Theory of reasoned action
The Theory of Reasoned Action (TRA) is a social psychological framework developed by Martin Fishbein and Icek Ajzen in 1975 to predict and explain volitional human behavior through the mediating role of behavioral intention.[1] The model posits that an individual's intention to engage in a specific behavior is primarily determined by two factors: their attitude toward performing the behavior, which reflects personal evaluations of the behavior's outcomes weighted by beliefs about those outcomes, and subjective norms, which capture perceived social pressures from relevant others weighted by motivation to comply with those pressures.[2] Mathematically, behavioral intention (BI) is expressed as BI = (attitude toward behavior) × w₁ + (subjective norm) × w₂, where w₁ and w₂ are empirically derived weights reflecting the relative importance of each component.[1] Under conditions of adequate control and stability, this intention directly predicts actual behavior, assuming the action is under volitional control and salient to the individual.[2] Empirical tests across diverse domains, including health behaviors like condom use and exercise adherence, have demonstrated that TRA accounts for substantial variance in intentions (often 40% or more) and, to a lesser extent, subsequent behaviors, with meta-analyses confirming its predictive utility in voluntary contexts.[3][4] The theory's emphasis on measurable beliefs and modifiable attitudes has facilitated targeted interventions in public health and consumer behavior, though its assumptions about volitional control have prompted extensions like Ajzen's 1991 Theory of Planned Behavior, which incorporates perceived behavioral control to address non-volitional influences.[2] While robustly supported for intention formation, TRA exhibits a consistent intention-behavior gap in real-world applications, attributable to factors such as unforeseen barriers or habit strength, underscoring the need for causal analyses beyond mere prediction.[3]Historical Development
Origins in Attitude-Behavior Research
In the mid-20th century, social psychologists widely assumed that attitudes directly influenced corresponding behaviors, a view rooted in foundational theories such as those from the Yale Communication Program in the 1950s, which emphasized persuasion and attitude change as precursors to action. However, accumulating empirical evidence challenged this assumption, revealing frequent discrepancies between expressed attitudes and observed behaviors across diverse domains like consumer choices, voting, and ethical decisions.[5] A pivotal synthesis came from Allan W. Wicker's 1969 review, which examined empirical studies from the prior decade and found a median correlation of only 0.15 between verbal attitudes and overt behaviors, with many coefficients near zero or negative. Wicker analyzed over 100 investigations, including church attendance surveys where pro-church attitudes correlated weakly with participation (r ≈ 0.10) and job satisfaction measures showing negligible links to turnover or productivity. This "attitude-behavior inconsistency" undermined confidence in attitude scales as reliable predictors and highlighted methodological issues, such as vague attitude targets and failure to account for situational constraints.[6][7] Martin Fishbein addressed these gaps through his expectancy-value formulation of attitudes, initially detailed in 1963 and refined in subsequent works, positing that attitudes reflect weighted beliefs about an object's attributes rather than simple evaluations. By the late 1960s, Fishbein contended that inconsistencies stemmed from measuring general object attitudes (e.g., toward a brand) instead of specific behavioral intentions (e.g., purchasing that brand next week under defined conditions). He proposed that behavioral intentions, formed via a reasoned evaluation of outcomes and normative pressures, serve as the immediate antecedent to action, thereby restoring predictive power when correspondence between measures is ensured. This reasoning, empirically tested in early studies like Fishbein's 1966 experiments on vaccination intentions, laid the groundwork for integrating attitudes, norms, and intentions into a cohesive model.[5][8]Formulation by Fishbein and Ajzen
The Theory of Reasoned Action was formally articulated by Martin Fishbein and Icek Ajzen in their 1975 book Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research, building on Fishbein's earlier expectancy-value models of attitudes to bridge the gap between attitudes and actual behavior.[1][9] The model specifies that voluntary behavior arises from reasoned deliberation, with behavioral intention serving as the proximal determinant of action, assuming volitional control, temporal stability of intentions, and absence of unforeseen barriers.[10] Fishbein and Ajzen emphasized that external variables, such as demographics or personality traits, influence behavior only indirectly through their effects on beliefs underlying attitudes and norms.[8] At the core of the formulation, behavioral intention (BI) is predicted by a linear combination of two constructs: the attitude toward the behavior (A_ACT) and the subjective norm (SN).[10] Attitude toward the behavior represents the individual's positive or negative evaluation of performing the action, calculated as the sum of products of salient behavioral beliefs (b_i, the subjective probability that the behavior leads to a specific outcome) and evaluations of those outcomes (e_i):A_ACT = Σ b_i e_i
This multiattribute formulation derives from Fishbein's prior work on expectancy-value theory, positing that attitudes stem from cognitively accessible beliefs about consequences rather than global object attitudes.[10]
The subjective norm captures perceived social pressure, computed as the sum of normative beliefs (NB_j, perception that referent others approve or disapprove) weighted by motivation to comply (MC_j):
SN = Σ NB_j MC_j
Fishbein and Ajzen's key equation integrates these as BI = w_1 A_ACT + w_2 SN, where weights w_1 and w_2 are regression-derived coefficients indicating the relative salience of personal attitudes versus social influences, often varying by behavior and population (e.g., attitudes typically weighting higher in individualistic contexts).[10] Empirical validation in the 1975 text drew from prior studies showing intention-behavior correlations around 0.63 across behaviors like voting and contraceptive use.[3] The formulation assumes rational actors weigh accessible beliefs and norms to form intentions, with behavior enactment requiring sufficient control; deviations occur if skills, resources, or habits intervene.[10] Fishbein and Ajzen operationalized measurement via Likert-scale surveys of beliefs (e.g., 5-7 salient outcomes identified via pilot elicitation), ensuring specificity to action, target, context, and time to enhance predictive validity.[11] This approach contrasted with earlier attitude models by focusing on behavior-specific constructs, addressing the "attitude-behavior inconsistency" observed in meta-analyses of pre-1975 research.[9]
Evolution Through the 1970s and 1980s
The Theory of Reasoned Action, formalized by Martin Fishbein and Icek Ajzen in their 1975 book Belief, Attitude, Intention and Behavior: An Introduction to Theory and Research, underwent initial empirical scrutiny in the late 1970s through targeted studies examining attitude-intention-behavior links in domains such as persuasion and consumer decision-making.[1] Early validations included Fishbein and Ajzen's 1973 review of 10 studies, which reported an average correlation of 0.63 between behavioral intentions and actual behaviors, underscoring the model's utility for volitional actions under conditions of sufficient control.[3] These efforts built on prior attitude research from the 1960s, emphasizing belief-based attitudes as proximal predictors while integrating subjective norms to account for social influences.[8] In 1980, Ajzen and Fishbein published Understanding Attitudes and Predicting Social Behavior, which synthesized accumulating evidence and refined operationalization guidelines for measuring components like belief strength and normative expectations, enhancing the model's applicability to health behaviors and moral actions.[12] Throughout the 1980s, the theory faced rigorous testing across diverse contexts, including contraceptive use, voting intentions, and ethical decision-making, with studies like those by Bagozzi (1982) confirming attitudes' derivation from belief evaluations.[8] A pivotal 1988 meta-analysis by Sheppard, Hartwick, and Warshaw aggregated findings from multiple investigations, revealing that intentions predicted behaviors with a correlation of 0.53, while attitudes and subjective norms jointly explained 66% of intention variance, affirming the framework's predictive power despite variability in non-volitional settings. Emerging critiques in the mid-1980s highlighted limitations in TRA's assumption of complete behavioral control, as evidenced by discrepancies in low-control scenarios like addiction or resource constraints, prompting preliminary refinements.[13] Ajzen's 1985 conceptualization of perceived behavioral control as an additional determinant marked an evolutionary bridge toward the Theory of Planned Behavior, though core TRA elements remained foundational for interventions in marketing and public health campaigns during the decade.[2] These developments reflected a shift from theoretical formulation to evidence-based iteration, with empirical correlations consistently outperforming earlier attitude models alone.[14]Core Theoretical Components
Attitudes Toward the Behavior
Attitudes toward the behavior in the Theory of Reasoned Action constitute an individual's overall positive or negative evaluation of performing a specific behavior, distinct from attitudes toward the behavioral target or object itself. Fishbein and Ajzen (1975) defined this construct as a function of salient behavioral beliefs—the perceived likelihood that the behavior will lead to particular outcomes—multiplied by the evaluations of those outcomes.[1] This specificity addresses prior inconsistencies in attitude-behavior relations by ensuring compatibility between the attitude measure and the behavior under consideration, such as the same action, target, context, and time frame.[12] The formation of these attitudes relies on a cognitive summation process: A = \sum b_i e_i, where b_i represents the subjective probability (ranging from -3 to +3, or similar scales) that the behavior produces outcome i, and e_i denotes the affective evaluation of that outcome (e.g., from -3 for undesirable to +3 for desirable).[8] Only salient or accessible beliefs, typically 5-9 per behavior as elicited through pre-testing, contribute meaningfully, emphasizing the role of readily available cognitions over exhaustive belief sets.[15] This belief-based approach aligns with expectancy-value models, positing that attitudes emerge from expected utilities rather than mere affective responses.[16] Empirical operationalization involves eliciting free responses to identify common outcomes, followed by scaling belief strengths and evaluations via Likert-type items, then computing the weighted sum. Studies applying this method, such as those reviewed by Fishbein and Ajzen, demonstrate that behavior-specific attitudes predict intentions with correlations often exceeding 0.50 when measurement correspondence is maintained.[3] However, the construct's predictive power diminishes if beliefs are not elicited from the relevant population or if external factors alter salient evaluations post-measurement.[10] In the broader TRA framework, these attitudes combine with subjective norms, weighted by individual importance (w_1), to form behavioral intentions, underscoring their proximal influence on volitional actions.[17]Subjective Norms
Subjective norms constitute one of the two primary determinants of behavioral intention in the Theory of Reasoned Action, capturing the perceived social pressures influencing an individual's decision to perform a behavior. Developed by Fishbein and Ajzen, subjective norm reflects the extent to which a person believes that significant others—such as family members, friends, or colleagues—approve or disapprove of the behavior, weighted by the person's motivation to align with those views.[18] This component emphasizes the interpersonal and normative influences on intention, distinct from personal evaluations captured by attitudes.[19] The construct decomposes into normative beliefs and motivation to comply. Normative beliefs represent the individual's subjective assessment of what specific referents expect regarding the behavior (e.g., "My spouse thinks I should exercise regularly"), typically rated on scales from disagree to agree. Motivation to comply measures the personal importance attached to each referent's expectations (e.g., "How much do I want to do what my spouse thinks I should do?"), often scaled from "not at all" to "very much."[10][20] Subjective norm is operationalized through an indirect multiplicative formula summing across salient referents:\text{SN} = \sum (b_i \times m_i)
where b_i is the normative belief for referent i and m_i is the corresponding motivation to comply; direct measures aggregate responses to items like "Most people important to me approve of my performing the behavior." Salient referents are elicited via open-ended questions in preliminary studies to ensure relevance to the target population and behavior.[18][20] In the model's behavioral intention equation, subjective norms contribute alongside attitudes, with regression-derived weights w_1 and w_2 indicating their relative predictive strength; empirical tests often reveal subjective norms accounting for 10-30% of intention variance, though this varies by domain, with stronger effects in socially constrained behaviors like condom use or voting.[18] The theory assumes these norms operate under volitional control, where perceived expectations causally shape intentions through reasoned compliance rather than automatic conformity.[21]