Phishing interrupted: The impact of task interruptions on phishing email classification

Email is a pervasive form of communication in both personal and professional settings. The extent to which a human user can accurately detect phishing emails impacts the amount of risk they are exposed to. Previous research has unfortunately shown that people often fall victim to phishing attacks, both in controlled laboratory and naturalistic settings, even if they have received phishing awareness training. While the impact of numerous participant, email, and task characteristics on phishing email classification performance have been assessed, the impact of various environmental factors must be considered as well. Interruptions are a widespread occurrence in daily life and have been shown to negatively impact performance on many tasks. The present studies aim to explore the effect of task interruptions on phishing email classification. Participants performed a classification task where they categorized emails as either phishing or legitimate. Participants in both experiments were occasionally interrupted with either a secondary task to complete (Experiments 1 & 2) or a blank box (Experiment 2). Results of both experiments indicated higher phishing classification accuracy on interrupted trials and an increase in response time, roughly equal to the amount of time the email was shown prior to the interruption, regardless of the type of interruption. Participants also showed an unbiased response and were more sensitive to the task when interrupted. Our findings suggest that being interrupted during this phishing classification task may improve performance in this limited capacity, though this benefit may not be due to the interruptions directly.

Automated summary

Email is a widely used form of communication, but people often fall victim to phishing attacks. Interruptions have been shown to negatively impact performance on tasks, and this study aimed to explore the effect of interruptions on phishing email classification. Participants who were occasionally interrupted had higher accuracy in classifying phishing emails and had longer response times. The type of interruption did not affect the results. These findings suggest that interruptions during the phishing classification task may improve performance.