Psychology by BogdanovaThe Encyclopedia of Psychology for the Artificial Era

RESEARCH DOSSIER

Acceptance and Commitment Therapy and Contextual Approaches

Abstract. Acceptance and Commitment Therapy (ACT) is a contextual behavioral intervention organized around psychological flexibility and the relations among private experience, language, values, and observable action. This critical narrative research monograph reconstructs the history of ACT and relational frame theory, examines the six proposed flexibility processes, evaluates the validity of major psychometric instruments and causal mediation claims, and synthesizes randomized trials and meta-analyses across anxiety, depression, chronic pain, obsessive-compulsive symptoms, and workplace distress. Comparative evidence is interpreted by control condition and outcome. A 2026 synthesis of 34 active-comparator randomized trials found no average superiority of full-model ACT over other bona fide psychotherapies, while studies against waitlist and other controls support benefits in several domains. The work identifies measurement overlap, mechanism uncertainty, trial power, treatment fidelity, context sensitivity, cultural validity, safety, and digital delivery as priorities. It distinguishes philosophical models, clinical hypotheses, empirically supported outcomes, and open questions in contextual behavioral science and the Artificial Era.

Keywords: acceptance and commitment therapy; ACT; contextual behavioral science; psychological flexibility; relational frame theory; experiential avoidance; cognitive defusion; acceptance; self-as-context; values; committed action; CompACT; MPFI; AAQ-II; psychotherapy randomized controlled trials; process-based therapy; clinical mechanisms; digital behavioral intervention

Introduction and the Scientific Object

Acceptance and Commitment Therapy (ACT) is a contextual behavioral intervention organized around psychological flexibility: the capacity to notice ongoing experience and act in directions that serve chosen values while circumstances, learning histories, and consequences change. Its therapeutic program includes willingness to experience difficult feelings, altered relationships to thoughts, flexible attention, a perspective on self that can hold changing experience, clarified values, and repeated committed action. These terms designate potentially measurable patterns of responding rather than interchangeable virtues. ACT developed within behavior analysis and cognitive behavioral therapy, yet its scientific questions extend into experimental language research, emotion regulation, motivation, social interaction, and the philosophy of scientific explanation. This monograph treats ACT as both an intervention model and a research program embedded within contextual behavioral science. Its central objects are behavioral functions, specified processes of change, intervention outcomes, and the conditions under which findings transfer to different problems and populations Hayes et al. (2006) Dindo et al. (2017).

The first research question concerns the proposed causal organization: under what circumstances do attempts to avoid unwanted inner experiences narrow life and maintain distress? The second concerns treatment specificity: what do acceptance, defusion, values, and committed action contribute beyond common therapeutic factors, exposure, expectancy, and ordinary behavioral activation? The third concerns comparative effectiveness: when ACT improves depression, anxiety, pain interference, obsessive-compulsive symptoms, or work functioning, how large and durable are benefits relative to credible alternatives? The fourth concerns scientific inference: do the standard questionnaires measure the processes named by the theory, and do measured process changes actually precede and mediate clinical improvements? These questions demand coordinated evidence from laboratory component experiments, randomized comparative trials, psychometrics, longitudinal observation, and intensive single-person assessment. Every design establishes a different proposition, and their convergence forms the evidential basis of the monograph Levin et al. (2012) Gloster et al. (2020) Cherry et al. (2021).

The investigation is a critical narrative scientific synthesis. It draws on accessible original studies, primary theoretical books, published trial abstracts and full-text reports where available, psychometric development articles, meta-analyses, and conceptual critiques. It is not a preregistered systematic review and does not pool new effect sizes. Numerical claims are restricted to values attributable to the identified reports, and differences between active and inactive comparators are preserved. Practical illustrations describe treatment logic without prescribing treatment for an individual. The parent discipline is Counseling Psychology and Psychotherapy, with substantive connections to Learning and Behavioral Psychology, Emotion and Motivation, and Clinical Psychology and Psychopathology. Dialectical behavior therapy, exposure therapy, traditional cognitive therapy, and process-based intervention science provide necessary comparison points, while retaining distinct treatment histories and protocols Hayes et al. (2011) Hofmann & Hayes (2019).

Historical Development of Acceptance-Based Behavioral Therapy

Behavioral psychotherapy originally emphasized experimentally observable relations between contexts, learning, and action. Subsequent cognitive approaches analyzed the role of interpretations, assumptions, and appraisal in distress, while behavioral activation and exposure remained central methods of change. ACT emerged from attempts to understand the paradox whereby controlling or suppressing unwanted private events can become another source of restriction. Its early clinical formulations asked whether the functional consequences of a thought, feeling, or memory might matter more for the person's next action than the formal truth or intensity of that event. The 1999 book by Hayes, Strosahl, and Wilson organized experiential avoidance, acceptance, valued directions, and behavioral commitment into a coherent treatment approach. This was an important historical synthesis of antecedent behavior-analytic and clinical ideas; the contemporary evidence for individual mechanisms must still be assessed independently Hayes et al. (1999) Hayes et al. (2006) Chawla & Ostafin (2007).

The 2006 model paper by Hayes and colleagues connected ACT explicitly to relational frame theory and differentiated process, component, and outcome evidence. An extended clinical movement subsequently situated acceptance and mindfulness practices among contextual cognitive behavioral approaches, often described historically as the third wave of behavior therapy. The label identifies a period of methodological and therapeutic innovation, but it does not establish one homogeneous school. Behavioral dialectics, mindfulness-based cognitive therapy, metacognitive approaches, and ACT differ in their manuals, explanatory commitments, and intervention targets. Hayes, Villatte, Levin, and Hildebrandt described a common orientation as open, aware, and active behavior, directed at the context and function of experience. That characterization helps place ACT historically while leaving actual treatment equivalence an empirical question Hayes et al. (2006) Hayes et al. (2011).

During the 2010s the field moved beyond demonstration projects toward direct comparison with established treatments, trials across multiple conditions, and increasingly sophisticated measures of psychological flexibility. Controlled studies in anxiety, pain, obsessive-compulsive disorder, and workplace distress supplied evidence that broadened the treatment domain. Meta-analyses aggregated these studies while also exposing variation in trial quality and comparator choice. In parallel, functional contextualism and relational learning inspired a process-oriented reorganization of psychotherapy research, emphasizing within-person change rather than sole dependence on syndrome-specific treatment names. The historical significance of ACT therefore includes its twofold influence: it stimulated outcome research on acceptance-based interventions and it helped redirect attention toward manipulable behavioral processes. Whether each proposed process is distinctive, causally primary, or best represented by a conventional six-part model remains unsettled Arch et al. (2012) Gloster et al. (2020) Hofmann & Hayes (2019) Hayes, Hofmann, & Ciarrochi (2020).

Functional Contextualism and Standards of Explanation

Functional contextualism evaluates psychological events in terms of situated histories, present settings, and consequential patterns of action. A functional description asks what particular behavior does under particular circumstances and how contingencies change its future probability. For ACT, a frightened person's refusal to enter a crowded room can be understood in relation to immediate reductions in fear and longer-term losses of valued participation. The same outward response can have different functions in different situations; even an outwardly calm behavior can depend on costly suppression or on flexible choice. Context accordingly includes discriminative stimuli, social expectations, verbal relations, incentives, available resources, and the history of reinforcement. This explanatory orientation is useful only when its proposed contextual variables are operationalized in ways allowing an investigator to test predictions about behavior Hayes et al. (2013) Hayes et al. (2006).

The pragmatic standard often associated with this approach is prediction and influence with precision, scope, and depth. Precision requires sufficiently clear terms to distinguish proposed mechanisms from the outcomes they seek to explain. Scope requires a model that applies across relevant situations without suppressing important differences. Depth requires compatibility with credible findings at other explanatory levels rather than merely asserting conceptual coherence. The broad proposition that avoidance maintains suffering is scientifically incomplete until avoidance is behaviorally defined, its relevant consequences are measured, and alternative accounts are considered. An observational relation between questionnaire avoidance scores and depression may arise from shared negative affect, common wording, or a causal feedback loop. A stronger investigation manipulates a hypothesized process and assesses whether subsequent functioning changes as predicted Chawla & Ostafin (2007) Tyndall et al. (2019) Ong et al. (2020).

Contextual analysis also places culture, income, disability, discrimination, family obligations, and institutional constraints within the explanatory field. A person may value meaningful work yet have limited access to a safe workplace, or may pursue intimacy in a setting that punishes disclosure. Calling such constraints psychological inflexibility would confuse feasible choice with inner process. A treatment model must examine both personal patterns and the practical affordances available for change. This attention to environment matters for ethical applications and for generalization beyond the samples on which scales and protocols were developed. Values-guided action is best evaluated against actual opportunities, costs, and supported choice rather than solely against therapist-preferred behavior Kashdan & Rottenberg (2010) Hayes et al. (2011) Hayes, Hofmann, & Ciarrochi (2020).

Relational Frame Theory and Verbal Behavior

Relational frame theory (RFT) proposes that humans learn to relate events in arbitrarily applicable ways, including comparison, opposition, temporal ordering, hierarchy, and perspective taking, and that the functions of one event can transform through derived relations with another. A word can evoke distress even when its referent is absent; an imagined failure may make an opportunity appear dangerous; a verbal rule may persist after contingencies change. RFT offers an explanatory vocabulary for why symbolic cognition can contribute both to adaptive planning and to rigid patterns of avoidance. Hayes, Barnes-Holmes, and Roche's 2001 edited volume articulated the conceptual and experimental research program, while subsequent ACT work described how derived relational responding might amplify evaluative thought and self-narratives. RFT functions here as one proposed basic-science foundation, rather than an established complete explanation for every therapeutic result Hayes et al. (2001) Hayes et al. (2006).

From the standpoint of mechanism, cognitive fusion refers to responding to the products of language as if their literal content exhausts what can be done. An anxious prediction can be treated as a binding instruction, and a self-description can be treated as a fixed property. Defusion methods alter the contextual control exercised by such descriptions: a person may notice the sentence as a sentence, speak it aloud, or observe how its appearance changes a response. The intended outcome is functional flexibility while the thought remains present. Such exercises should be compared with exposure, attentional redeployment, decentering, reappraisal, and experimenter demand. An induced change in subjective believability does not alone establish that relational transformation caused downstream behavior. The value of RFT as a mechanism theory depends on experiments that isolate relevant relational operations and link them prospectively to clinical outcomes Hayes et al. (2006) Villatte et al. (2016) Levin et al. (2012).

Language also sustains values and personally meaningful patterns through verbal commitments extending beyond immediate reinforcement. A person may choose to endure discomfort for a relationship, scientific task, or social responsibility because long-range consequences are represented symbolically. ACT aims to preserve this constructive function of language while weakening rigid and avoidant forms of verbal control. The conceptual architecture thereby connects self-reference, temporal perspective, rule governance, and behavior in settings where delayed consequences cannot be directly perceived. The theory is promising because these claims can be made experimentally specific; it remains vulnerable when language-based interpretation becomes a universal explanation without direct evidence. RFT-informed treatment and clinical ACT should therefore be evaluated as partly connected research programs with separate empirical tests Hayes et al. (2001) Villatte et al. (2016) Hayes et al. (2013).

Psychological Flexibility as a Multidimensional Construct

Psychological flexibility has been described as the capacity to contact present circumstances more fully and persist or change behavior when doing so serves valued ends. This formulation contains several dimensions: discrimination of relevant environmental events, tolerance for private experiences, flexible verbal control, behavioral adaptability, and goal-sensitive action. It also contains normative content, because the identification of values affects what counts as adaptive persistence. A person who continues one activity despite fear may display flexibility in relation to a chosen aim, whereas the same persistence may be rigid in an altered setting. Flexibility therefore needs to be measured in relation to the demands of a context and the ability to switch behavior when consequences change. The broader literature on adaptation identifies partly overlapping flexibility constructs, making construct definition especially important Hayes et al. (2006) Kashdan & Rottenberg (2010) Cherry et al. (2021).

The conventional ACT hexaflex presents six interrelated processes: acceptance, cognitive defusion, present-moment awareness, self-as-context, values, and committed action. Their opposing patterns are often described as experiential avoidance, fusion, inflexible attention, attachment to a conceptualized self, disconnection from values, and ineffective or avoidant action. This configuration is a useful clinical map and a guide for constructing testable hypotheses. It does not follow that the six processes are statistically independent, that they are equally causally powerful, or that every successful intervention must change all of them. Psychometric models frequently reduce these constructs to fewer factors, and behavior may reorganize across multiple processes at once. The hexaflex is best treated as a theoretical specification requiring continued experimental and longitudinal evaluation Hayes et al. (2012) Francis et al. (2016) Rolffs et al. (2018).

Some investigators use psychological flexibility more broadly to describe emotion regulation, executive set shifting, goal pursuit, or accommodation to changing environments. A 2021 scoping review identified 23 flexibility constructs across 220 articles and emphasized problems of definitional overlap and instrument quality. This finding argues for explicit boundary conditions when ACT studies use the term. Researchers need to distinguish flexible attention from cognitive control tests, flexibility of behavior from psychological distress, and openness to experience from passive resignation. A measure built around the frequency of unwanted thoughts is not equivalent to one built around changing behavior under contextual demands. Accurately stating what an instrument observes is therefore indispensable to evaluating whether ACT changes the process its theory identifies Cherry et al. (2021) Bond et al. (2011) Ong et al. (2020).

Experiential Avoidance and the Problem of Control

Experiential avoidance involves efforts to alter, escape, or prevent contact with unwanted thoughts, memories, sensations, or emotions, especially when those efforts persist despite costly consequences. It can be overt, as when someone cancels a valued activity to avoid anticipated anxiety, or covert, as when repetitive mental checking temporarily relieves doubt. Avoidance is sometimes beneficial: leaving a dangerous environment, treating acute pain, or taking necessary rest can be adaptive. The relevant research object is rigid and context-insensitive avoidance that reduces opportunities, reinforces fear, or restricts values-guided action. This functional definition makes ACT a theory about patterns of response and their consequences. It also prevents a moralized demand that people endure every aversive condition without attempting to change it Chawla & Ostafin (2007) Hayes et al. (2006).

Negative reinforcement offers one candidate mechanism: immediate relief following escape can increase the future likelihood of escape, even if it produces long-term deprivation of valued experiences. Verbal processes can expand the range of avoided events by turning remembered, imagined, or symbolically related possibilities into triggers. Yet this explanatory story has multiple competitors. Avoidance may be determined by actual danger, low perceived efficacy, economic barriers, habitual motor routines, insufficient social support, or prospective calculation under uncertainty. Observational data cannot isolate these alternatives. A convincing process model examines within-person transitions between triggers, avoidance acts, short-term relief, and later functioning, ideally coupled to randomized intervention components Chawla & Ostafin (2007) Hayes et al. (2013) Kashdan & Rottenberg (2010).

The literature historically treated broad questionnaire scores as evidence of experiential avoidance, but the tendency of items to reflect distress raises serious identification problems. If respondents endorse an avoidance statement because they are already more distressed, correlations with symptoms can be partly tautological. More rigorous work requires direct behavioral tasks, ecological records of actual choices, discriminant psychometric analysis, and prospective tests showing that altered avoidance predicts future valued behavior above baseline symptom change. ACT's proposal retains scientific value as a falsifiable functional account: sustained avoidance under conditions where safe alternative action is possible should narrow behavioral repertoires, while increased willingness in those contexts should allow wider participation. Whether each individual intervention follows this sequence must be empirically established Tyndall et al. (2019) Ong et al. (2020) Cherry et al. (2021).

Acceptance and Willingness

Acceptance in ACT is an active willingness to make room for private experiences when doing so supports workable action. The relevant question is what a person can do while pain, anxiety, sadness, embarrassment, or uncertainty is present. An acceptance exercise can include observing physical sensation, identifying a habitual escape response, and exploring the consequences of behaving differently in an adequately safe context. It does not ask that an external injustice or avoidable hazard be tolerated. Acceptance is functional because the presence of discomfort ceases to dictate the entire response repertoire. It may become visible through behavioral choices such as returning to meaningful activities despite unpleasant sensation, while the person's reported distress remains unchanged. The construct is therefore distinguishable from symptom relief, resignation, emotional numbing, and the absence of coping efforts Hayes et al. (2006) McCracken & Vowles (2014).

In clinical research, the proposed mechanisms of acceptance overlap with exposure and inhibitory learning. Repeated contact with feared cues without habitual avoidance may change predictions about danger, increase tolerance for uncertainty, or create new context-dependent response associations. ACT may facilitate this contact by explaining it through willingness and chosen purpose rather than emphasizing symptom reduction. A treatment could then obtain benefits through processes common to other approaches while fostering a different subjective orientation. Mechanism research must compare components under equivalent exposure and expectancy conditions, examine temporal sequencing, and measure durable behavioral repertoires. Without such comparisons, the observation that clients report more acceptance after therapy supports only association with improvement Arch et al. (2012) Wetherell et al. (2011) Levin et al. (2012).

Acceptance is also affected by social conditions. People may be asked to accept distress generated by discrimination, violence, work exploitation, or inadequate access to care. An ethically coherent contextual model distinguishes unavoidable private experience from modifiable external conditions and permits values-guided resistance to harmful arrangements. In clinical formulation, questions about safety, power, bodily vulnerability, and feasible behavioral alternatives precede assumptions that an avoidant response is excessive. These distinctions matter for medical and trauma-related contexts, where emotional exposure can interact with pain, fatigue, medication, memory, and relationships. The therapeutic target remains contextually flexible responding, evaluated in light of both individual objectives and the environment in which action occurs Dindo et al. (2017) Hayes et al. (2011).

Cognitive Defusion and Thought–Behavior Relations

Cognitive defusion refers to responding to thoughts as events within ongoing experience rather than allowing their literal content to determine behavior automatically. A person may have the sentence 'I will fail' and still proceed with a valued experiment, conversation, or application. Defusion strategies can involve repeating a phrase until its verbal form becomes salient, labeling it as a mental prediction, or noticing how a description influences posture and action. Their purpose is to create flexible distance between a verbal event and behavioral control. The exercise does not require concluding that the thought is false. This emphasis makes the ACT vocabulary distinct from methods that explicitly evaluate the evidential accuracy of beliefs, although real interventions can contain overlapping attention and exposure elements Hayes et al. (2006) Villatte et al. (2016).

A credible scientific model must specify which variable defusion manipulates: believability, attention to verbal form, conditioned emotional function, behavioral avoidance, or a more general repertoire of flexible rule following. Laboratory component experiments suggest that brief exercises may alter immediate responses to difficult private experiences, but their outcomes can reflect novelty, repetition, demand, and short testing intervals. Clinical evidence requires proof that defusion changes a process before clinical benefit, that change predicts a consequential behavioral difference, and that alternative mechanisms have been sufficiently constrained. The 2012 laboratory component meta-analysis offers evidence relevant to ACT processes while also demonstrating the limits of extrapolating brief experiments to long-term psychotherapy Levin et al. (2012) Hayes et al. (2013).

The relation between defusion and cognitive restructuring is an important comparative question. Restructuring can reduce rigid belief dominance by evaluating predictions against evidence, and defusion can improve belief flexibility by changing how a person responds to persistent cognitions. Both may involve metacognitive observation and revised appraisals. Distinct explanations require designs in which treatment dose, therapist contact, behavioral practice, and credibility are matched; process measurements must have demonstrated discriminant validity. In a case in which an inaccurate belief signals real danger, evidence evaluation can be valuable. In another case, a repeated rational argument may function as reassurance seeking. Functional assessment must therefore determine the consequences of each operation rather than classify techniques solely by their labels Forman et al. (2007) Arch et al. (2012) Ong et al. (2020).

Present-Moment Attention and Contact With Context

Present-moment awareness concerns flexible attention to internal and external events as they unfold. In ACT this includes sensations, thoughts, environmental cues, interpersonal signals, and action opportunities. A clinician may guide a client to notice a tightening chest, the sound of another person's voice, and the impulse to withdraw before choosing a response. The scientific hypothesis is that expanded, discriminating contact with current contingencies enables more responsive action than rigid absorption in verbal predictions and memories. Attention is neither neutral nor cost-free: monitoring bodily sensations may heighten distress in some populations, and increased awareness without supportive behavioral choices may become rumination. Assessment should examine both attentional changes and what follows from them Hayes et al. (2006) Hayes et al. (2011).

Mindfulness is represented differently across contemplative, cognitive, clinical, and behavior-analytic traditions. In contextual behavioral work it is often analyzed as a combination of attentional contact, openness, and reduced rigid control by private events. Evidence for mindfulness training broadly cannot be transferred directly to a particular ACT component without matched intervention tests. The ability to describe a thought, return attention to a task, or sustain observation of a feeling may depend on different skills and environmental structures. Designs should therefore separate attention allocation, emotional acceptance, perceived control, and behavioral choice. Intensive measures of daily attentional shifts could complement standard self-report scales and help identify which element changes first Hayes et al. (2011) Francis et al. (2016) Rolffs et al. (2018).

Social interaction is a particularly important context for attention. Listening during disagreement can require detecting another person's signals while noticing one's own impulses to argue or retreat. In a clinic, the therapist's pacing, questions, and responsiveness can shape attention without formal mindfulness exercises. Outside therapy, access to quiet spaces, predictable routines, language accommodation, and safe relationships influences what flexible attention can accomplish. The empirical unit is an episode of attention in relation to an opportunity for action, not simply a trait adjective. This approach suggests repeated sampling of real settings with complementary behavioral and self-report indicators Villatte et al. (2016) Kashdan & Rottenberg (2010) Hayes, Hofmann, & Ciarrochi (2020).

Self-as-Context and Perspective Taking

ACT distinguishes the conceptualized self, expressed as relatively stable verbal descriptions, from a perspective from which changing experiences can be noticed. Self-as-context refers to flexible perspective taking that permits the recognition of thoughts, memories, and roles as contents of experience rather than total definitions of personal identity. A person can notice a self-judgment after failure while remaining capable of learning, repairing harm, and making choices. This clinical construct is linked to relational accounts of deictic responding, including relations such as I–you, here–there, and now–then. These links are theoretical and researchable; the self-as-context vocabulary is not evidence of a special metaphysical entity or an autonomous observer outside ordinary psychological processes Hayes et al. (2001) Hayes et al. (2006) Villatte et al. (2016).

Empirical assessment of self-as-context is demanding because questionnaires may mix decentering, self-compassion, dissociation, emotional detachment, and general distress. A person's endorsement of an observer metaphor may express comprehension of a therapist's language rather than an enduring difference in response to adversity. Behavioral tasks can explore whether perspective training improves flexible use of self-relevant information or reduces the grip of rigid identity rules. Longitudinal clinical designs should assess whether perspective changes occur before improvements in meaningful action, with controls for altered mood and expectations. Existing multidimensional measures are a methodological starting point, while claims of a distinct causal factor require stricter validation Rolffs et al. (2018) Cherry et al. (2021).

The topic acquires additional importance when treatment concerns stigmatized identities, disability, age, or traumatic experiences. A contextual approach can help examine the difference between an imposed social label and the person's range of current possibilities, while retaining the material consequences of discrimination. The aim is increased agency within realistically described circumstances. A formulation that attributes structural restrictions to excessive attachment to identity would misuse the theory. Research must account for whether new perspective-taking abilities lead to safer relationships, better social participation, or more autonomous choice, and whether such effects survive outside the supportive therapy setting Hayes et al. (2011) Dindo et al. (2017).

Values as Behavioral Directions

Values in ACT refer to chosen qualities of ongoing action, such as responding with care, curiosity, honesty, responsibility, or contribution. They differ from specific goals, which can be completed, and from social expectations that may be accepted without reflection. A person might value learning while pursuing multiple concrete goals whose feasibility changes over time. The scientific function of values is to provide a flexible source of motivational organization when immediate affective consequences discourage behavior. Clinical work seeks to identify actions that express a meaningful direction and to make those actions observable in daily contexts. The distinction is consequential because a client can have a valued direction and still face real constraints on particular goals Hayes et al. (1999) Hayes et al. (2006) Hayes et al. (2012).

Values clarification may involve structured choices, exploration of personally significant relationships, or examination of the consequences of current behavior. Its mechanisms may include shifts in attention to delayed rewards, changes in rule-governed behavior, self-determination, commitment to identity-relevant projects, and social reinforcement. These are candidate explanations that require direct tests. A posttreatment rise in an inventory score of valued action can result from improved mood, changed expectations, or reinterpretation of the questionnaire. Studies should include independently observable behavior and longitudinal patterns rather than treat a verbal declaration of values as proof of their enactment Rolffs et al. (2018) Hayes, Hofmann, & Ciarrochi (2020).

Values research also faces normative questions. Some goals may conflict with the welfare of others, and individuals disagree about what constitutes a meaningful life. Ethical delivery requires informed, client-led articulation of values within legitimate safety and relational responsibilities. Clinicians cannot assume that their preferred work ethic, family form, or cultural conception of flourishing belongs to every client. An ACT-derived program should distinguish process claims about sustained chosen action from a moral claim about which choices are superior. That distinction makes values-guided therapy compatible with cultural pluralism and supports the study of how social supports and constraints shape implementation Hayes et al. (2011) Dindo et al. (2017).

Committed Action and Flexible Persistence

Committed action is the repeated construction, revision, and maintenance of behavior serving chosen values. It includes skill acquisition, specific goals, planning, graded practice, obstacle management, and willingness to change strategy when conditions require. Persistence can express flexibility when it continues a meaningful project through predictable discomfort, but the same persistence becomes rigid when evidence shows that the project has become harmful or impossible. ACT therefore requires both commitment and responsiveness to consequences. The unit of analysis is situated behavior over time, not an isolated intention or an affective state. This emphasis makes behavioral outcomes—attendance, task completion, communication, mobility, and participation—central tests of therapeutic impact Hayes et al. (2006) Hayes et al. (2012).

Committed action overlaps with behavioral activation, exposure, problem solving, goal setting, and rehabilitation. The shared behavioral elements make comparative trials informative: if two programs contain similarly effective activity scheduling, improvement cannot simply be credited to ACT's conceptual vocabulary. Investigators must identify what treatment adds, whether values clarification changes the likelihood or persistence of action, and which patients benefit from particular methods. Functional measures should distinguish the number of actions attempted from their relevance to the person's freely chosen aims. In pain or anxiety care, activity can improve even while symptoms persist, making functioning a particularly meaningful endpoint Wetherell et al. (2011) McCracken & Vowles (2014) Dindo et al. (2017).

Commitment unfolds across uncertain timescales. Immediate experiments can demonstrate willingness, yet broad life changes depend on habit formation, opportunity, resources, relationships, and sustained practice. High dropout, changing circumstances, or unstable living conditions may interrupt treatment independently of a person's psychological flexibility. Intensive longitudinal designs can test transitions from intention to enacted behavior and distinguish short-lived enthusiasm from repeated adaptation. They can also reveal when an intervention increases effort without improving quality of life. The strongest evidence of successful committed action is meaningful, autonomously selected participation sustained under realistically varied conditions Hayes, Hofmann, & Ciarrochi (2020) Hofmann & Hayes (2019).

Measuring Flexibility and the Acceptance–Action Questionnaire Tradition

Research on ACT faces a fundamental measurement problem: the explanatory construct must be distinguished from the distress and functional impairment it is intended to predict. The Acceptance and Action Questionnaire and its second version, the AAQ-II, were developed to index psychological inflexibility and experiential avoidance using brief self-report items. Bond and colleagues' 2011 psychometric investigation reported useful reliability and associations with clinically relevant criteria across multiple samples. Such evidence supports the scale's capacity to organize covariance among responses, yet it does not establish a direct measure of a specific behavioral mechanism. In particular, statements about unwanted private experience can be endorsed because distress is frequent or severe rather than because the respondent employs rigid avoidance. Item-level analyses and comparisons with measures of negative affect are indispensable before causal interpretation Bond et al. (2011) Tyndall et al. (2019).

Later psychometric investigations raised challenges to the scale's discriminant validity. Tyndall and colleagues examined whether the AAQ-II measures a separable avoidance tendency, while Ong and colleagues compared AAQ-II scores with other inflexibility instruments and measures of emotional distress. Ong and colleagues' study identified stronger discriminant validity for the CompACT than for some alternatives in the populations investigated. A 2024 analysis by Ruiz and colleagues found discriminant item structure for the Spanish-language AAQ-II in Colombian samples and argued for its interpretation specifically as psychological inflexibility rather than as an interchangeable measure of flexibility or experiential avoidance Ruiz et al. (2024). The coexistence of these findings makes construct specification and population-specific validation central to measurement selection. The more precise question asks what unique predictive and mechanistic information each scale supplies above general distress, response style, and baseline impairment Tyndall et al. (2019) Ong et al. (2020).

ACT outcome researchers should specify measurement roles in advance. Psychological flexibility might be a proposed mediator, a prognostic covariate, a treatment target, or a patient-reported outcome; these roles entail different estimands. A scale used to establish clinical improvement should not automatically be reused as evidence that the treatment's hypothesized mechanism has been confirmed. Multimethod designs can pair questionnaire scores with laboratory behavioral choices, ecological momentary reports, observer ratings, and objectively recorded participation. Measurement invariance across languages, diagnoses, and service contexts should be investigated rather than assumed. The eventual standard is a process measure that predicts future contextually appropriate behavior while remaining sufficiently distinct from symptom severity Cherry et al. (2021) Bond et al. (2011) Rolffs et al. (2018).

Multidimensional Process Instruments and Construct Representation

The Comprehensive Assessment of Acceptance and Commitment Therapy Processes (CompACT) was developed to represent multiple facets of the intervention model, including openness to experience, behavioral awareness, and valued action. The Multidimensional Psychological Flexibility Inventory (MPFI) sought a more differentiated representation of the six flexibility and six inflexibility processes associated with the hexaflex. These measures have different conceptual levels and practical burdens. A brief global scale may be useful for repeated clinical monitoring but obscures potentially distinctive components; a long multidimensional inventory can support detailed modeling yet increases participant burden and may produce correlated factors difficult to separate. Factor structure is an empirical property of item responses, not proof that an underlying process is causally unitary or divisible Francis et al. (2016) Rolffs et al. (2018).

Validation requires reliability appropriate to the inference: internal consistency, temporal stability when expected, sensitivity to genuine change, and discriminant performance relative to adjacent constructs. Conventional coefficient alpha is insufficient for a multidimensional intervention model because correlated items can yield high consistency without capturing the targeted process. Item response analysis can investigate measurement precision across severity ranges, while longitudinal factor models can test whether items mean approximately the same thing before and after therapy. If treatment changes the meaning respondents assign to an item, apparent process change may reflect response shift. Cultural adaptation likewise requires cognitive interviews and differential item functioning analyses, especially when values and acceptance translate imperfectly into local languages Francis et al. (2016) Rolffs et al. (2018) Cherry et al. (2021).

The validity challenge becomes more acute when the model predicts functional, situation-sensitive change. A respondent may report low acceptance in a period involving genuine threat, or high acceptance during a period with few distressing events. A single global score collapses these differences. Scenario-specific items, event-contingent sampling, and behavioral tasks can distinguish response to difficult events from the frequency of exposure to them. Hierarchical models can estimate flexibility as a within-person coupling between circumstances and action rather than a static trait. Such designs would align measurement more closely with functional contextualism while preserving direct comparability with existing scales Hayes, Hofmann, & Ciarrochi (2020) Ong et al. (2020) Cherry et al. (2021).

Experimental Components and Laboratory Mechanism Research

Laboratory component studies seek to isolate the effects of brief acceptance, defusion, mindfulness, or values instructions. A typical paradigm elicits unpleasant thoughts, sensations, or emotional responses and compares a contextual instruction with control or alternative instruction conditions. Outcome measures may include willingness to continue a task, tolerance of discomfort, believability of thoughts, escape behavior, or performance under distraction. Levin and colleagues' 2012 meta-analysis assembled laboratory-based investigations of components suggested by the psychological flexibility model. These experiments provide useful evidence that processes can be manipulated under controlled conditions. They do not by themselves demonstrate that a few minutes of altered responding mediate improvement in chronic depression or pain Levin et al. (2012) Hayes et al. (2013).

Constructive experimental tests equate instruction length, experimenter contact, expectancy, task novelty, and emotional stimulus intensity across conditions. Some forms of apparent defusion may result from repeated stimulus exposure, habituation, or shifts in attention; an acceptance instruction may also function as an invitation to remain in an exposure task longer. A study can compare these specific explanations by measuring immediate avoidance, subsequent transfer to new stimuli, and delayed re-exposure. Factorial component designs can determine whether a values exercise modifies the effect of willingness training. Such designs are more informative about theoretical mechanisms than a simple comparison of an ACT-branded exercise and a passive waiting condition Levin et al. (2012) Arch et al. (2012).

Clinical experiments add important complexity because therapists adjust interventions in response to patients' behavior. Components are interdependent: values clarification can increase willingness, and defusion can permit committed action. A dismantling trial that removes one component may change treatment coherence, engagement, and therapist confidence as well as the intended process. Researchers should therefore combine component randomization with fidelity assessment, manipulation checks, and repeated mechanism measurement. Explanatory strength increases when an experimentally targeted process changes before the downstream behavioral outcome and the change is distinguishable from broad affective improvement Hayes et al. (2006) Gower et al. (2026) Hofmann & Hayes (2019).

Mediation, Temporal Ordering, and Causal Inference

A statistical mediator is a variable through which an intervention may influence an outcome, conditional on the assumptions of the chosen causal model. Conventional analyses estimate the association between treatment and process change, between process change and outcome, and the resulting indirect effect. For ACT, common candidate mediators include acceptance, inflexibility questionnaire scores, valued action, and avoidance. An indirect effect can support a process hypothesis, but it cannot establish the mechanism when mediator and outcome are measured simultaneously or share item content. Symptoms may improve first, subsequently making private events seem more tolerable; an unmeasured factor such as alliance may change both. Temporal ordering and measurement specificity are therefore essential Forman et al. (2007) Hayes et al. (2013) Hofmann & Hayes (2019).

Repeated assessment can improve identification by establishing when an intervention component occurs, when the presumed process changes, and when functioning changes. Cross-lagged associations are often used for this purpose but require assumptions about stationarity, unobserved confounding, and the meaning of between-person versus within-person effects. Intensive longitudinal data make it possible to distinguish short-term feedback loops: a day's increased willingness might enable social participation, which then reduces distress and reinforces willingness. Dynamic structural equation models or carefully specified within-person models can represent such reciprocal processes more faithfully than a two-wave mediation model. They remain observational unless the putative mediator is manipulated or confounding assumptions are otherwise defensible Hayes, Hofmann, & Ciarrochi (2020) Cherry et al. (2021).

A strong causal research program can randomize both treatment and process-focused components, prespecify mediation estimands, and capture repeated behavioral outcomes. It should ask whether changing a particular process is sufficient, necessary, or merely one of several routes to improvement. Treatment may work through exposure learning, renewed activity, social reinforcement, changed verbal interpretation, and therapeutic alliance simultaneously. Causal mediation studies must also consider missingness, treatment adherence, process-outcome confounding affected by earlier treatment, and the risk of selecting mediators after viewing results. Mechanism claims should be proportional to the experimental control and temporal resolution achieved Gower et al. (2026) Levin et al. (2012) Hofmann & Hayes (2019).

Randomized Trials, Comparators, and Clinical Outcome Selection

Randomized clinical trials provide evidence about the consequences of offering or delivering ACT under defined conditions. Their interpretation depends heavily on the comparison. Waitlist control can estimate improvement relative to delayed access, but it leaves differences in therapeutic attention, expectations, and activity unbalanced. Treatment-as-usual comparisons depend on what usual care consists of in the relevant health system. An active credible psychotherapy comparator places greater pressure on claims of superiority because both conditions receive structured therapeutic attention. Outcomes likewise require definition: symptom reduction, functional participation, quality of life, adherence, and valued behavior are related but not identical domains. ACT's distinctive emphasis on functioning suggests that trials should assess meaningful activity alongside distress A-Tjak et al. (2015) Gower et al. (2026).

Treatment fidelity determines whether the intervention actually tests the intended model. Manuals, session recordings, independent raters, and documented process targets can help distinguish full-model ACT from a short acceptance exercise or a mixed intervention carrying an ACT label. A trial without adequate adherence monitoring cannot resolve whether an ineffective treatment result challenges the underlying model or reflects incomplete implementation. Conversely, a successful multimodal program cannot prove the unique efficacy of ACT processes if exposure, activity scheduling, or therapeutic support account for improvements. Researchers should report dose, therapist training, patient selection, concurrent treatments, and recruitment settings, along with prespecified endpoints Gower et al. (2026) Arch et al. (2012) Wetherell et al. (2011).

Statistical power and precision become especially important in comparative designs. A failure to reject a difference between ACT and cognitive behavioral therapy does not establish equivalence; equivalence or noninferiority must be assessed against a prospectively justified margin and an adequately powered design. Intention-to-treat analysis preserves the randomized comparison under appropriate missing-data handling, while per-protocol analyses address different questions and are sensitive to postrandomization selection. Follow-up periods should reflect the claimed durability of effect. Underpowered trials can still identify promising patterns, but pooled analyses and clinical recommendations must reflect the limits of their precision Gower et al. (2026) Gloster et al. (2020).

Meta-Analysis, Overlap, and Evidence Grading

Meta-analysis synthesizes results from studies that measure sufficiently similar clinical outcomes, while permitting variation in populations and implementations. A-Tjak and colleagues' 2015 synthesis identified 39 randomized clinical trials involving 1,821 participants across mental and physical health conditions. Its broad inclusion criteria provided an important early overview of the field. Later reviews expanded the evidence base and explored outcomes by disorder and comparator. Gloster and colleagues' 2020 overview covered 20 meta-analyses encompassing 133 studies and 12,477 participants. Studies were sometimes counted within more than one constituent meta-analysis; therefore, the 20 syntheses cannot be treated as 20 fully independent confirmations. Overlap matters when summaries are used to estimate the strength or breadth of cumulative evidence A-Tjak et al. (2015) Gloster et al. (2020).

The most recent comparative synthesis shifts the central question from benefit over minimal intervention to added benefit over established psychotherapy. Gower, Styler, and Litz's 2026 systematic review restricted eligibility to randomized trials comparing therapist-delivered, multisession, full-model ACT with bona fide psychotherapies. Its 34 eligible trials yielded a three-level random-effects estimate of g = −0.01 (p = .86) for mental and behavioral health outcomes, providing no evidence of ACT's superiority on average. The review identified weaknesses including limited preregistration, underpowered contrasts, incomplete intention-to-treat methods, and interpretive overreach. Its exploratory equivalence analysis was not a substitute for a definitive equivalence trial. The result supports ACT as a treatment option while discouraging a general claim that it outperforms established comparators Gower et al. (2026).

Review conclusions should be interpreted relative to outcomes, comparator intensity, diagnosis, time point, and confidence in the included studies. Small positive differences against waitlists can coexist with negligible average differences against bona fide psychotherapy because the estimands differ. Meta-analytic heterogeneity may reflect intervention fidelity, patient severity, delivery format, follow-up, and measurement decisions rather than a single source of bias. Researchers should examine risk-of-bias domains, publication bias, evidence certainty, and clinical importance, and should report separate symptom and functioning results where possible. Aggregate efficacy is important but cannot replace explanations of for whom a treatment works, under which contexts, through which processes, and at what cost Gloster et al. (2020) Gower et al. (2026) A-Tjak et al. (2015).

Depression, Behavioral Engagement, and Recurrent Withdrawal

Depression frequently involves reduced activity, hopeless verbal predictions, self-critical rumination, and diminished responsiveness to available rewards. ACT addresses these patterns through behavioral engagement, willingness to have difficult thoughts and moods, and the identification of meaningful directions. Its behavioral emphasis has substantive overlap with behavioral activation, while acceptance and defusion provide a different explanation for sustaining action in the presence of persistent negative affect. The conceptual account proposes that people need not await the disappearance of low mood before taking practical steps toward relationships, care, work, and valued routines. The clinical question is whether this framework enables greater or more durable action than available alternatives, not whether it offers an appealing vocabulary Forman et al. (2007) Dindo et al. (2017).

Forman and colleagues' 2007 randomized effectiveness study compared ACT with cognitive therapy among adults experiencing anxiety and depression. The treatments produced broadly favorable outcomes and examined potentially different pathways of change, though the exploratory process findings cannot conclusively establish distinct causal mechanisms. A 2025 systematic review of ACT for patients diagnosed with depression included ten randomized trials and reported improvements in depressive symptoms relative to the included controls, while grading the certainty of evidence for that result as low. Such conclusions are encouraging but require attention to the diversity of comparators, clinical severity, co-occurring diagnoses, and risks of bias. A pooled symptom benefit does not resolve whether flexibility, renewed activity, or therapeutic support drove the difference Forman et al. (2007) Kong et al. (2025).

Research on recurrent depression needs longer follow-up, assessment of functioning in actual environments, and transparent reporting of concurrent medication and care. Because ACT may target adaptation during recurring distress, relapse prevention and the capacity to resume meaningful action following symptom recurrence are particularly important. Studies also must distinguish depression severity from the experience of demoralization associated with chronic social adversity. A person may need social protection, material assistance, or medical evaluation alongside psychological treatment. Values-guided action should never be substituted for clinical risk assessment or interventions addressing external harm Dindo et al. (2017) Gower et al. (2026).

Chronic Pain, Functional Participation, and Disability

Chronic pain is a major research setting for ACT because pain intensity and pain-related disability are only partially coupled. Avoidance of movement, fear of exacerbation, fatigue, loss of role participation, and social withdrawal can amplify the functional consequences of persistent pain. A contextual model asks whether acceptance and values-guided activity permit safer participation without requiring complete elimination of pain sensations. The primary therapeutic outcomes may therefore include pain interference, mobility within medical limits, everyday activity, emotional functioning, and quality of life. It is essential to retain appropriate medical assessment and to differentiate adaptive pacing from fear-driven withdrawal; overactivity can be harmful in some medical conditions McCracken & Vowles (2014) Wetherell et al. (2011).

Wetherell and colleagues randomized 114 adults with chronic nonmalignant pain to eight weekly group sessions of ACT or CBT. The study assessed outcomes through six months and found improvement in several measures within the ACT group, including pain interference and depression, but detected no significant between-group differences in improvement. The result supports ACT as a plausible structured intervention within chronic pain management while supplying no general proof of superiority to CBT. It also illustrates the importance of studying functioning separately from pain intensity and of documenting treatment engagement, concurrent medical care, and expectations. Clinical relevance ultimately depends on whether the intervention enhances feasible, safe participation in the particular patient population Wetherell et al. (2011).

The broader evidence includes in-person and digitally delivered formats, but delivery changes intervention dose, support, attrition, and accessibility. A 2021 meta-analysis of online ACT for chronic pain aggregated five randomized trials including 746 participants; it reported favorable outcomes versus included controls for several pain-related and psychological measures, while calling for more rigorous active-comparator research. Remote care may improve access for individuals with mobility constraints yet also exclude those lacking connectivity, accessible interfaces, or privacy. Future pain research should combine patient-chosen activity endpoints with clinical safety measures and compare ACT against other credible interdisciplinary pain programs Trindade et al. (2021) McCracken & Vowles (2014) Gower et al. (2026).

Obsessive–Compulsive Symptoms and Response Flexibility

Obsessive-compulsive disorder presents repetitive responses to unwanted thoughts, uncertainty, and perceived threats. ACT theorizes that rigid attempts to control intrusive thoughts and relieve distress can maintain ritual behavior. A treatment may cultivate willingness to experience obsessions while refraining from compulsive responses, alongside values-directed engagement in everyday life. Exposure and response prevention (ERP) already targets ritual reduction through learning processes, so the clinical distinction between ACT-oriented work and established ERP must be defined carefully. ACT can use willingness language to support exposure, but claims that its verbal framework produces additional benefit require direct, well-controlled comparisons with high-quality ERP Twohig et al. (2010) Dindo et al. (2017).

In a randomized trial published in 2010, Twohig and colleagues assigned 79 adults with obsessive-compulsive disorder to eight sessions of ACT or progressive relaxation training, with no in-session exposure. Independent assessors found greater reductions in obsessive-compulsive symptom severity in the ACT condition at posttreatment and three-month follow-up. The trial is an important demonstration that a structured ACT package can outperform the particular active comparator used. It does not establish that ACT is superior to ERP or every form of CBT, because relaxation training and ERP target different mechanisms. The size, demographic composition, and limited follow-up also constrain generalization Twohig et al. (2010).

The most informative next comparisons would examine ACT with and without formal ERP elements, ERP delivered with different rationales, and process-focused treatments matched for duration and credibility. Outcomes should include symptom severity, the frequency and functional role of compulsions, participation in chosen activities, and relapse over longer periods. Investigators can test whether defusion and willingness improve response prevention, engagement, or adherence while accounting for baseline severity and comorbid depression. Mechanism findings would strengthen clinical selection if they predict who benefits from a particular treatment component rather than only reproduce the overall symptom improvement Twohig et al. (2010) Gower et al. (2026).

Workplace Stress and Other Transdiagnostic Settings

ACT has also been evaluated outside disorder-specific psychotherapy, including work-related distress, health care, and preventive programs. These settings test whether psychological flexibility has explanatory value for common challenges involving uncertainty, emotion, interpersonal conflict, and competing goals. Workplace programs may focus on continuing effective and meaningful work while handling stress, but organizations must also address workload, discrimination, unsafe practices, and other contextual causes of distress. Individual coping programs cannot substitute for organizational responsibility. Outcome measures should therefore include worker well-being, participation, safety, and institutional indicators rather than only self-rated symptoms Flaxman & Bond (2010) Dindo et al. (2017).

Flaxman and Bond's 2010 workplace trial randomized 107 adults with elevated distress to ACT, stress inoculation training, or waitlist. ACT and stress inoculation training showed similar improvements in psychological distress over the reported follow-up. The investigators' mediation analyses offered preliminary support for differing process pathways, including change in reported psychological flexibility in the ACT group. Those analyses cannot eliminate all unmeasured mediators, but they demonstrate the value of comparing intervention rationales under matched real-world conditions. For large-scale implementation, treatment selection should consider staff time, acceptability, organizational support, and access for workers with different schedules and needs Flaxman & Bond (2010).

Transdiagnostic delivery creates opportunities and challenges. ACT processes can be presented in diverse symptom contexts, but a broad label may conceal differences in safety, preferred outcomes, and intervention intensity. A brief classroom or digital exercise may be useful for certain skills while having little evidential relevance to a full protocol for major depression or complex comorbidity. Researchers should define the intervention with enough detail that its mechanisms and clinical appropriateness are auditable. Strong claims of transdiagnostic efficacy require replication across distinct populations and active comparators rather than extrapolation from conceptual breadth Dindo et al. (2017) Gower et al. (2026) Gloster et al. (2020).

Comparative Clinical Meaning and Treatment Selection

Clinical treatment selection requires simultaneous attention to average trial outcomes, a person's presentation, preferences, available services, and the qualifications of the provider. ACT can be attractive when someone seeks to maintain participation while distress persists, but preference does not establish superiority. Likewise, a well-validated conventional CBT protocol may be preferable where a particular condition has a stronger condition-specific evidence base and better trained providers. A current synthesis of full-model ACT against bona fide psychotherapies gives little support to an across-the-board superiority claim. That result does not erase meaningful improvements observed within ACT groups or its conceptual contribution to process-based inquiry Gower et al. (2026) Arch et al. (2012).

Different outcomes answer different clinical questions. A treatment that lowers fear but leaves daily participation restricted may be less valuable for some individuals than one improving work, family involvement, or mobility with modest symptom change. Conversely, severe depression or obsessive-compulsive symptoms may require close attention to symptom thresholds and safety alongside participation measures. Clinical significance, reliable change, and patient-defined functioning should be reported together. In underpowered comparisons, even apparently similar group means can be compatible with a range of clinically meaningful differences. Researchers should report confidence intervals and prospective decision criteria instead of equating nonsignificance with equivalence Gower et al. (2026) Gloster et al. (2020).

Variation within the patient population may be more actionable than a single global ranking of therapeutic schools. Baseline avoidance, treatment preference, comorbid pain, social support, therapist expertise, and practical barriers may all affect outcomes. A well-designed trial can examine prespecified interactions and, eventually, validate individualized treatment-selection rules in independent samples. At present, the responsible scientific synthesis recognizes ACT as a supported approach for several contexts, with unequal evidence certainty and important gaps in mechanism specificity. The next advance lies in identifying which procedures change which behaviors for whom, without overgeneralizing the branded package Hayes, Hofmann, & Ciarrochi (2020) Gower et al. (2026) Zhang et al. (2026).

Therapeutic Formulation and Functional Assessment

Functional formulation begins with a description of the behavior of concern, the circumstances in which it occurs, and the short- and long-term consequences maintaining it. In ACT, relevant observations can include a recurring internal trigger, an avoidance response, the immediate relief the response produces, and the opportunities it displaces. A clinically meaningful formulation might trace how anticipation of embarrassment prompts withdrawal from social events, reducing immediate anxiety while limiting relationships valued by the person. The description should remain open to competing explanations, including realistic danger, poor environmental accessibility, or limited material resources. A formulation is a testable map of influences, revised when observed behavior contradicts predictions, rather than a diagnosis generated from the presence of an ACT-related term Hayes et al. (2013) Dindo et al. (2017).

The therapist may move among acceptance, attention, values clarification, and action planning according to the client's response and current circumstances. An experiential exercise is useful when it changes the functional relationship between a cue and behavior, not because it reproduces a characteristic script. Treatment planning includes collaboratively chosen target behaviors, opportunities for rehearsal, measures of meaningful progress, and routine monitoring of adverse effects. Process fidelity involves documenting the function of an intervention rather than counting the number of times a preferred metaphor is used. This approach calls for trained professional judgment because the same technique may be helpful, irrelevant, or destabilizing under different clinical and interpersonal conditions Villatte et al. (2016) Hayes et al. (2011).

The explanatory advantage of functional analysis depends on precision and possible falsification. If a proposed avoidance mechanism is maintained by negative reinforcement, the clinician-researcher should be able to describe what relief follows escape, what behavior recurs, and what would count as evidence against the hypothesis. When a client persistently avoids a risky environment, safety may be a sufficient explanation and exposure may be inappropriate. When withdrawal continues despite reliable safety and undermines chosen aims, willingness-based exercises may be plausible. Treatment relevance arises from observing a concrete response pattern and evaluating alternatives, not simply identifying a symptom category Chawla & Ostafin (2007) Kashdan & Rottenberg (2010).

Delivery Formats, Adherence, and Therapeutic Relationship

ACT is delivered through individual and group psychotherapy, brief structured programs, workplace workshops, and internet-mediated interventions. Each format changes the treatment that participants actually receive. Individual therapy allows detailed functional formulation; groups offer social modeling and shared practice but limit individual time; digital programs enable flexible access while often relying more heavily on self-directed engagement. Studies cannot assume that interventions sharing the ACT label are equivalent when their intensity, therapist involvement, and opportunities for behavioral rehearsal differ. Reporting should include contact hours, attendance, training, adherence measures, modifications, and clinically relevant cointerventions Dindo et al. (2017) Trindade et al. (2021).

The therapeutic relationship can shape the operation of acceptance-based exercises. A client may attempt a difficult behavioral task because the therapist creates a predictable interpersonal environment, demonstrates respect, and calibrates the task to the person's circumstances. Such support can strengthen engagement in ACT and in other effective therapies, complicating claims that outcomes result exclusively from distinctive flexibility processes. It is therefore important to assess alliance and therapist effects independently and to compare clinician-delivered interventions with comparable attention and credibility. An intervention's specific theory may still contribute, but its unique incremental value must be established under controls sensitive to these shared mechanisms Gower et al. (2026) Hofmann & Hayes (2019).

Adherence and dropout are also scientific outcomes. A program that produces a favorable mean response among completers may be impractical if a large proportion cannot access it or leave early. Digital ACT may reduce travel burdens yet introduce barriers involving device literacy, confidentiality, attention, or data costs. Group formats can increase exposure to social evaluation and may be unacceptable for some people. Intention-to-treat analyses, transparent missing-data methods, and patient-preference reporting help determine real-world effectiveness. Implementation research should measure the quality of the support system as well as change in the individual participant Trindade et al. (2021) Gower et al. (2026).

Development, Culture, and Population Diversity

The psychological flexibility model is often presented as transdiagnostic, but its operational meanings are influenced by age, language, developmental capacities, and social context. A child, an older adult with cognitive change, and a young professional facing workplace insecurity differ in the kinds of abstract verbal rules they use and the opportunities available for values-guided action. Developmentally adapted interventions may require concrete language, supported communication, shorter tasks, and different measures. Evidence from adult outpatient samples cannot be assumed to generalize to pediatric, geriatric, neurodivergent, or severely medically ill groups without direct trials and culturally appropriate outcome definitions Dindo et al. (2017) Cherry et al. (2021).

Culture affects how acceptance, self, autonomy, and values are expressed. A questionnaire item emphasizing individual choice may measure something different where obligations to family or community organize the person's motivational field. Values can be relational and collective while still being meaningful to an individual. Translation of instrument words cannot alone establish comparable factor structure, item functioning, or relevance of the intervention. Research should involve locally knowledgeable investigators and participants in the design of examples, language, and behavioral outcomes. Evaluation of measurement invariance and qualitative understanding can reveal whether an observed cross-group difference reflects flexibility or a mismatch between an instrument and the population Rolffs et al. (2018) Francis et al. (2016).

Clinical trials should report who was recruited and who was excluded, especially when the intervention requires English fluency, reliable internet, transportation, stable housing, or a capacity to complete written exercises. Exclusion of highly distressed participants, those with acute safety needs, or people with significant comorbidity can make trial samples unlike ordinary clinical caseloads. Pragmatic research can evaluate supported delivery while keeping safety procedures and clinician expertise appropriate to severity. Equity of access is part of the contextual theory's explanatory remit: opportunity to perform meaningful behavior varies across institutions and populations, and it can be modified by environmental changes Gower et al. (2026) Hayes, Hofmann, & Ciarrochi (2020).

Clinical Safety, Risk Assessment, and Limits of Self-Help

Acceptance-based therapy concerns clinically meaningful distress and may involve exposure to feared sensations, memories, or situations. Careful screening and monitoring are necessary when participants have suicidal thoughts, severe mood episodes, psychosis, substance-related emergencies, serious medical conditions, or exposure to ongoing violence. ACT language about willingness must be applied within an appropriate plan for safety and stabilization. A person experiencing severe depression may require immediate diagnostic evaluation and individualized care beyond a general values exercise. The outcome of an exercise should be monitored for functional benefit and adverse response. Research protocols should distinguish transient discomfort, expected exposure-related distress, deterioration, and serious adverse events Dindo et al. (2017) Gower et al. (2026).

Clinical risk cannot be reduced to a score on a flexibility measure. Assessment should identify the severity of current symptoms, comorbid conditions, impaired functioning, vulnerability, and the availability of social and emergency support, according to the professional standards of the relevant jurisdiction. Exposure-like work and behavioral commitments must respect medical restrictions and realistic risks. When avoidance protects someone from actual abuse or danger, an intervention directed at reducing that avoidance can increase harm. Contextual case formulation thus requires assessment of the environment and not merely the internal experience of fear Chawla & Ostafin (2007) McCracken & Vowles (2014).

Scientific publications about ACT can explain treatment principles and evidence without substituting for clinical consultation or individual evaluation. The present monograph compares research designs and treatment findings; it offers no patient-specific diagnosis or medication recommendation. For interventions with substantial clinical claims, an independent qualified reviewer should assess the accuracy of diagnostic, risk, and treatment statements before public release. This requirement is especially pertinent where research targets people with serious psychiatric illness or medical disability. The record of review must name only a person who actually completed it and describe precisely what was evaluated Dindo et al. (2017) Gower et al. (2026).

Digital ACT, Personalization, and Data Ethics

Digital adaptations of ACT can offer structured psychoeducation, acceptance and defusion exercises, journaling, and support for selected behavioral commitments. Online programs vary from self-guided modules to therapist-supported protocols and cannot be assigned the same empirical status solely because they share component names. The chronic-pain meta-analysis by Trindade and colleagues synthesized five online ACT trials and reported generally favorable average outcomes against the controls included. Its modest evidence base also highlighted the need for stronger comparators, lower-bias measurements, and longer follow-up. Research must separate the effects of the therapeutic content from reminders, human coaching, app usability, and participants' access to stable technology Trindade et al. (2021).

Digital monitoring can make contextual analysis more precise. Ecological momentary assessment may record an episode of distress, the person's response, the accessibility of a valued activity, and a later outcome. With appropriate consent, repeated observations can reveal conditions under which willingness supports action or under which a particular exercise becomes burdensome. Such data are sensitive because they may disclose clinical symptoms, relationships, routines, and vulnerabilities. Responsible research requires data minimization, transparent retention and access practices, careful security, and consent that does not make participants feel compelled to disclose more than is needed for treatment Hayes, Hofmann, & Ciarrochi (2020) Trindade et al. (2021).

Adaptive algorithms may eventually help clinicians identify contextual patterns and choose process-focused exercises. Their validity depends on representative data, prospective evaluation, and evidence that the suggested decision improves outcomes compared with ordinary care. A model trained on self-reported distress may learn to predict distress rather than functional flexibility, and personalized recommendations may amplify biases present in training populations. The Artificial Era makes distributed monitoring and human–artificial partnerships relevant research questions, while the evidence required for clinical use remains behavioral, comparative, and safety-centered. The status of any artificial system as a conscious or independent psychological subject cannot be inferred from a digital treatment program's performance Hofmann & Hayes (2019) Hayes, Hofmann, & Ciarrochi (2020).

Education, Prevention, and Applied Behavioral Change

Acceptance and commitment methods also appear in educational, organizational, and community contexts where the immediate objective may be learning, resilience, or constructive action rather than treatment of a diagnosed disorder. The scientific question is whether skills involving openness, attention, and values-guided behavior improve meaningful performance under stress. Educational interventions should distinguish observed attendance, help-seeking, learning progress, and social participation from satisfaction with the training. Programs offered universally also need to account for participants who experience the exercises as irrelevant, culturally mismatched, or emotionally difficult Flaxman & Bond (2010) Dindo et al. (2017).

Brief exercises can be appealing because they are scalable, yet evidence from a single laboratory manipulation or a small uncontrolled workshop is insufficient to support large population-wide benefits. A school-based intervention should measure developmental appropriateness, attendance, comparison instruction, adverse responses, and academic or relational consequences over time. In a workplace, institutional work design can determine whether meaningful action is possible; encouraging acceptance while leaving dangerous policies unchanged can misallocate responsibility. A contextual approach gains practical credibility when it evaluates both individual behavior and the conditions enabling that behavior Levin et al. (2012) Flaxman & Bond (2010).

The strongest applied programs will specify the population, the problems addressed, and the outcomes that justify the intervention. They will include real comparisons, accessible materials, independent outcome assessment where possible, and transparent reporting of implementation costs. Follow-up should test whether skills generalize from training exercises to ordinary demands rather than merely improve participants' familiarity with the program's terminology. A scientifically integrated ACT model can inform applied behavioral change, but its breadth must be earned through replication in each domain Gower et al. (2026) Hayes, Hofmann, & Ciarrochi (2020).

Treatment Specificity and the Problem of Common Mechanisms

ACT is defined by a distinctive explanatory framework, but intervention effects can arise through both named and shared psychological mechanisms. Behavioral activation, exposure, self-monitoring, corrective interpersonal experiences, attention training, and improved confidence can operate across therapies. The finding that two treatments produce similar average benefits is compatible with several possibilities: they may recruit the same underlying processes, they may use different processes with equal net effectiveness, or different subgroups may respond through different routes. A direct comparison between ACT and CBT therefore provides information about relative outcomes without resolving every question about explanatory specificity. Component studies and process assessment are needed to test distinctive pathways Forman et al. (2007) Arch et al. (2012) Gower et al. (2026).

Mechanism research should look for evidence of change in a theoretically specified variable that is independently operationalized, temporally prior to an outcome, experimentally influenced when feasible, and comparatively predictive after plausible shared influences are accounted for. A questionnaire labeled 'psychological flexibility' can support such research but cannot perform all of those functions alone. In particular, decreases in distress may alter responses to items asking about the costs of unwanted thoughts, creating a statistical mediator even where flexible behavior has not changed. Stronger process indicators include choices made during safe but difficult tasks, persistence linked to chosen goals, and response shifts when contexts change. This expands the theory's potential for falsification while improving clinical relevance Ong et al. (2020) Tyndall et al. (2019) Hayes, Hofmann, & Ciarrochi (2020).

Some critiques interpret the absence of treatment superiority as evidence against ACT's unique usefulness. A more precise reading distinguishes comparative effectiveness, process validity, and therapeutic utility. The 2026 synthesis of full-model ACT and bona fide treatments does not demonstrate an average superiority effect, while process studies and individual trials may identify meaningful routes to behavior change. These propositions can coexist without weakening the obligation to test each. Progress requires preregistered hypotheses that predict how a component works, when it will be beneficial, and when another treatment should be preferable. Explanatory confidence depends on such discriminatory predictions Gower et al. (2026) Gloster et al. (2020).

Replication, Publication Bias, and Researcher Allegiance

Clinical psychotherapy research faces challenges related to researcher allegiance, publication selection, flexibility in analysis, and heterogeneity in treatment delivery. Investigators who developed an intervention may possess unusual therapeutic expertise and strong expectations about its effects; later independent trials can test whether results generalize. Publication bias can inflate estimates when positive small trials appear more readily than null trials. Unregistered outcome switching and selective reporting can further complicate interpretation, especially when a study collects numerous symptom and process measures. Meta-analytic conclusions depend on the completeness and quality of their trial corpus Gower et al. (2026) Gloster et al. (2020).

The 2026 comparative review identified recurrent shortcomings, including preregistration gaps, inadequate power for differences between active psychotherapies, and limited use of intention-to-treat analysis. These are methodological concerns across a set of studies rather than proof that each individual result is invalid. The appropriate response is a program of independent replication, transparent study protocols, clearly specified primary endpoints, accessible analytical materials where ethically permissible, and follow-up periods chosen in advance. Process measures should undergo independent construct validation before they are used as decisive explanations of treatment effects Gower et al. (2026) Cherry et al. (2021).

Outcome heterogeneity must be handled thoughtfully. Researchers sometimes combine anxiety, depression, functioning, and distress into broad categories, but aggregate measures can obscure clinically important distinctions. Multiple reports from the same trial and overlapping systematic reviews also risk unintentionally counting participants more than once. Gloster and colleagues directly investigated overlap among meta-analyses, showing why bibliographic independence and trial identification matter for evidence synthesis. Future cumulative reviews should preserve trial identifiers, protocol versions, comparator descriptions, and study-level risk-of-bias judgments. Scientific progress is cumulative when contradictory and null results remain visible in the record Gloster et al. (2020) A-Tjak et al. (2015).

Individual Differences, Dynamic Processes, and Precision Treatment

An average trial effect is the mean of heterogeneous individual trajectories. Some participants improve quickly, some change later, and others obtain little benefit or deteriorate. Psychological flexibility theory implies that the relevant process may depend on the present functional context: the same acceptance exercise might facilitate action in one situation while attentional redirection or practical problem solving is more helpful in another. Identifying these contingencies is a promising use of intensive longitudinal and repeated-treatment designs. Such research must be conducted with the same safeguards expected in conventional psychotherapy studies, including safety monitoring, consent, and transparent outcome definitions Hayes, Hofmann, & Ciarrochi (2020) Hofmann & Hayes (2019).

Within-person process studies can sample relevant antecedents, avoidance responses, value-linked activity, and later consequences across time. They can model lags and feedback loops and may identify predictors of when an exercise is followed by more effective behavior. Their limits include measurement reactivity, missing episodes, confounding by daily conditions, and the difficulty of distinguishing patterns learned from data from patterns prospectively validated. A statistical model that predicts tomorrow's distress is useful for forecasting but does not automatically justify choosing a specific clinical intervention. Randomized micro-interventions and independently tested decision rules can move from prediction toward evidence for improved care Hayes, Hofmann, & Ciarrochi (2020) Cherry et al. (2021).

Precision treatment requires more than matching a person to the therapy they prefer. It involves reproducible process measures, clinically meaningful adaptive rules, outcome comparisons against reasonable alternatives, and attention to unequal access. If a person has already mastered willingness but lacks practical resources, additional acceptance exercises may add little. If rigid avoidance is the central barrier, a well-supported behavioral exposure program may help regardless of whether it is presented as ACT or CBT. Identifying the most useful procedure in a given context is more scientifically informative than maintaining an unconditional allegiance to a named school Gower et al. (2026) Hofmann & Hayes (2019).

Contextual Behavioral Science and the Artificial Era

The transition toward the Artificial Era introduces new environments in which human behavior is organized through recommendation systems, digital communication, adaptive interfaces, and artificial cognitive tools. These technologies can shape prompts, expectations, avoidance opportunities, and routes to valued activity. A person may use an application to prepare for a difficult conversation, track a chronic condition, or practice recognizing a self-critical thought. A functional analysis asks whether the technology extends meaningful agency or maintains repetitive reassurance seeking and withdrawal. The therapeutic question remains empirical: how does a particular digital affordance alter human action in a particular environment? Hayes et al. (2013) Hayes, Hofmann, & Ciarrochi (2020).

Research on human–artificial interaction can investigate cognitive offloading, adaptive support, technology-mediated reassurance, self-monitoring, and the changing ecology of privacy and social attachment. ACT-derived language offers candidate constructs for describing these interactions, but using the construct does not establish that an artificial system possesses a human-like psyche. Artificial cognition can be studied through reproducible behavior, memory architecture, contextual sensitivity, and decision policies, with operational criteria specified for each claim. Philosophical frameworks about Artificial Sapiens and nonbiological reason belong to a separate explanatory level unless an empirical protocol tests a relevant hypothesis. Interdisciplinary integration is strongest when the boundaries between proposed ontology, psychological measurement, and clinical outcome are explicit Hofmann & Hayes (2019) Hayes, Hofmann, & Ciarrochi (2020).

New systems could eventually support personalized behavioral exercises, but they also introduce risks of inappropriate escalation, covert monitoring, dependence on unreliable outputs, and unequal treatment across populations. Clinical deployment needs robust comparison with ordinary care, clear boundaries between automated prompts and professional judgment, privacy protection, and procedures for handling safety-critical situations. An AI-generated exercise should be treated as a delivered intervention whose benefits and harms require evaluation, rather than as evidence that the system understands the patient. The contextual orientation contributes a framework for asking which interactions improve behavior in lived settings while demanding observable answers Gower et al. (2026) Trindade et al. (2021).

A Research Agenda for Mechanistic and Comparative Progress

The next generation of ACT science should first define a limited set of falsifiable functional hypotheses. Researchers can specify an environmental trigger, a hypothesized avoidant response, an experimentally manipulated process, and a later choice relevant to the participant's goals. The resulting design should distinguish the process from general negative affect, expectancy, therapist alliance, and familiar exposure mechanisms. A preplanned primary outcome could be verified attendance at a valued activity under safe conditions, complemented by symptom and quality-of-life measures. Such operational detail gives psychological flexibility greater explanatory force than an undefined global capacity Hayes et al. (2013) Ong et al. (2020) Gower et al. (2026).

Comparative clinical research should preregister protocols, recruit enough participants for realistic active-treatment contrasts, document manual fidelity, and analyze randomized participants under principled missing-data assumptions. Longer-term follow-up should evaluate relapse, maintained functioning, patient satisfaction, and adverse experiences. Trials could compare full-model ACT, mechanism-matched CBT, and carefully designed component combinations rather than rely on delayed-treatment controls alone. Independent replication and prospective equivalence or noninferiority margins are needed before claims that two treatments are interchangeable. The literature also needs tests in underrepresented languages, cultural settings, and multimorbid clinical populations Gower et al. (2026) Zhang et al. (2026).

Measurement research should combine dimensional questionnaires with event-level behavior, ecological experience sampling, and psychometric tests of invariance and response shift. Research syntheses must link unique trial records, report overlapping inclusion, describe controls precisely, and distinguish within-group improvement from between-group effects. Cross-disciplinary work should maintain the same specificity when analyzing workplace stress, chronic medical disability, and digital support systems. The overall goal is a science of contextually organized behavior that can identify useful procedures, measure their consequences, and revise its theories in response to reliable evidence Gloster et al. (2020) Cherry et al. (2021) Hayes, Hofmann, & Ciarrochi (2020).

Conclusion

Acceptance and Commitment Therapy established a significant contextual research program focused on the functional relations among inner experience, language, learning history, chosen values, and observable action. Its historical synthesis connected behavior analysis, relational frame theory, acceptance-based practice, and clinical behavior change into a coherent treatment framework. The hexaflex specifies six complementary process domains and provides a useful vocabulary for formulation and intervention design. Their scientific status depends on construct validity, causal testing, and evidence of context-sensitive functioning. Measures of psychological inflexibility have generated substantial research, while psychometric criticism underscores the need to separate reported distress from the behavioral processes proposed to explain it Hayes et al. (2006) Rolffs et al. (2018) Ong et al. (2020).

Across clinical conditions, ACT trials and meta-analyses document beneficial outcomes in several populations, especially compared with waitlist or minimally active controls. Randomized comparisons of ACT with CBT for anxiety and chronic pain show broadly comparable outcomes in influential trials, and the 2026 full-model comparative meta-analysis found no average superiority of ACT over other bona fide psychotherapies. Evidence certainty varies with the diagnosis, comparator, trial quality, and outcome. This pattern supports ACT's role among therapeutic options while directing research away from unqualified school-level superiority claims. It also demonstrates why improvement in symptoms, quality of life, and valued participation must be assessed with their distinct meanings intact Arch et al. (2012) Wetherell et al. (2011) Gower et al. (2026) Zhang et al. (2026).

The strongest continuing contribution of ACT is its insistence that human behavior be studied in the situations in which it unfolds. Acceptance, defusion, awareness, self-perspective, values, and committed action become scientifically informative when each is tied to observable contingencies, independently validated measures, and prospective predictions. Contextual behavioral science can integrate methodological advances in process-based therapy, intensive longitudinal research, and human–artificial interaction without sacrificing rigor. Its future will be determined by precise mechanisms, independent replication, comparative clinical utility, safety, and meaningful behavioral outcomes. In this form, ACT becomes a sustained inquiry into how people organize action in a changing world Hayes et al. (2013) Hayes, Hofmann, & Ciarrochi (2020) Gower et al. (2026).

References

  • A-Tjak, J. G. L., Davis, M. L., Morina, N., Powers, M. B., Smits, J. A. J., & Emmelkamp, P. M. G. (2015). A meta-analysis of the efficacy of acceptance and commitment therapy for clinically relevant mental and physical health problems. Psychotherapy and Psychosomatics, 84(1), 30–36. https://doi.org/10.1159/000365764
  • Arch, J. J., Eifert, G. H., Davies, C., Plumb Vilardaga, J. C., Rose, R. D., & Craske, M. G. (2012). Randomized clinical trial of cognitive behavioral therapy (CBT) versus acceptance and commitment therapy (ACT) for mixed anxiety disorders. Journal of Consulting and Clinical Psychology, 80(5), 750–765. https://doi.org/10.1037/a0028310
  • Bond, F. W., Hayes, S. C., Baer, R. A., Carpenter, K. M., Guenole, N., Orcutt, H. K., Waltz, T., & Zettle, R. D. (2011). Preliminary psychometric properties of the Acceptance and Action Questionnaire–II: A revised measure of psychological inflexibility and experiential avoidance. Behavior Therapy, 42(4), 676–688. https://doi.org/10.1016/j.beth.2011.03.007
  • Chawla, N., & Ostafin, B. (2007). Experiential avoidance as a functional dimensional approach to psychopathology: An empirical review. Journal of Clinical Psychology, 63(9), 871–890. https://doi.org/10.1002/jclp.20400
  • Cherry, K. M., Vander Hoeven, E., Patterson, T. S., & Lumley, M. N. (2021). Defining and measuring psychological flexibility: A narrative scoping review of diverse flexibility and rigidity constructs and perspectives. Clinical Psychology Review, 84, Article 101973. https://doi.org/10.1016/j.cpr.2021.101973
  • Dindo, L., Van Liew, J. R., & Arch, J. J. (2017). Acceptance and commitment therapy: A transdiagnostic behavioral intervention for mental health and medical conditions. Neurotherapeutics, 14(3), 546–553. https://doi.org/10.1007/s13311-017-0521-3
  • Flaxman, P. E., & Bond, F. W. (2010). A randomised worksite comparison of acceptance and commitment therapy and stress inoculation training. Behaviour Research and Therapy, 48(8), 816–820. https://doi.org/10.1016/j.brat.2010.05.004
  • Forman, E. M., Herbert, J. D., Moitra, E., Yeomans, P. D., & Geller, P. A. (2007). A randomized controlled effectiveness trial of acceptance and commitment therapy and cognitive therapy for anxiety and depression. Behavior Modification, 31(6), 772–799. https://doi.org/10.1177/0145445507302202
  • Francis, A. W., Dawson, D. L., & Golijani-Moghaddam, N. (2016). The development and validation of the Comprehensive assessment of Acceptance and Commitment Therapy processes (CompACT). Journal of Contextual Behavioral Science, 5(3), 134–145. https://doi.org/10.1016/j.jcbs.2016.05.003
  • Gloster, A. T., Walder, N., Levin, M. E., Twohig, M. P., & Karekla, M. (2020). The empirical status of acceptance and commitment therapy: A review of meta-analyses. Journal of Contextual Behavioral Science, 18, 181–192. https://doi.org/10.1016/j.jcbs.2020.09.009
  • Gower, T., Styler, Z. R., & Litz, B. T. (2026). Randomized controlled evidence of the efficacy of acceptance and commitment therapy: A systematic review and meta-analysis. Clinical Psychology Review, 128, Article 102786. https://doi.org/10.1016/j.cpr.2026.102786
  • Hayes, S. C., Barnes-Holmes, D., & Roche, B. (Eds.). (2001). Relational frame theory: A post-Skinnerian account of human language and cognition. Plenum Press. https://contextualscience.org/publications/hayes_barnesholmes_roche_2001
  • Hayes, S. C., Hofmann, S. G., & Ciarrochi, J. (2020). A process-based approach to psychological diagnosis and treatment: The conceptual and treatment utility of an extended evolutionary meta model. Clinical Psychology Review, 82, Article 101908. https://doi.org/10.1016/j.cpr.2020.101908
  • Hayes, S. C., Levin, M. E., Plumb-Vilardaga, J., Villatte, J. L., & Pistorello, J. (2013). Acceptance and commitment therapy and contextual behavioral science: Examining the progress of a distinctive model of behavioral and cognitive therapy. Behavior Therapy, 44(2), 180–198. https://doi.org/10.1016/j.beth.2009.08.002
  • Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., & Lillis, J. (2006). Acceptance and commitment therapy: Model, processes and outcomes. Behaviour Research and Therapy, 44(1), 1–25. https://doi.org/10.1016/j.brat.2005.06.006
  • Hayes, S. C., Pistorello, J., & Levin, M. E. (2012). Acceptance and commitment therapy as a unified model of behavior change. The Counseling Psychologist, 40(7), 976–1002. https://doi.org/10.1177/0011000012460836
  • Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). Acceptance and commitment therapy: An experiential approach to behavior change. Guilford Press. https://contextualscience.org/publications/hayes_strosahl_wilson_1999
  • Hayes, S. C., Villatte, M., Levin, M., & Hildebrandt, M. (2011). Open, aware, and active: Contextual approaches as an emerging trend in the behavioral and cognitive therapies. Annual Review of Clinical Psychology, 7, 141–168. https://doi.org/10.1146/annurev-clinpsy-032210-104449
  • Hofmann, S. G., & Hayes, S. C. (2019). The future of intervention science: Process-based therapy. Clinical Psychological Science, 7(1), 37–50. https://doi.org/10.1177/2167702618772296
  • Kashdan, T. B., & Rottenberg, J. (2010). Psychological flexibility as a fundamental aspect of health. Clinical Psychology Review, 30(7), 865–878. https://doi.org/10.1016/j.cpr.2010.03.001
  • Kong, Q., Yan, S., Huang, K., Han, B., Han, R., Jiao, Y., Yang, H., Pu, Y., Li, S., & Jia, Y. (2025). The efficacy of acceptance and commitment therapy (ACT) for depression: A systematic review and meta-analysis. Psychiatry Research, 352, Article 116701. https://doi.org/10.1016/j.psychres.2025.116701
  • Levin, M. E., Hildebrandt, M. J., Lillis, J., & Hayes, S. C. (2012). The impact of treatment components suggested by the psychological flexibility model: A meta-analysis of laboratory-based component studies. Behavior Therapy, 43(4), 741–756. https://doi.org/10.1016/j.beth.2012.05.003
  • McCracken, L. M., & Vowles, K. E. (2014). Acceptance and commitment therapy and mindfulness for chronic pain: Model, process, and progress. American Psychologist, 69(2), 178–187. https://doi.org/10.1037/a0035623
  • Ong, C. W., Pierce, B. G., Petersen, J. M., Barney, J. L., Fruge, J. E., Levin, M. E., & Twohig, M. P. (2020). A psychometric comparison of psychological inflexibility measures: Discriminant validity and item performance. Journal of Contextual Behavioral Science, 18, 34–47. https://doi.org/10.1016/j.jcbs.2020.08.007
  • Rolffs, J. L., Rogge, R. D., & Wilson, K. G. (2018). Disentangling components of flexibility via the hexaflex model: Development and validation of the Multidimensional Psychological Flexibility Inventory (MPFI). Assessment, 25(4), 458–482. https://doi.org/10.1177/1073191116645905
  • Ruiz, F. J., Bianchi, J. M., Bastidas-Suárez, D. M., Ramírez, E. S., & Peña-Hernández, V. (2024). Is the AAQ-II that bad? Journal of Contextual Behavioral Science, 34, Article 100854. https://doi.org/10.1016/j.jcbs.2024.100854
  • Trindade, I. A., Guiomar, R., Carvalho, S. A., Duarte, J., Lapa, T., Menezes, P., Nogueira, M. R., Patrão, B., Pinto-Gouveia, J., & Castilho, P. (2021). Efficacy of online-based acceptance and commitment therapy for chronic pain: A systematic review and meta-analysis. The Journal of Pain, 22(11), 1328–1342. https://doi.org/10.1016/j.jpain.2021.04.003
  • Twohig, M. P., Hayes, S. C., Plumb, J. C., Pruitt, L. D., Collins, A. B., Hazlett-Stevens, H., & Woidneck, M. R. (2010). A randomized clinical trial of acceptance and commitment therapy versus progressive relaxation training for obsessive-compulsive disorder. Journal of Consulting and Clinical Psychology, 78(5), 705–716. https://doi.org/10.1037/a0020508
  • Tyndall, I., Waldeck, D., Pancani, L., Whelan, R., Roche, B., & Dawson, D. L. (2019). The Acceptance and Action Questionnaire-II (AAQ-II) as a measure of experiential avoidance: Concerns over discriminant validity. Journal of Contextual Behavioral Science, 12, 278–284. https://doi.org/10.1016/j.jcbs.2018.09.005
  • Villatte, M., Villatte, J. L., & Hayes, S. C. (2016). Mastering the clinical conversation: Language as intervention. Guilford Press. https://www.guilford.com/books/Mastering-the-Clinical-Conversation/Villatte-Villatte-Hayes/9781462542161
  • Wetherell, J. L., Afari, N., Rutledge, T., Sorrell, J. T., Stoddard, J. A., Petkus, A. J., Solomon, B. C., Lehman, D. H., Liu, L., Lang, A. J., & Atkinson, J. H. (2011). A randomized, controlled trial of acceptance and commitment therapy and cognitive-behavioral therapy for chronic pain. Pain, 152(9), 2098–2107. https://doi.org/10.1016/j.pain.2011.05.016
  • Zhang, Y., Zhao, B., Wang, Q., Ma, T., Xu, M., Chen, J., & Cheng, X. (2026). Effect of acceptance and commitment therapy for anxiety disorders: A systematic review and meta-analysis. BMC Psychiatry. Advance online publication. https://doi.org/10.1186/s12888-026-08804-5