Category: Tier 2 - Emotional Traits Scale: 0.0 (low) to 1.0 (high)

Definition

Self-Efficacy measures belief in one's ability to do what is needed to succeed. In web contexts, it controls whether users think they can finish tasks, how many paths they try before quitting, and how they explain failures.

High self-efficacy users approach unfamiliar interfaces with confidence. They persist through challenges and see obstacles as solvable. Low self-efficacy users doubt their abilities. They quit tasks early and may avoid complex interactions entirely.

Research Foundation

Primary Citation

"Efficacy expectations determine how much effort people will expend and how long they will persist in the face of obstacles and aversive experiences. The stronger the perceived self-efficacy, the more active the efforts. Those who persist in subjectively threatening activities will eventually eliminate their fear." -- Bandura, A., 1977, p. 194

Full Citation (APA 7): Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191-215.

DOI: https://doi.org/10.1037/0033-295X.84.2.191

Supporting Research

"Computer self-efficacy was found to be a significant determinant of behavioral intention and perceived ease of use. Individuals with higher computer self-efficacy were more likely to use computers and perceived them as easier to use." -- Compeau, D.R., & Higgins, C.A., 1995, p. 192

Full Citation (APA 7): Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19(2), 189-211.

DOI: https://doi.org/10.2307/249688

Key Numerical Values

Metric	Value	Source
Persistence increase (high vs low)	3x more attempts	Bandura (1977)
Task completion rate difference	35-40% higher for high self-efficacy	Compeau & Higgins (1995)
Abandonment speed (low self-efficacy)	40% faster on first error	Derived from behavioral research
Computer Self-Efficacy Scale reliability	alpha = 0.95	Compeau & Higgins (1995)
Effort expenditure correlation	r = 0.62 with self-efficacy	Bandura (1977)

Behavioral Levels

Value	Label	Behaviors
0.0-0.2	Very Low	Abandons 40% faster on first error; avoids complex tasks entirely; says "I can't do this" internally; attributes all failures to personal inadequacy; seeks help immediately or gives up; unwilling to try unfamiliar UI patterns; clicks only on familiar elements; avoids forms with many required fields
0.2-0.4	Low	Hesitates before attempting new interactions; gives up after 1-2 failed attempts; blames self for unclear error messages; seeks external validation before proceeding; avoids "advanced" or "expert" features; prefers guided wizards over open-ended interfaces; may complete simple tasks but abandons at first complexity
0.4-0.6	Moderate	Attempts new interactions with some hesitation; tries 2-3 solution paths before seeking help; balanced attribution between self and system; willing to explore but needs periodic success to continue; can complete moderately complex tasks; may pause to plan approach before difficult sections
0.6-0.8	High	Approaches unfamiliar interfaces with confidence; tries 4-6 solution paths before abandoning; attributes failures to system issues or temporary obstacles; actively seeks solutions rather than help; comfortable with trial-and-error exploration; interprets error messages as debugging information; assumes tasks are achievable
0.8-1.0	Very High	Tries 6+ solution paths; views all tasks as solvable; treats errors as informative feedback; may override warnings believing they know better; enjoys mastering complex interfaces; assumes ability to complete any task; may underestimate actual difficulty leading to overconfident behavior; rarely seeks help even when warranted

Trait Implementation in CBrowser

Solution Path Attempts

CBrowser models self-efficacy through the number of alternative approaches attempted:

// Number of solution paths tried before abandoning
const solutionAttempts = Math.floor(1 + (selfEfficacy * 7));
// Low self-efficacy: 1-3 attempts
// High self-efficacy: 6-8 attempts

// Willingness to try unfamiliar elements
const explorationConfidence = 0.3 + (selfEfficacy * 0.6);
// Low: 30% base willingness
// High: 90% willingness

First-Error Response

// Speed of abandonment after first error
const firstErrorPersistence = 1 - (0.4 * (1 - selfEfficacy));
// Low self-efficacy: 40% reduction in persistence (abandons faster)
// High self-efficacy: minimal impact

// Attribution style after error
const selfBlameRatio = 0.7 - (selfEfficacy * 0.5);
// Low: 70% self-attribution ("I messed up")
// High: 20% self-attribution ("The interface is unclear")

Self-Efficacy State Tracking

interface SelfEfficacyState {
  currentEfficacy: number;      // Dynamic efficacy level (0-1)
  recentSuccesses: number;      // Count in current session
  recentFailures: number;       // Count in current session
  domainConfidence: Map<string, number>;  // Task-specific confidence
}

// Efficacy updates based on outcomes
function updateEfficacy(state: SelfEfficacyState, success: boolean): void {
  if (success) {
    state.currentEfficacy = Math.min(1, state.currentEfficacy + 0.05);
    state.recentSuccesses++;
  } else {
    state.currentEfficacy = Math.max(0, state.currentEfficacy - 0.08);
    state.recentFailures++;
  }
}

Estimated Trait Correlations

Correlation estimates are derived from related research findings and theoretical models. Empirical calibration is planned (GitHub #95).

Research and theory suggest these correlations:

Related Trait	Correlation	Research Basis
Resilience	r = 0.56	Both serve as protective factors against failure impact
Persistence	r = 0.48	Self-efficacy fuels sustained effort (Bandura, 1977)
Risk Tolerance	r = 0.42	Confident users take more interface risks
Comprehension	r = 0.35	Some correlation; competence builds confidence
Curiosity	r = 0.38	Confident users explore more freely
Anxiety (inverse)	r = -0.52	Self-efficacy buffers against performance anxiety

Interaction Effects

Self-Efficacy x Comprehension: High efficacy + low comprehension creates overconfident users who attempt tasks beyond their ability
Self-Efficacy x Patience: Low efficacy + high patience may lead to prolonged ineffective attempts without trying alternatives
Self-Efficacy x Resilience: Combined high values create maximally persistent users

Persona Values

Persona	Self-Efficacy Value	Rationale
power-user	0.85	Experts have extensive mastery experiences building confidence
first-timer	0.35	No prior success to build confidence; uncertain of abilities
elderly-user	0.40	May doubt abilities with "modern" technology despite other competencies
impatient-user	0.55	Moderate; impatience not related to self-doubt
mobile-user	0.60	Familiar with touch interfaces; moderate confidence
screen-reader-user	0.70	Developed high competence navigating accessibility challenges
anxious-user	0.25	Anxiety undermines belief in ability to succeed
skeptical-user	0.50	Skepticism about sites, not about own abilities

UX Design Implications

For Low Self-Efficacy Users (< 0.4)

Early wins: Design easy initial steps that build confidence
Progress indicators: Show how far they've come to reinforce capability
External attribution: Error messages should blame the system, not the user
Guided paths: Provide step-by-step wizards instead of open interfaces
Social proof: Show that others successfully completed the task
Help accessibility: Make help easily visible without stigma

For High Self-Efficacy Users (> 0.7)

Challenge engagement: Provide complex options for those who seek them
Autonomy: Allow skipping tutorials and guided flows
Power features: Surface advanced capabilities
Warning calibration: Ensure warnings are credible; overconfident users may dismiss weak warnings
Error details: Provide technical information for self-diagnosis

Sources of Self-Efficacy (Bandura, 1977)

Design interventions can leverage the four sources:

Source	Description	UX Application
Mastery experiences	Prior successes at similar tasks	Progressive complexity, early wins
Vicarious experience	Observing others succeed	Video demos, user testimonials
Verbal persuasion	Encouragement from others	Encouraging microcopy, supportive error messages
Physiological states	Reduced anxiety and stress	Calm visual design, clear layouts

Bibliography

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

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Prentice-Hall.

Bandura, A. (1997). Self-efficacy: The exercise of control. W.H. Freeman.

Compeau, D. R., & Higgins, C. A. (1995). Computer self-efficacy: Development of a measure and initial test. MIS Quarterly, 19(2), 189-211. https://doi.org/10.2307/249688

Gist, M. E., & Mitchell, T. R. (1992). Self-efficacy: A theoretical analysis of its determinants and malleability. Academy of Management Review, 17(2), 183-211. https://doi.org/10.5465/amr.1992.4279530

Marakas, G. M., Yi, M. Y., & Johnson, R. D. (1998). The multilevel and multifaceted character of computer self-efficacy: Toward clarification of the construct and an integrative framework for research. Information Systems Research, 9(2), 126-163. https://doi.org/10.1287/isre.9.2.126

Stajkovic, A. D., & Luthans, F. (1998). Self-efficacy and work-related performance: A meta-analysis. Psychological Bulletin, 124(2), 240-261. https://doi.org/10.1037/0033-2909.124.2.240

Venkatesh, V. (2000). Determinants of perceived ease of use: Integrating control, intrinsic motivation, and emotion into the technology acceptance model. Information Systems Research, 11(4), 342-365. https://doi.org/10.1287/isre.11.4.342.11872

Copyright: (c) 2026 Alexa Eden.

License: MIT License

Contact: [email protected]

Trait SelfEfficacy