This study conducted a psychometric evaluation of a 20-item short form of the Cybersecurity Resilience Scale (CSRS-20), designed to reduce respondent burden while retaining the core conceptual coverage. The short form focuses on individual-level cybersecurity resilience (CSR), defined as the capacity to recognize, respond to and recover from cybersecurity threats through increasingly advanced protective behaviors. The purpose of this study was to conduct a psychometric evaluation of a 20-item short form of the Cybersecurity Resilience Scale (CSRS-20).
A two-stage validation procedure was used. First, item reduction and exploratory analyses with a Turkish sample (n = 805) produced a 20-item version. Because the advanced behavioral domains did not separate empirically once the item pool was reduced, the short form resolved into three dimensions rather than the five of the parent scale. Second, confirmatory factor analysis (CFA) with an independent international sample (n = 1,040) from Pakistan, Kuwait and Türkiye tested the stability of this three-factor structure. Additional analyses assessed reliability, convergent validity, discriminant validity, measurement invariance and nomological validity.
A three-factor structure fit the data well and outperformed one-factor and bifactor alternatives. Internal consistency was strong for all three dimensions. Convergent and discriminant validity were supported, and the reduction to three dimensions resolved the discriminant-validity overlap observed among the advanced domains of the full-length scale. Multi-group CFA supported configural and metric invariance across the three national samples, suggesting that the factor structure and factor loadings were comparable across groups. Nomological analyses showed theoretically consistent associations between CSRS-20 scores, theory of planned behavior constructs and perceptions of educational sustainability in AI-supported learning contexts.
The CSRS-20 offers a concise, three-dimensional short form of the parent CSRS for settings where a brief measure of individual CSR is needed. Rather than reproducing the five domains of the full scale, it captures three empirically robust dimensions and provides initial multi-country evidence in young-adult, university-based samples. It is particularly suited to multi-construct survey designs where questionnaire length is a constraint.
