Self-Efficacy and Security Behavior: Results from a Systematic Review of Research Methods

Title of the Paper

Self-Efficacy and Security Behavior: Results from a Systematic Review of Research Methods

Paper Information

• Subject Area: Cybersecurity, Psychology, and User Behavior
• Keywords: Self-efficacy, Cybersecurity, Systematic review, Research methods, Security behavior, Psychometrics

Research Background and Issues

• Issues or Challenges:

  • The psychological foundation of cybersecurity behavior has yet to be standardized, particularly in the measurement and theoretical mechanisms of "self-efficacy."
  • Inconsistent definitions and methods in existing literature lead to fragmented data and difficulties in reproducing research findings.

• Importance of the Research:

  • Data privacy and IT security have become key public concerns, and the formation of secure habits lies at the intersection of technology and psychology.
  • Self-efficacy is considered the only psychological variable that significantly predicts cybersecurity behavior, making it crucial to study its role and operational potential.

• Motivation and Related Work:

  • Building on previous studies (e.g., He et al., 2014), this paper aims to comprehensively evaluate research methods and existing issues regarding self-efficacy in the cybersecurity domain.

Solution

• Proposed Methods or Solutions:

  • This paper systematically reviewed 174 studies conducted between 2010 and 2021, analyzing:
    1. The quality and methods of measuring cybersecurity self-efficacy.
    2. The role of self-efficacy in theoretical frameworks.
    3. Intervention designs targeting self-efficacy.
  • Multi-database searches, dual coding, and strict exclusion criteria were employed to minimize selection bias.

• Innovative Contributions:

  • Conducted a systematic evaluation of the fragmentation issue across interdisciplinary literature and proposed recommendations for data standardization.
  • Provided an extensive variable analysis, categorizing variables as "causes," "outcomes," or "both."

• Implementation Steps and Techniques:

  • Data collection utilized 18 databases, covering both quantitative and qualitative studies.
  • Variables coded included sample information, study types, measurement tools, psychometric properties, and intervention methods.
  • Both quantitative and qualitative analysis methods were applied, including data aggregation, network analysis, and exploratory statistics.

Research Findings

• Specific Findings:

  • Identified 173 unique tools for measuring cybersecurity self-efficacy, most of which were used only once.
  • Weak psychometric standards: While average reliability was high, validation analyses (e.g., validity testing) were generally lacking.
  • Extracted and analyzed 55 outcome variables and 51 causal variables, highlighting the diverse yet non-standardized roles of self-efficacy.
  • Only 13 studies designed interventions targeting self-efficacy, none of which were replicated.

• Comparison with Existing Solutions:

  • Compared to previous reviews (e.g., He et al., 2014), the issues remain unresolved, with measurement tools becoming even more diverse and fragmented.
  • Emphasized the need for improved consistency in psychometric standards and integration of theoretical frameworks.

• Experimental or Evaluation Results:

  • The weighted average Cronbach α coefficient for newly introduced scales was approximately 0.87, indicating good reliability.
  • Intervention methods primarily focused on training and activity design, suggesting "mastery experiences" as the main mechanism of action.
  • Theoretical frameworks lacked consensus in variable definitions, with multiple pathways or dual causal relationships observed among variables.

• Limitations and Future Directions:

  • This paper only evaluated data up to 2021, requiring updates over time.
  • Transparency and detailed reporting in studies were insufficient, particularly regarding scale development processes and intervention mechanisms.
  • Future work should focus on scale construction, theoretical simplification, and replication of interventions.