Decomposed TPB - Taylor & Todd (1995)

Theory Publication Information

Citation Information

Why Was the Model Created?

Taylor and Todd developed the Decomposed Theory of Planned Behavior to advance beyond the original Theory of Planned Behavior (TPB) in predicting IT adoption and usage. While Ajzen’s TPB was a robust general theory of human behavior, it operated at a relatively high level of abstraction with broad categories of beliefs (attitudes, subjective norms, and perceived control). Taylor and Todd recognized that IT adoption required more technology-specific theorizing with decomposed belief constructs that captured the unique facets influencing information systems decisions.

The authors observed that prior IT adoption models like TAM offered narrower attitudinal constructs (usefulness and ease of use) but lacked the social (subjective norm) and control (perceived behavioral control) determinants of intention present in TPB. They conceptualized DTPB as decomposing the three TPB belief categories into multiple technology-specific dimensions. Attitudinal beliefs decompose into perceived usefulness, perceived ease of use (complexity), and compatibility (from Rogers); normative beliefs decompose into peer influence and superior influence; control beliefs decompose into self-efficacy, resource facilitating conditions, and technology facilitating conditions.

The empirical research compared three models (TAM, TPB, DTPB) using 786 university students at a business school computing facility. This large-scale study enabled comprehensive model comparison revealing that DTPB provided superior fit and explained more variance in IT usage intentions than both TAM and the original TPB. The research demonstrated that decomposing general behavioral constructs into technology-specific dimensions substantially improved prediction of IT adoption behavior.

Core Concepts and Definitions

The DTPB decomposes the three TPB belief structures into technology-specific multi-dimensional constructs:

Attitudinal Belief Decomposition

• Perceived Usefulness (Relative Advantage):The degree to which using the system enhances job performance. Analogous to Rogers’ relative advantage and Davis’ perceived usefulness.
• Perceived Ease of Use (Complexity):The degree to which the system is perceived as easy to understand and use. Analogous to Rogers’ complexity (inverse direction) and Davis’ ease of use.
• Compatibility:The degree to which the system fits with the potential adopter’s existing values, previous experience, and current needs. From Rogers’ (1983) innovation characteristics.

Normative Belief Decomposition

• Peer Influence:Influence of peers (colleagues, other students) on the individual’s adoption decision.
• Superior Influence:Influence of superiors (managers, professors) on the individual’s adoption decision.

Control Belief Decomposition

• Self-Efficacy:The individual’s confidence in their ability to use the system, based on Bandura’s (1977) self-efficacy theory.
• Resource Facilitating Conditions: Availability of resources needed to use the system (time, money), based on Triandis’ (1979) facilitating conditions concept.
• Technology Facilitating Conditions: Compatibility issues that may constrain usage (such as a lack of appropriate hardware, software, or technology support). Taylor and Todd separate these from resource conditions because technology compatibility operates independently of resource availability in the empirical data.

What Does the Model Measure?

Taylor and Todd (1995) decompose each TPB belief structure into multiple technology-specific constructs. The measured constructs are:

• Attitude (from decomposed attitudinal beliefs):
  • Perceived Usefulness: Rogers-style relative advantage / TAM PU adapted to the technology context.
  • Perceived Ease of Use: Davis 1989 PEOU.
  • Compatibility: Rogers 1983 compatibility with values, experiences, and needs.
• Subjective Norm (from decomposed normative beliefs):
  • Peer Influence: Influence of colleagues.
  • Superior’s Influence: Influence of supervisors/managers.
• Perceived Behavioral Control (from decomposed control beliefs):
  • Self-Efficacy:Bandura-style belief in one’s capability.
  • Resource Facilitating Conditions: Access to money, time, and other resources.
  • Technology Facilitating Conditions: Access to compatible technology and support.
• Behavioral Intention and Behavior: Standard TPB dependent measures.

Taylor and Todd (1995) report a field study of information center users; they provide reliability and validity evidence for the decomposed scales and compare the decomposed TPB with TAM and pure TPB.

Preceding Models or Theories

The DTPB synthesized multiple prior theoretical traditions:

• Theory of Planned Behavior (Ajzen, 1991): Foundational framework establishing attitude, subjective norm, and perceived behavioral control as intention predictors.
• Technology Acceptance Model (Davis, 1989): IS-specific model demonstrating perceived usefulness and ease of use effectiveness in predicting IT adoption.
• Diffusion of Innovations (Rogers, 1983): Provided compatibility construct and concepts of relative advantage relevant to technology adoption.

Describe the Model

DTPB specifies that behavioral intentions to use IT are determined by attitudes, subjective norms, and perceived behavioral control. Critically, each of these three intention predictors is decomposed into multiple specific belief dimensions. Attitudes toward IT use are predicted by perceived usefulness, perceived ease of use (complexity), and compatibility beliefs. Normative beliefs are predicted by peer influence and superior influence. Perceived behavioral control is predicted by self-efficacy, resource facilitating conditions, and technology facilitating conditions. Per Table 3, the model explained 60% of variance in behavioral intention (R²=.60), compared to TAM (R²=.52) and original TPB (R²=.57). The study also measured actual usage behavior by monitoring 3,780 visits to the computing resource center over a 12-week period (Methods, p.157).

What does the model measure?

• Attitudinal beliefs: Perceived usefulness, ease of use (complexity), and compatibility capturing instrumental, cognitive, and value-alignment motivations for adoption.
• Normative beliefs: Peer influence and superior influence capturing horizontal and vertical social influence pathways.
• Control beliefs: Self-efficacy, resource facilitating conditions, and technology facilitating conditions capturing individual capability confidence, resource availability, and technology compatibility constraints.
• Behavioral intentions: Likelihood of system adoption and usage predicted through multiple belief pathways.

Main Strengths

• Theoretical comprehensiveness: Integrates three distinct theoretical traditions (TPB, TAM, diffusion theory) creating more complete framework than any single theory.
• Technology-specific decomposition: Decomposes general TPB constructs into IT-relevant dimensions (usefulness, ease of use/complexity, compatibility, peer/superior influence, self-efficacy, resource and technology facilitating conditions) improving precision.
• Superior predictive power: Explains 60% of intention variance (R²=.60), outperforming TAM (R²=.52) and original TPB (R²=.57) in this sample.
• Multi-pathway modeling: Recognizes attitudes, norms, and control operate through separate pathways, none fully mediating others.
• Large-scale validation: Tested with 786 university users (582 undergraduate, 204 MBA) at a single business school computing resource center, providing adequate statistical power for complex model estimation and direct comparison of three competing theoretical models in the same sample.
• Comparative design: Direct comparison with competing theories (TAM and TPB) in same sample strengthens conclusions about relative model performance.
• Organizational relevance: Includes peer/superior influence and resource facilitating conditions reflecting real organizational adoption contexts.

Main Weaknesses

• University laboratory context: Study conducted in university computing facility with voluntary student users, limiting generalization to mandatory workplace contexts.
• Single-time measurement: Cross-sectional design at single time point prevents understanding of actual adoption trajectories or sustained usage patterns.
• Intention-behavior gap: While actual usage was monitored (3,780 visits over 12 weeks by 451 of 786 respondents), the decomposed belief structure primarily predicts intentions. The link from intention to actual behavior showed moderate strength.
• Model complexity:13-variable decomposed model (per Taylor & Todd, 1995, p.169) with multiple paths is substantially more complex than simpler alternatives like the 5-variable TAM, and only improves behavior variance explained by 2% over TAM (36% vs. 34%), raising a parsimony vs. understanding tradeoff.
• Limited moderator exploration: Does not examine whether belief-intention relationships differ by user experience, task type, or system characteristics.
• Self-report limitations: The belief and intention constructs relied on self-reported perceptions, although actual usage was tracked separately through system visit logs rather than direct behavioral observation.
• Sample homogeneity: Student sample may not represent diverse organizational roles, skill levels, and adoption motivations of heterogeneous workforces.

Key Contributions

• Theoretical integration model:Illustrated how to combine TPB’s behavioral-intention framework with TAM’s IT-specific constructs into a unified model.
• Decomposition methodology: Articulated an approach for making general behavioral theories technology-specific through construct decomposition.
• Superior predictive framework: Provided empirical evidence that decomposed, technology-specific theories outperform both TAM and original TPB.
• Multi-belief-pathway recognition: Articulated how attitudes, normative pressures, and control beliefs operate as independent intention predictors.
• Decomposed social and control constructs: Clarified how normative beliefs split into peer and superior influence, while control beliefs split into self-efficacy, resource facilitating conditions, and technology facilitating conditions in technology adoption models.
• IT adoption theory advancement: Created framework synthesizing behavioral intention theory with information systems adoption insights.

Internal Validity

The researchers established internal validity through comprehensive measurement and structural equation modeling:

• Multi-item measurement scales: Operationalized all constructs with multiple survey items enabling measurement error and reliability estimation.
• Reliability assessment:Reported Cronbach’s alpha coefficients for all 13 scales (Table 1). Ten scales were above the conventional 0.70 threshold, but three scales fell below: PBC (α=0.68), Ease of Use (α=0.60), and Resource Facilitating Conditions (α=0.52), limiting confidence in those constructs.
• Discriminant validity: Provided evidence that model constructs are empirically distinct through correlation analysis and variance extracted comparisons.
• Structural equation modeling:Used LISREL 8 with weighted least squares (WLS) estimation (Joreskog & Sorbom, 1993) for simultaneous estimation of measurement and structural models, with fit assessed by χ², AGFI, RNI, and RMSEA (p.158).
• Model fit comparison: Systematically compared three competing models (TAM, TPB, DTPB) using chi-square and fit indices enabling model selection.
• Variance explained: Reported R-squared values showing 60% of intention variance explained by DTPB model (Table 3).

External Validity

External validity considerations require acknowledging both strengths and limitations:

• University sample composition: Students may be younger, more educated, and more computer-proficient than typical organizational IT users.
• Voluntary adoption context: University lab context with optional system use differs from mandatory workplace IT implementations.
• Single-system limitation: Tested on only one computing resource center (document/presentation production facility) in a single university setting, limiting generalization to other types of IT systems and organizational contexts.
• Structural generalization: Three-component belief structure of TPB is theoretically robust and generalizable across populations and technologies.
• Cross-national applicability: While specific belief dimensions may vary, the decomposition approach transfers to different cultural contexts.
• Temporal generalizability: Cross-sectional design limits understanding of whether belief-intention relationships persist over time or change as use experience accumulates.

Relevance to Technology Adoption

DTPB is directly relevant to technology adoption because it identifies multiple specific barriers and levers operating through distinct belief pathways. The model suggests that adoption strategies must address not only instrumental utility but also peer and superior influences, self-efficacy, resource facilitating conditions, and technology facilitating conditions simultaneously.

Barriers to Technology Adoption Identified

• Low perceived usefulness: When users believe systems do not enhance job performance or outcomes, attitude-based resistance emerges.
• Poor perceived compatibility: Systems perceived as misaligned with existing work practices and values face adoption friction despite potential utility.
• Weak social support: Absence of peer adoption, manager endorsement, or opinion leader influence undermines normative adoption motivation.
• Insufficient resource facilitating conditions: Inadequate training, technical help, equipment access, and management support reduce adoption likelihood by undermining perceived behavioral control.
• Perceived high complexity: Systems requiring substantial learning effort create control barriers reducing adoption intentions.
• Negative social influence: Peer skepticism, visible non-adoption by respected users, and management indifference create normative barriers.

Leadership Actions the Model Prescribes

• Establish clear usefulness narrative: Communicate specific productivity gains, time savings, or capability enhancements building usefulness beliefs.
• Demonstrate compatibility fit: Show how systems integrate with existing workflows, tools, and job tasks reducing friction and workflow disruption.
• Mobilize social influence: Identify opinion leaders, early adopters, and respected figures to publicly endorse and visibly use systems.
• Secure visible management commitment: Ensure managers and executives visibly adopt and use systems, creating strong normative pressure.
• Provide comprehensive facilitating conditions: Ensure training programs, help desk support, documentation, and easy access to required resources.
• Reduce perceived complexity: Design intuitive interfaces, provide hands-on practice, and create clear learning pathways reducing cognitive burden.
• Implement multi-level strategies: Address attitudinal barriers (usefulness, compatibility, complexity, ease-of-use), normative barriers (peer and superior influences), and control barriers (self-efficacy, resource facilitating conditions, and technology facilitating conditions) simultaneously.

Following Models or Theories

Taylor and Todd’s decomposed approach influenced subsequent IT adoption research directions:

• TAM 2 (Venkatesh & Davis, 2000):Extended TAM with social influence and cognitive instrumental processes, reflecting DTPB’s more differentiated treatment of adoption determinants.
• TAM 3 (Venkatesh & Bala, 2008):Further elaborated TAM by decomposing determinants of perceived usefulness and perceived ease of use in ways that parallel DTPB’s decomposition approach.
• Unified Theory of Acceptance and Use of Technology (UTAUT): Integrated multiple theories including TPB, TAM, and DTPB insights into comprehensive model.
• Technology Readiness Index expansions: Applied decomposition methodology to user predispositions and technology readiness dimensions.
• Implementation science frameworks: Adopted decomposed multi-pathway approach examining fidelity, adoption, and sustainability determinants.
• Organizational change models: Incorporated multi-level decomposition acknowledging individual beliefs, group norms, and organizational structures.

References

1. Taylor, S., & Todd, P. A. (1995). Understanding information technology usage: A test of competing models. Information Systems Research, 6(2), 144-176. https://doi.org/10.1287/isre.6.2.144
2. Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211.↩︎ https://doi.org/10.1016/0749-5978(91)90020-T
3. Davis, F. D. (1989). Perceived usefulness, perceived ease of use, and user acceptance of information technology. MIS Quarterly, 13(3), 319-340.↩︎ https://doi.org/10.2307/249008
4. Rogers, E. M. (1983). Diffusion of innovations (3rd ed.). Free Press.↩︎
5. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191-215.↩︎ https://doi.org/10.1037/0033-295X.84.2.191
6. Triandis, H. C. (1979). Values, attitudes, and interpersonal behavior. In H. E. Howe & M. M. Page (Eds.), Nebraska Symposium on Motivation, Vol. 27 (pp. 195-259). University of Nebraska Press.↩︎

Further Reading

1. Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
2. Venkatesh, V., Morris, M. G., Davis, G. B., & Davis, F. D. (2003). User acceptance of information technology: Toward a unified view. MIS Quarterly, 27(3), 425-478. https://doi.org/10.2307/30036540
3. Davis, F. D., Bagozzi, R. P., & Warshaw, P. R. (1992). Extrinsic and intrinsic motivation to use computers in the workplace. Journal of Applied Social Psychology, 22(14), 1111-1132. https://doi.org/10.1111/j.1559-1816.1992.tb00945.x
4. Venkatesh, V., & Davis, F. D. (2000). A theoretical extension of the technology acceptance model: Four longitudinal field studies. Management Science, 46(2), 186-204.
5. Venkatesh, V., & Bala, H. (2008). Technology acceptance model 3 and a research agenda on interventions. Decision Sciences, 39(2), 273-315. https://doi.org/10.1111/j.1540-5915.2008.00192.x

Series Navigation