Who Is At Risk? Examining the Prevalence of Digital-Safety Attacks and Contextual Risk Factors in the United States

Paper Title

Who Is At Risk? Examining the Prevalence of Digital-Safety Attacks and Contextual Risk Factors in the United States

Publication Info

• Topic area: Digital safety, risk factors, and population-level analysis in the U.S.
• Keywords: Digital-safety attacks, contextual risk factors, cybersecurity, online harassment, scams, account hacking, population-level survey, risk perception, HCI, U.S. adults

Background and Problem

• Problem / challenge: Despite qualitative insights into digital-safety risks, there is a lack of quantitative data on how contextual risk factors are distributed across populations, limiting targeted interventions for improving digital safety.
• Significance: Understanding the prevalence and distribution of risk factors is crucial for policymakers and technology designers to create effective, evidence-based safety measures.
• Motivation and related work: Prior research has identified ten contextual risk factors that elevate digital-safety risks, including societal, relational, and personal circumstances. However, existing studies lack population-level data and focus primarily on qualitative insights, leaving gaps in understanding the generalizability of these risks.

Solution

• Proposed approach: A nationally representative survey of 5,001 U.S. adults to measure the prevalence of digital-safety attacks and self-identification with ten contextual risk factors.
• Novelty:
  1. First large-scale quantitative analysis linking contextual risk factors to digital-safety attacks.
  2. Identification of demographic groups disproportionately affected by digital-safety risks.
  3. Expansion of qualitative insights into how contextual risk factors are interpreted by individuals.
  4. Logistic regression modeling to assess correlations between risk factors, demographics, and attack likelihood.
• Procedure and key techniques:
  • Survey design included modules on demographics, digital-safety attacks, and contextual risk factors.
  • Respondents rated their agreement with 18 statements related to risk factors on a Likert scale.
  • Logistic regression and qualitative coding were applied to analyze correlations and interpret open-ended responses.

Results

• Concrete findings:
  • 74% of respondents experienced at least one digital-safety attack, with scams (49.7%), account hacking (41.3%), and exposure to explicit content (32.8%) being most common.
  • 85% identified with at least one contextual risk factor; the most prevalent were being resource/time-constrained (47.9%), reliant on a third party (38%), and having access to other at-risk users (34.4%).
  • Disabled individuals and younger adults (18–44) exhibited higher odds of experiencing attacks and identifying with multiple risk factors.
• Advantage over baselines:
  • Provides population-level data on contextual risk factors, addressing gaps in prior qualitative studies.
  • Demonstrates that digital-safety risks are a normative experience, not limited to edge cases.
• Experiments / evaluation:
  • Logistic regression models revealed significant correlations between risk factors and attack types (e.g., prominence correlated with stalking; resource constraints correlated with financial theft).
  • Qualitative analysis added depth to understanding how respondents interpret risk factors.
• Limitations and future work:
  • Self-reported data may be prone to recall bias.
  • Survey instrument needs further validation, especially for non-WEIRD contexts.
  • Future research should explore longitudinal changes in risk profiles and expand the framework globally.

Summary

This study provides the first quantitative analysis of contextual risk factors and their relationship to digital-safety attacks in a U.S. population. Key findings reveal that 74% of adults have experienced at least one attack, and 85% identify with at least one risk factor, with disabled individuals and younger adults facing heightened risks. Logistic regression models and qualitative insights highlight the interplay between demographics, risk factors, and attack likelihood. The results underscore the need for targeted, evidence-based interventions and methodological improvements in studying digital safety. Future work should validate the framework globally and explore dynamic risk profiles over time.