The Gender Gap and Cybercrime: An Examination of College Students’ Online Offending

Abstract

As computer and Internet technologies have continued to advance over the last few decades, a considerable amount of research has been dedicated to exploring online behavior. Criminologists, in particular, have been interested in examining cybercrime and online deviance. Contributing to this growing body of scholarship, the current research investigated the effects of gender on cybercrime and whether socialization can account for the gender gap in online offending. Data were collected from a sample of 522 college students and three cybercrime categories were examined: digital piracy, cyberharassment, and hacking offenses. The findings revealed that men were more likely to engage in online offending and that this gender gap was fairly persistent across the socialization variables of self-control and immersion into the cyberenvironment. Specifically, only three gender-invariant findings emerged: men and women commit statistically similar rates of digital piracy when self-control levels are low, and they also engage in similar rates of digital piracy and hacking offenses when Internet use is high. Detailed results, implications for theory and policy, study limitations, and directions for future research are discussed.

Keywords:

• cyberbullying
• gender
• offenders
• technology

ACKNOWLEDGMENT

I would like to thank Jerry Banfield and Wesley Jennings for their data collection and data analysis efforts. I would also like to thank guest editor Loretta Stalans and the anonymous reviewers who provided supportive and helpful comments for earlier versions of this manuscript.

Notes

1. While digital piracy and hacking offenses may both be committed for economic purposes (and, thus, could be paired together as a single, economic-based category of cybercrime), they are separated here because hacking offenses draw on a variety of other motivations (e.g., malicious intentions; see Yar, 2005).

2. Likelihood ratio alpha tests confirmed the use of negative binomial regression.

3. Median-splits were used for these models. For Internet use, the median was 5.00. For self-control, the median was 3.31.