Two attitude clusters differ significantly on preparedness, confidence, and interest (p<.0001), with Cluster 1 lower on Pew and GSE but not SeBIS, and reporting more word-of-mouth learning¶

Claim · All claims
Evidence · 1 study (3 entries) · associational r1 · q2 · i0 negligible
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? K-means analysis identified two clusters of 119 and 117 participants differing significantly on all three attitude factors (p<.0001): one less prepared, quite unconfident, and very interested in learning more; the other prepared, confident, and somewhat interested. → Adama 2026 q2 i0 Cluster 1 scored significantly lower than Cluster 2 on Pew (r = 0.1) and GSE (d = −0.81), but the SeBIS difference was not significant (d = −0.15, p = .123). → Adama 2026 (2) q2 i? Cluster 1 reported significantly more word-of-mouth knowledge sourcing than Cluster 2 (χ2(1) = 4.641, p = 0.031, without multiple-test correction). → Adama 2026 (3)

Evidence¶

Adama 2026¶

Adama, V., Biddle, R., Arachchilage, N., & Lottridge, D. (2026). Confident yet Concerned: Inconsistencies in Computing Students' Attitudes on Cybersecurity. https://arxiv.org/abs/2606.18541 doi:10.48550/arxiv.2606.18541

q2 · i? · associational · r1

K-means cluster analysis of attitude factor scores for 236 participants, shown in Figure 8. The article describes the two clusters of 119 and 117 participants and reports "For each factor, the difference between clusters is significant (p<.0001)".

"In our sample, 119 feel prepared, confident and some- what interested in learning more and the other 117 feel less prepared, quite unconfident, and very interested in learning more."

Adama 2026 (2)¶

Adama, V., Biddle, R., Arachchilage, N., & Lottridge, D. (2026). Confident yet Concerned: Inconsistencies in Computing Students' Attitudes on Cybersecurity. https://arxiv.org/abs/2606.18541. doi:10.48550/arxiv.2606.18541

q2 · i0 · associational · r1

One-sided t-tests for SeBIS and GSE and a Wilcoxon test for Pew comparing clusters (Table 2). The article prints SeBIS d = −0.15 (p = .123, not significant), Pew r = 0.1 (p = 0.040), and GSE d = −0.81 (p < .001); the SeBIS scatter plot showed mixed scores in each cluster.

"The differences between clusters for Pew and GSE scores showed significant difference, but those for the SeBIS did not: see Table 2."

Adama 2026 (3)¶

Adama, V., Biddle, R., Arachchilage, N., & Lottridge, D. (2026). Confident yet Concerned: Inconsistencies in Computing Students' Attitudes on Cybersecurity. https://arxiv.org/abs/2606.18541. doi:10.48550/arxiv.2606.18541

q2 · i? · associational · r1

Chi-square tests of knowledge sources by cluster. The article reports word-of-mouth as the only significant source difference, "without correcting for multiple tests", and found no significant cluster differences for go-to persons, victimization, number of sources, gender, employment, or age.

"Of these, the only significant difference was for word-of-mouth, with Cluster 1 people reporting signifi- cantly more use than those in Cluster 2: WoM:χ 2(1) = 4.641,p= 0.031"

Discussion¶

Related Claims¶

• Computing students show high objective cybersecurity knowledge, with over 75% answering five or more of seven Pew questions correctly — related
• Computing students learn cybersecurity mainly from informal sources outside the classroom, with word of mouth, social media, and financial institutions most common — related
• Security intentions and objective knowledge correlate with preparedness, and self-efficacy correlates with confidence — related
• Self-efficacy predicts cluster membership, with the article concluding self-efficacy may play a foundational role in the cluster difference — related
• Computing students report limited cybersecurity assistance and substantial personal victimization, with 61% lacking a trusted person and 21% having fallen victim — related