopedia of Communication (2008), doi:10.1002/9781405186407.wbieci041, https://onlinelibrary.wiley.com/doi/abs/10.1002/9781405186407.wbieci041. Original author-authored abstract and bibliography inspected; full entry not inspected. registry ↩a ↩b ↩c ↩d ↩e ↩f
[3] Kory Floyd and Judee K. Burgoon, “Reacting to nonverbal expressions of liking: A test of interaction adaptation theory,” Communication Monographs 66, no. 3 (1999): 219–239, doi:10.1080/03637759909376475, https://www.koryfloyd.com/s/reacting-to-nonverbal-expressions-of-liking.pdf. Full original publisher PDF hosted by coauthor inspected; printed pp. 221–225 (theory and open weighting), 225–227 (method), 233–236 (findings and limits). registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x
[4] Chatterjee and Ricks, “Adapting to expectancy violations during telehealth interactions with clients: Communication and technological implications for counselors,” Health (2026), doi:10.1177/13634593261471913, https://journals.sagepub.com/doi/10.1177/13634593261471913. Full original publisher article inspected; adaptation framework, methods, RED findings, reported responses, and discussion of future direct observation. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r
[5] Ashley Duggan and Ylisabyth Bradshaw, “Mutual Influence Processes in Physician-Patient Communication: An Interaction Adaptation Perspective,” Communication Research Reports 25, no. 3 (2008), doi:10.1080/08824090802237618, https://www.tandfonline.com/doi/abs/10.1080/08824090802237618. Original publisher abstract only; cited solely as abstract-level corroboration of convergent and divergent behavior. registry ↩a ↩b