Cognitive Load and Phishing Detection in Trained Cybersecurity Students: A Dual-Task N-back Paradigm Validation Study

Read full text Summarise with VIRAT
Authors: Wafiy Ammar
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Phishing attacks remain one of the leading vectors in cybersecurity breaches worldwide, and Malaysia is no exception. In 2024, according to CyberSecurity Malaysia, phishing accounted for 77% of fraud cases within the country. Malaysia's response has been primarily reactive, empowering anti-scam call centres, while the main proactive measure for the general public has been awareness campaigns limited in both reach and impact. The central question this research addresses is: does technical cybersecurity training effectively shield practitioners from phishing under realistic cognitive conditions? Currently, no research appears to address this gap. This study aims to bridge it by examining whether N-back induced cognitive load degrades phishing detection accuracy amongst trained Malaysian cybersecurity students. The research also aims to validate the N-back task as an effective proxy for workplace-induced cognitive load in a security context. We present a within-subjects, repeated-measures, counterbalanced experimental design (N=40) recruiting students from a Malaysian technical university. Participants will evaluate 20 constructed email stimuli — 10 phishing and 10 legitimate — distributed across three difficulty tiers in a 3:4:3 ratio (Tier 1: predictable, Tier 2: intermediate, Tier 3: convincing). From the 20 emails, 10 will be evaluated under a concurrent auditory 2-back task, while the remaining 10 will be evaluated under a control condition, with a minimum 48-hour washout period between conditions. Participants are split into two counterbalanced groups, with one group completing the cognitive load condition first and the other completing the control condition first. The NASA-TLX survey will be administered post-condition to assess perceived cognitive load and validate the N-back task as a workplace load proxy. Data will be analysed using mixed-effects logistic regression as the primary statistical model, supplemented by repeated-measures ANOVA and paired response-time analysis. A pilot study will precede main data collection to assess baseline accuracy and calibrate stimulus difficulty against a target of 65–80% correct under control conditions, mitigating ceiling effects common to trained-sample research. Individual 2-back calibration will also be conducted prior to the experiment to control for working memory capacity differences across participants. This study constitutes, to the best of our knowledge, the first empirical investigation of N-back task validity as a cognitive load model for phishing detection in trained cybersecurity practitioners. Findings are expected to inform both behavioural cybersecurity theory and the design of cognitively realistic security training simulations.

Fields of Study & Subjects

Spam and Phishing Detection Information and Cyber Security User Authentication and Security Systems

Publication details

Publication date
1 January 2026
Language
EN
Document type
Preprint