Keywords
Email classification, task interruptions, eye tracking, phishing
Abstract
With the continuous rise in email use, the prevalence and sophistication of phishing attacks have increased. Expanding cybersecurity awareness and strengthening email practices will help reduce the dangers posed by phishing emails, but ultimately, the extent to which a user can accurately detect phishing emails directly impacts the amount of risk to which they are exposed. Being interrupted while reading and replying to emails is a consequence of working in a dynamic world. Interruptions are often identified to be disruptive, both in terms of time costs and performance changes; they reliably increase a task's completion time, but their impact on accuracy is less consistent. The present three studies manipulated the length (Experiment 1), difficulty (Experiment 2), and similarity (Experiment 3) of interruptions in accordance with the memory for goals (MFG) model, which aims to explain why interruptions may be disruptive. Participants classified emails as either phishing or legitimate, while periodically being interrupted with a secondary task. Across all three experiments, interruptions did not affect classification accuracy, but they did reliably increase classification response time. Oculomotor analyses indicated that interruptions, regardless of type, impaired memory of previously encoded email information. This was evidenced across all three experiments by an increase in refixations and an increase in the distance between fixations pre- and post-interruption. MFG can account for some of these findings, but not all. Interruptions did not impair performance on an email classification task when participants could review the interrupted information, yet overall classification accuracy was still low. These results may suggest a pathway toward improving email classification performance however, as participants exhibited behaviors known to improve performance on other tasks, such as revisiting previously viewed areas of an email.
Completion Date
2024
Semester
Spring
Committee Chair
Neider, Mark
Degree
Doctor of Philosophy (Ph.D.)
College
College of Sciences
Department
Psychology
Degree Program
Human Factors and Cognitive Psychology
Format
application/pdf
Identifier
DP0028327
URL
https://purls.library.ucf.edu/go/DP0028327
Language
English
Rights
In copyright
Release Date
May 2025
Length of Campus-only Access
1 year
Access Status
Doctoral Dissertation (Campus-only Access)
Campus Location
Orlando (Main) Campus
STARS Citation
Slifkin, Elisabeth, "Gone Phishing: How Task Interruptions Impact Email Classification Ability" (2024). Graduate Thesis and Dissertation 2023-2024. 158.
https://stars.library.ucf.edu/etd2023/158
Accessibility Status
Meets minimum standards for ETDs/HUTs
Restricted to the UCF community until May 2025; it will then be open access.