Exploring the Moderating Role of Gender in Juvenile Hacking Behaviors

Abstract

Criminological research has demonstrated the significant relationship between deviant peer associations, a lack of self-control, and individual delinquency. These relationships also account for involvement in cybercrime, though these results are based largely on adult samples. There is less research considering juvenile offending online, particularly examining involvement in property-based offenses such as computer hacking. This study utilized an international sample of 48,327 juvenile respondents in the Second International Self-Report of Delinquency (ISRD-2) study to examine the role of gender as a moderating factor in the relationship between deviant peer associations, self-control, opportunity, and self-reported computer hacking behavior. The findings demonstrated different correlates associated with hacking for males and females, as well as differences on the basis of urban and rural residency.

Keywords

cybercrime deviant peers differential association hacking low self-control

There is interest from policy-makers (National Crime Agency [NCA], 2017) and researchers as to the drivers influencing juvenile involvement in more serious forms of cybercrime, especially computer hacking (Holt & Bossler, 2016; Leukfeldt, 2017). Over the past three decades, research has consistently demonstrated the significant relationship between juvenile delinquency and two key criminological frameworks: the general theory of crime (Gottfredson & Hirschi, 1990) and social learning (Akers, 1998). These competing theories argue there are different factors that increase the likelihood of offending. Specifically, the general theory of crime is a neoclassical theory asserting that individuals with low self-control are more likely to act upon opportunities to engage in crime (Gottfredson & Hirschi, 1990). An individual’s level of self-control is formed in early childhood, with low self-control resulting from poor parental monitoring, recognition of deviance, and punishment of problematic behavior.

Social learning theory is rooted in Sutherland’s (1939) differential association framework, arguing that crime is a learned behavior gleaned primarily through interactions with deviant peers and intimate others. These associations provide individuals with models for offending that can be imitated, as well as definitions and justifications that reinforce involvement in delinquency and crime (Pratt et al., 2010). Criminological research has shown that association with and exposure to deviant peers is one of the most salient risk factors affecting both delinquent behavior and adult criminality (Pratt et al., 2010).

Despite the differences between these theories, both have received substantial empirical support to account for various forms of delinquency (Pratt & Cullen, 2000; Pratt et al., 2010; Vazsonyi, Mikuška, & Kelley, 2017), and cybercrimes, or offenses performed through Internet-connected devices (Holt, Bossler, & May, 2012; Holt, Burruss, & Bossler, 2010; Li, Holt, Bossler, & May, 2016; Marcum, Higgins, & Nicholson, 2017; Marcum, Higgins, Ricketts, & Wolfe, 2014). These studies also demonstrate that both self-control and deviant peer associations must be present in theoretical models to fully account for their dynamic relationships (Holt et al., 2012; Li et al., 2016; Marcum et al., 2017; Pratt & Cullen, 2000). At the same time, offenses enabled by technology present a unique challenge to our understanding of the general theory of crime, as the ubiquity of technology and access to information make opportunities to offend constant (Holt & Bossler, 2016). The complexity of cybercrimes and specialized knowledge required to complete the acts also necessitate the role of deviant peers who may be able to provide information as to how to offend (Bossler & Burruss, 2010; Skinner & Fream, 1997).

Although the relationships between these theories appear consistent within college student samples examining cybercrime, there is less research testing these dynamics within juvenile populations (Back, Soor, & LaPrade, 2018; Holt et al., 2012; Marcum et al., 2014; Udris, 2016). As youth gain access to computers, the Internet, and mobile devices at increasingly earlier ages, it is plausible that juveniles are more likely to engage in various forms of cybercrime. There is particular emphasis on the need for research examining cybercrimes that target data and computer systems like computer hacking, which involves the application of knowledge of computers and the Internet to gain access to systems and information with or without the owners’ permission (Holt, 2007; NCA, 2017; Steinmetz, 2015; Taylor, 1999). The ubiquity of technology and e-commerce enables hackers to directly compromise government and private industry, as well as to consumers (Schell & Dodge, 2002; Wall, 2007). In fact, hacking-related crimes cause billions of dollars in economic harm globally (McAfee, 2018), with hundreds of dollars of losses to individuals due to fraud and theft (Norton, 2017).

Qualitative research demonstrates individuals who engage in hacking gain an interest in this activity typically during early adolescence (Holt, 2007; Jordan & Taylor, 1998; Steinmetz, 2015). Evidence from college samples suggest that both low self-control and differential peer associations increase the likelihood of involvement in hacking (Bossler & Burruss, 2010; Holt et al., 2010; Skinner & Fream, 1997). Limited research with juvenile samples found support for an association between low self-control, deviant peer associations with individuals engaged in cybercrime, and hacking behaviors (Back et al., 2018; Holt et al., 2012; Marcum et al., 2014; Udris, 2016).

There are, however, numerous questions that must be addressed related to juvenile participation in hacking to better address gaps in the existing empirical literature (Back et al., 2018; Holt et al., 2012; Marcum et al., 2014; NCA, 2017; Udris, 2016). First, it is unclear what relationships may exist between deviant peer associations related to forms of off-line delinquency and individual involvement in hacking activities. Second, there is a gender gap observed in hacking activity (Holt, 2007; Taylor, 1999), but few have considered whether there are different predictors associated with participation in hacking for males and females using quantitative data. Third, there is limited knowledge as to the extent to which rural and urban residency differentially shape the risk of hacking behaviors due to limited juvenile sample sizes (Holt & Bossler, 2016; Holt et al., 2012; Leukfeldt, 2017; Marcum et al., 2014). These findings may not be generalizable to broader samples of adolescent populations, especially those outside of the United States (Back et al., 2018; Holt & Bossler, 2016; Udris, 2016).

To address these questions, this exploratory study examined the relationships between low self-control, delinquent peer associations, and self-reported hacking in a sample of 48,327 juveniles from 30 countries. Binary logistic regression analyses were conducted, demonstrating substantial support for aspects of both the general theory of crime and differential associations generally. Additional regression models were conducted parsing the models by gender and city size to identify differences in the significant predictors for hacking. The implications of this study for our understanding of juvenile delinquency online and off-line as well as policies to reduce involvement in offending are explored in detail.

Defining the Act of Hacking and the Demographic Composition of Hackers

Over the last few decades, social scientists have examined the act of hacking and its historical evolution from benign act to criminal activity (e.g., Meyer, 1989; Steinmetz, 2015; Taylor, 1999). In fact, the term hacker developed from an extremely benign source: the Massachusetts Institute of Technology (MIT) Model Railroad Club (Marcum, 2015). During this period in the 1950s and 1960s, hacking was associated with individuals who expressed a fervent desire to understand the ins and outs of technology to control and manipulate it for the better (Holt & Schell, 2013). Referring to someone as a hacker during this period was akin to awarding a badge of honor for displaying exceptional knowledge and intelligence in terms of computers and technology.

During the early 1980s, the development of personal home computers and early Internet connectivity provided young people with unparalleled access to complex technology. Concern grew over juveniles with technical skills being able to gain remote access to financial information and sensitive computer systems, driven in part by the popularity of the 1984 film WarGames (Schell & Dodge, 2002; Taylor, 1999). A number of high-profile hacks of banks, businesses, and government networks occurred during this period, some of which were attributed to groups of teenage hackers using names such as the Masters of Deception and the Legion of Doom (Furnell, 2002; Slatalla & Quittner, 1995). As a consequence, the positive connotations of hacking within the computer using community became conflated with the criminal definition accepted by the general public (Furnell, 2002; Kinkade, Bachmann, & Smith-Bachmann, 2016; Taylor, 1999).

Today, the act of hacking is largely perceived in a negative context by the general public and refers primarily to an individual who manipulates or utilizes technology to engage in criminal activity (Holt, 2007; Taylor, 1999). Hacking, like many forms of cybercrime, range from complex tasks, such as the creation and distribution of malicious software, to simple breaches of security performed through password guessing or other means (Holt, 2007; Rogers, Smoak, & Liu, 2006). Evidence suggests individuals typically engage in minor, simplistic forms of hacking as they gain an understating of computers and their behaviors may escalate in both severity and complexity over time (Holt, 2007; NCA, 2017). The onset of hacking appears to occur in early adolescence, similar to other forms of crime, with some desistence into adulthood (Holt, 2007; Jordan & Taylor, 1998; Steinmetz, 2015; Taylor, 1999). As a result, there is substantive value in examining hacking during adolescence to better understand the etiology of this activity compared with traditional forms of offending (Holt & Bossler, 2016; NCA, 2017; Udris, 2016).

Quantitative research to date has focused primarily on hacking perpetration among college student populations, possibly capturing the peak of participation in hacking generally (Bossler & Burruss, 2010; Holt et al., 2010; Holt & Kilger, 2008; Morris, 2010; Rogers et al., 2006; Skinner & Fream, 1997). A smaller body of research have developed juvenile samples, though the rates of hacking are variable depending on the populations sampled (Back et al., 2018; Holt et al., 2012; Marcum et al., 2014; Udris, 2016). Scholars have suggested that the high rate of hacking within adolescence and young adulthood may be related to the general turbulence often seen in this period of development. Indeed, young adolescence and early adulthood is a time where there is usually psychological confusion and a general ethical deficit, which may manifest in the form of cyberdeviance (Muncie, 1999; Rogers et al., 2006).

In that respect, multiple studies demonstrate the overall explanatory power of traditional criminological theories to account for cybercrimes, including computer hacking (see Holt & Bossler, 2016; Leukfeldt, 2017 for review). Although acts like hacking depend upon technology, many of the social and behavioral factors associated with traditional acts of crime and delinquency are also evident in predicting cybercrimes. For instance, there is a significant relationship observed between low self-control and participation in simple forms of hacking, such as password guessing and manipulation of data (Bossler & Burruss, 2010; Holt & Kilger, 2008; Holt et al., 2012; Marcum et al., 2014). There is also evidence that deviant peer associations influence the likelihood of hacking generally. Qualitative research demonstrates hacking is a skill largely developed through experiential learning on one’s own (Holt, 2007; Steinmetz, 2015). Peer associations also play a critical role in learning methods and justifications to hack (Bossler & Burruss, 2010; Holt, 2009; Holt et al., 2010; Morris, 2010; Skinner & Fream, 1997). Most quantitative studies consider peer involvement in various forms of cybercrime to capture the potential for imitation and modeling of support for these offenses. There is also limited evidence that actors involved in hacking may also engage in traditional forms of off-line criminality, including drug and alcohol abuse (Schell & Dodge, 2002; Taylor, 1999) and more serious property crimes and drug sales (Leukfeldt, 2017; Slatalla & Quittner, 1995).

Both the general theory of crime and peer relationships appear to predict involvement in hacking among youth samples as well (Back et al., 2018; Holt et al., 2012; Marcum et al., 2014; Udris, 2016). Individuals with low self-control are likely to associate with deviant peers whom also have low self-control (Baron, 2003; Chapple, 2005; Higgins & Makin, 2004; Hinduja & Ingram, 2009; Holt et al., 2012), in keeping with the argument with differential association that “birds of a feather flock together” (Sutherland, 1939). These studies, however, only measure peers’ participation in online offenses, calling to question how peers engaging in traditional forms of off-line offending affect individual involvement in hacking.

Recent research demonstrates that there is a tie between off-line delinquency and online victimization risks, suggesting behavior in one environment impact activities in another (Holt, Turner, & Exum, 2014; McCuddy & Esbensen, 2017). Qualitative studies of hacking emphasize that individuals interested in hacking during adolescence have few peers who hack in their larger social networks (Holt, 2009). From a social learning perspective, associations with peers involved in traditional forms of delinquency targeting persons or property may offset an absence of associations with peers who hack. Delinquent peers can still expose an individual to general definitions supportive of offending which may increase their willingness to engage in hacks and other forms deviance. The more time they spend with these delinquent peers, regardless of where they offend, may increase their willingness to engage in hacks generally. Given that hackers appear to engage in both virtual and real offenses (e.g., Leukfeldt, 2017; Schell & Dodge, 2002), it is plausible that there may be a relationship between peer engagement in various forms of person and property crimes and individual online deviance. The limited body of scholarship, however, demands further research to examine the relationship between off-line deviance and online deviance within this broader theoretical framework (Hinduja & Ingram, 2009; Holt et al., 2012).

In addition, prior research has identified a persistent relationship between gender and involvement in hacking (Gilboa, 1996; Hutchings & Chua, 2016; Jordan & Taylor, 1998; Schell & Dodge, 2002; Taylor, 1999). Males are overwhelmingly more likely to self-report involvement in hacking, which may be a function of differential gender role socialization of boys toward technology compared with girls during early adolescence (Hutchings & Chua, 2016; Taylor, 1999). It is unclear, however, whether there are differences in the risk factors that account for participation in hacking between males and females during adolescence similar to the patterns observed in off-line delinquency (Daigle, Cullen, & Wright, 2007; Lanctôt & Guay, 2014; Tracy, Kempf-Leonard, Abramoske-James, 2009). For instance, female offending during adolescence is less common compared with males, particularly with respect to violence and property offenses. Similarly, females are more likely to be supervised and monitored by parents or guardians which may decrease their potential for deviant peer interactions and limit opportunities to offend (Daigle et al., 2007; Lanctôt & Guay, 2014). Thus, greater research is needed examining whether these relationships persist with respect to juvenile hacking behaviors.

Limited research also suggests individuals who engage in hacking tend to come from middle class backgrounds and more sizable cities due to differential access to technology and high-speed Internet connectivity (Holt, 2007; Schell & Dodge, 2002; Steinmetz, 2015). These concepts have been largely ignored in quantitative research, either due to difficulty measuring socioeconomic status or limited geographic distribution of respondent populations (Holt & Bossler, 2016; Leukfeldt, 2017; Marcum et al., 2014). It is possible that individuals in higher socioeconomic status groups have increased access to technology thereby increasing potential opportunities to hack. At the same time, as targets in online spaces are available at nearly all times, those opportunities should be constant to any actor and limited only by the actor’s knowledge of its presence (Maimon, Kamerdze, Cukier, & Sobesto, 2013; Yar, 2005).

The Current Study

Although the etiological understanding of hacking has vastly improved, there are several issues requiring empirical exploration. This study attempted to test multiple hypotheses related to computer hacking and examine mediating relationships that have been underexamined in the literature generally. First, it is expected that individuals with low self-control and unsupervised access to technology, which creates opportunities to offend, will be more likely to report engaging in hacking consistent with prior research (e.g., Back et al., 2018; Holt et al., 2012; Marcum et al., 2014; Udris, 2016).

Second, prior research has identified a consistent relationship between deviant peer associations and hacking, such that peer involvement in hacking increases the likelihood an individual will hack (e.g., Bossler & Burruss, 2010; Holt et al., 2012; Skinner & Fream, 1997). As juveniles who hack may have limited physical contacts to other hackers (e.g., Holt, 2007, 2009), associating with peers who engage in any form of deviance and crime may provide justifications for offending that cut across the digital divide. Thus, this analysis tested the hypothesis that time spent with peers engaged in delinquent behavior may be more likely to hack. Multiple forms of off-line delinquency and deviant behavior were included to investigate the association between offense types and hacking.

Third, it is expected males would be more likely to report hacking behaviors in keeping with prior research (Bossler & Burruss, 2010; Holt, 2007; Holt et al., 2012; Jordan & Taylor, 1998; Steinmetz, 2015). There is, however, little research regarding the moderating relationships between gender and specific risk factors for hacking. Thus, this study tested the hypothesis that there would be differential predictors for males compared with females, with more technology access factors associated with males’ hacking behaviors (e.g., Hutchings & Chua, 2016; Taylor, 1999).

Fourth, limited inquiry and anecdotal evidence suggests that juvenile hackers are typically from middle class families (Holt, 2007; Schell & Dodge, 2002; Slattalla & Quittner, 1995; Steinmetz, 2015). This study tested the hypothesis that individuals living in families with higher socioeconomic status would have increased odds of self-reported hacking. Fifth, evidence from qualitative studies suggests that hackers primarily reside in urban areas (e.g., Holt, 2007; Jordan & Taylor, 1998; Steinmetz, 2015). Thus, this study tested the hypothesis that urban residency would be more associated with self-reported hacking. In addition, the moderating relationship between urban and rural residency was examined to consider any differences in the predictors for hacking by place. The implications of this analysis for our understanding of criminological theory, as well as effective prevention and intervention efforts to combat juvenile hacking were discussed in depth.

Data and Methods

This analysis used data developed from the Second International Self-Report of Delinquency study (ISRD-2, Junger-Tas, 2010; Junger-Tas & Marshall, 2012). This data set consists of a sample of juveniles in Grades 7 through 9 found in 30 nations, including the United States, Latin America, and representation of most all European nations. The students were selected using probability sampling in classrooms across small and large cities in each nation (see Marshall & Enzmann, 2012 for more detail on the sampling and methodology). Such a data set is essential to examine the prevalence and correlates of computer hacking as the majority of existing research utilized small samples reflecting single schools or regional populations (except Back et al., 2018; Udris, 2016). In addition, little is known about the global prevalence of hacking behaviors in the general population (Holt & Bossler, 2016; Taylor, 1999). A more diverse and international sample, as with the ISRD-2, provides a robust population to examine the behavioral and demographic correlates of hacking generally.

The full data set contained 68,507 respondents, though our final sample consisted of 48,327 due to incomplete or unclear answer as well as missing responses. Approximately 24% of the total population was excluded from this analysis, though the final sample was demographically consistent to the full data set based on gender (49.3% female and 48.5% male), age (M = 1.08 in both samples), and geographic distribution across all national samples and city sizes (22.4% and 27.0% of the sample respondents came from cities with less than 100,000 residents).

Dependent Variable

Respondents were asked whether they ever used a computer for “hacking?” and to specify “did you do it during the last 12 months?” (hacking: 0 = “no”; 1 = “yes”; see Table 1 for detail). A relatively small proportion of the respondents engaged in hacking (5.4%), which is in keeping with prior rates reported in both juvenile (Holt et al., 2012; Marcum et al., 2014) and college samples (Bossler & Burruss, 2010; Holt et al., 2010; Rogers et al., 2006; Skinner & Fream, 1997). Given that this measure does not define what acts are hacking as in college-based surveys (Bossler & Burruss, 2010), it is not possible to assess the technical skill needed to perform the hack or the severity of the activity. At the same time, this measure allows the respondent to self-identify what they consider to be a hack. Such a measure is consistent with the broad range of activities defined as hacks in qualitative research, including minor system modifications to serious criminal activity such as malicious software distribution (Holt, 2007; Jordan & Taylor, 1998; Steinmetz, 2015; Taylor, 1999).

Table 1.

Descriptive Statistics (n = 48,327).

	M	SD	Minimum	Maximum
Hacking	0.054	.001	0	1
Own computer	0.856	.008	0	1
Own mobile	0.898	.005	0	1
Own room	0.757	.005	0	1
Tech use	4.185	.014	1	6
Piracy	0.492	.007	0	1
Self-control	61.107	.204	0	100
Time peers	4.208	.022	1	6
Peer drugs	1.287	.008	1	4
Peer vandalism	1.193	.004	1	4
Peer shoplift	1.107	.007	1	4
Peer computer	2.444	.009	1	4
Peer frighten	1.348	.006	1	4
Age	1.087	.004	0	3
Car ownership	0.874	.0062	0	1
Small town	0.270	.0174	0	1
Male	0.485	.004	0	1

Independent Variables

A series of two binary measures were included to assess opportunities to engage in hacking based on access to technology: (a) “Do you have a computer at home that you are allowed to use?” (own computer) and (b) “Do you own a mobile phone?” (own mobile). An additional question was provided to measure opportunity to potentially hack: do you have a room of your own? (own room: 0 = “no”; 1 = “yes”). It is thought that access to a private space, like a bedroom, may decrease guardianship and increase the risk of cybercrime offending (Holt & Bossler, 2016; Holt et al., 2012). Access to a computer is also intrinsically essential to hack (Holt, 2007; Jordan & Taylor, 1998; Steinmetz, 2015), though the increased computing power of mobile devices may also be associated with hacking activities.

Two measures were also included to capture technology use and online activities. First, respondents were asked a single question capturing multiple forms of technology use: “Outside school how much time do you spend on an average school day on each of these activities: watching tv, playing games, or chatting on the computer?” A six-item response was provided (tech use: 1 = “none”; 2 = “30 minutes”; 3 = “one hour”; 4 = “two hours”; 5 = “three hours”; and 6 = “four hours plus”). A second binary measure was included to identify involvement in digital piracy: “when you use a computer did you ever download music or films during the last 12 months?” (piracy: 0 = “no”; 1 = “yes”). Prior research has noted a relationship between piracy and hacking behaviors in both the qualitative (Holt, 2007; Holt & Copes, 2010) and quantitative literature (Bossler & Burruss, 2010; Holt et al., 2012). Thus, this measure is included here to identify whether that relationship is consistent in an international sample of juveniles.

Self-control was measured using an abbreviated version of the original 24-item scale developed by Grasmick, Tittle, Bursik, and Arneklev (1993). The ISRD-2 contains 12 of the original 24 items, which capture four of the six dimensions of self-control (impulsivity, risk-taking, volatile temperament, and self-centeredness) and has been validated in cross-national research (Marshall & Enzmann, 2012). To create this measure, the Percentage of Maximum Possible (POMP) scoring method was used to rescale the 12 item measures from 0 to 100 and create an average score for each respondent (Botchkovar, Marshall, Rocque, & Posick, 2015; Cohen, Cohen, Aiken, & West, 1999). Lower scores are indicative of lower levels of self-control. The reliability of this measure is .83, which has been shown in prior research to have substantive validity and be positively associated with delinquency (Botchkovar et al., 2015; Marshall & Enzmann, 2012; Rocque, Posick, & Zimmerman, 2013). Thus, this scale is reliable and comparable with other measures used in self-control research.

A series of six measures were used to assess the relationship between differential peer associations and individual involvement in hacking. First, a single item was used to capture overall time spent with peers, asking: “Outside school how much time do you spend on an average school day . . . hanging out with friends” (time peers: 1 = “none”; 2 = “30 minutes”; 3 = “one hour”; 4 = “two hours”; 5 = “three hours”; 6 = “four hours plus”). This measure assesses both the frequency with which individuals spend time with potential peers as per Akers (1998), and potential opportunities to offend as a function of proximity to other motivated offenders (Haynie & Osgood, 2005; Hoeben, Meldrum, Walker, & Young, 2016; Osgood, Wilson, O’Malley, Bachman, & Johnston, 1996).

Five measures were also used to assess the relationship between peer deviance, online activities, and individual offending. Respondents were asked,

when you hang out with your friends, we usually: 1) drink a lot of beer/alcohol or take drugs (peer drugs); 2) smash or vandalize things just for fun (peer vandalism); 3) shoplift just for fun (peer shoplift); 4) play computer games or chat on the computer (peer computer), and 5) frighten or annoy people around us just for fun (peer frighten).

A four-item response was used (1 = “never”; 2 = “sometimes”; 3 = “often”; 4 = “always”).

Four demographic control variables were included in this analysis. Age was measured as a categorical variable (0 = “less than 12”; 1 = “12 to 15”; 2 = “16 to 17”; 3 = “18 and older”). A binary measure for family car ownership (0 = “no”; 1 = “yes”) was included as a partial proxy for socioeconomic status and access to technology and resources. A binary measure was included to capture the size of the community in which the respondent lived. Cities and towns larger than 100,000 residents or considered to be an important city within the country were coded as 0, and those smaller than 100,000 or not an important city were coded as 1 (small city). This was necessary due to the variations in countries sampled, as some nations may have smaller populations, but still contain major cities from an international perspective. Finally, gender was included (male: 0 = “female”; 1 = “male”) as a control due to the gendered nature of hacking identified in the larger empirical literature (Bachmann, 2010; Gilboa, 1996; Hutchings & Chua, 2016).

Results

To examine the relationships between hacking behaviors and the various factors identified, a binary logistic regression model was estimated. Due to the large number of respondents with variations in the sample size by country, the analyses were conducted using STATA statistical software using the cluster command by school (n = 1,183) to minimize the intra-cluster correlations and standard errors. There was no evidence of multicollinearity as no variance inflation factor (VIF) was higher than 1.62 and no tolerance was lower than 0.618.

Model 1 provides the results of individual-level opportunity and self-control variables on hacking behaviors (see Table 2 for detail). There was partial support for opportunity, as those who owned a computer, mobile phone, and spent more time actively using computers or watching television were more likely to engage in hacking. Having one’s own room was nonsignificant. Juveniles who engaged in piracy were also more likely to have hacked in the last year in keeping with previous research showing a positive correlation between these offenses (Holt et al., 2012).

Table 2.

Binary Logistic Regression Model Results (n = 48,327; Adjusted for 1,183 Clusters by School).

	Model 1			Model 2
	b	SE	Exp(b)	b	SE	Exp(b)
Own computer	0.276**	.121	1.318	0.272**	.123	1.313
Own mobile	0.172*	.102	1.188	0.131	.098	1.140
Own room	−0.002	.057	0.997	−0.018	.057	0.982
Tech use	0.130***	.021	1.139	0.076***	.021	1.078
Piracy	1.900***	.473	6.690	1.808***	.436	6.099
Self-control	−0.019***	.000	0.981	−0.012***	.001	0.987
Time peers				0.036*	.015	1.036
Peer drugs				0.208***	.049	1.231
Peer vandalism				0.091*	.044	1.095
Peer shoplift				0.187***	.054	1.206
Peer computer				0.176**	.032	1.192
Peer frighten				0.078**	.033	1.081
Age	0.036	.041	1.037	0.003	.046	1.003
Car ownership	0.311***	.111	1.364	0.286***	.110	1.332
Small town	0.346***	.103	1.414	0.353***	.104	1.423
Male	1.076***	.159	2.933	1.024***	.153	2.784
Constant	−5.221***	.001	0.005	−6.492***	.000	0.001
Pseudo R²	.150			.166

Note. Model 1: chi-square = 1,816.67*** and −2LL of −9,356.687; Model 2: chi-square = 2,086.20*** and −2LL of −8,513.636.

p < .05. **p < .01. ***p < .001.

Low self-control was also significantly associated with involvement in hacking, which aligns with prior research (Bossler & Burruss, 2010; Holt et al., 2012; Marcum et al., 2014). Males were also more likely to hack (Gilboa, 1996; Hutchings & Chua, 2016) as were those whose families owned a car. In addition, those living in smaller cities were more likely to hack, though it is not clear why this relationship may be present.

Model 2 includes both the individual and peer measures to test their association with hacking. Only one difference was observed between the models: owning a mobile phone was no longer significant in Model 2. Owning a computer and use of technology remained significant predictors and at the same levels of significance. An individual’s level of self-control is still significant, despite the inclusion of multiple peer association measures, similar to prior research (Holt et al., 2012). With respect to deviant peers, those who spent more time with friends were more likely to engage in hacking. All five measures of specific peer activities were also significant, providing support for the notion that association with deviant peers increases individual deviance online (Akers, 1998; Holt & Bossler, 2016; Marcum et al., 2014).

Given the significance of gender in both Models 1 and 2, the model was partitioned by sex (see Table 3). The findings demonstrated that individuals with low self-control and deviant peer associations were significant predictors for both males and females. Specifically, those whose peers used drugs, shoplifted, and played computer games were more likely to engage in hacking. In addition, individual involvement in piracy, parents owning their own vehicle, and living in small towns were consistent predictors across both sexes. There were also distinct factors affecting the risk of offending, as females were more likely to hack if they spent more time with their peers generally, and also had peers who frightened others (see Table 3, Model 1). Males were more likely to hack if they owned their own computer and mobile device, spent more time watching TV and playing computer games, and had peers who engaged in vandalism (see Table 3, Model 1).

Table 3.

Binary Logistic Regression Model Segmented By Gender (n = 48,327; Adjusted for 1,183 Clusters by School).

	Females (n = 24,862; 1,154)			Males (n = 23,465; 1,164)
	b	SE	Exp(b)	b	SE	Exp(b)
Own computer	0.030	.161	1.030	0.385***	.106	1.470
Own mobile	−0.073	.171	0.928	0.189*	.099	1.208
Own room	0.010	.105	1.010	−0.029	.068	0.971
Tech use	−0.003	.036	0.996	0.101***	.023	1.106
Piracy	1.651***	.121	5.216	1.862***	.085	6.436
Self-control	−0.015***	.002	0.984	−0.011***	.001	0.988
Time peers	0.078**	.029	1.082	0.024	.016	1.024
Peer drugs	0.230***	.063	1.259	0.198***	.036	1.219
Peer vandalism	0.043	.086	1.044	0.117*	.046	1.124
Peer shoplift	0.327***	.075	1.387	0.134**	.049	1.144
Peer computer	0.157**	.051	1.170	0.179***	.031	1.196
Peer frighten	0.174**	.056	1.191	0.047	.037	1.048
Age	−0.017	.087	0.982	0.005	.052	1.005
Car ownership	0.519***	.162	1.680	0.210*	.095	1.234
Small town	0.265*	.127	1.304	0.381***	.074	1.464
Constant	−6.061***	.361	0.002	−5.641***	.235	0.003
Pseudo R²	.126			.132

Note. Model 1: chi-square = 704.34*** and −2LL of −2,566.534; Model 2: chi-square = 1,079.67*** and −2LL of −5,926.238; note that any statistically significant difference in coefficients is highlighted.

p < .05. **p < .01. ***p < .001.

There were two variables that were statistically significantly different between males and females in these models on the basis of Z-tests comparing regression coefficients (Clogg, Petkova, & Haritou, 1995). Specifically, more time spent watching TV and playing computer games increased the likelihood of offending among males, whereas females whose peers engaged in shoplifting were more likely to hack. These relationships suggest there may be some differential pathways to offending, as noted in prior research (Gilboa, 1996; Hutchings & Chua, 2016; Taylor, 1999).

To further explore the relationship observed between city size and hacking, the model was again partitioned by large and small cities (see Table 4 for detail). Low self-control and having peer relationships with individuals who use drugs and play computer games were significant across both models. In addition, involvement in piracy and being male were significant predictors across both cities. Several differences were also observed across the models, as individuals living in large cities who had access to mobile phones, time spent watching TV or playing games, having peers who shoplift and frighten others, and have parents who own their own cars were more likely to hack (see Table 4, Model 1). Individuals living in small towns were more likely to hack if they had access to a computer and spent more time with their peers generally (see Table 4, Model 2).

Table 4.

Binary Logistic Regression Model Segmented By City Size (n = 48,327; Adjusted for 1,183 Clusters by School).

	Large City (n = 35,238; 915)			Small Town (n = 13,089; 269)
	b	SE	Exp(b)	b	SE	Exp(b)
Own computer	0.229	.119	1.258	0.415**	.143	1.515
Own mobile	0.290**	.104	1.336	−0.130	.152	0.877
Own room	−0.031	.068	.969	0.014	.110	1.014
Tech use	0.092***	.023	1.096	0.042	.036	1.043
Piracy	1.834***	.087	6.260	1.775***	.124	5.904
Self-control	−0.012***	.001	.987	−0.013***	.001	0.986
Time peers	0.012	.017	1.012	0.090***	.027	1.095
Peer drugs	0.229***	.037	1.257	0.163**	.062	1.177
Peer vandalism	0.079	.048	1.083	0.118	.077	1.125
Peer shoplift	0.201***	.053	1.223	0.157	.082	1.170
Peer computer	0.181**	.033	1.199	0.162***	.044	1.176
Peer frighten	0.107**	.036	1.113	0.006	.057	1.006
Age	−0.017	.053	.982	0.093	.098	1.098
Car ownership	0.320***	.093	1.377	0.141	.181	1.151
Male	0.994***	.064	2.702	1.098***	.104	3.000
Constant	−6.679***	.252	.001	−5.872***	.369	0.002
Pseudo R²	.162			.174

Note. Model 1: chi-square = 1,624.47*** and −2LL of −5,956.528; Model 2: chi-square = 546.63*** and −2LL of −2,547.279; note that any statistically significant difference in coefficients is highlighted.

p < .05. **p < .01. ***p < .001.

There were only two significant differences observed on the basis of Z-test results, specifically mobile phone access, which was significant for those in large cities only. This may be a factor of either socioeconomic access, or possibly a function of poor mobile phone reception in smaller cities and rural areas. In addition, the fact that time spent with peers was a significant predictor only for small towns, suggests there may be something unique about the nature of youth activities by place type.

Discussion and Conclusion

This study adds to the hacking literature and provides new insight into risk factors associated with juvenile hackers on an international level. Currently, there is sparse information on hacking perpetrated by juveniles compared with college-aged individuals (Holt & Bossler, 2016; Leukfeldt, 2017; NCA, 2017) and very little that considers hacking in a cross-national context. Aside from this contribution, this study examined risk factors of hacking engagement from a gender-neutral framework (see Table 2) as well as from a gendered framework (see Table 3). This pursuit yielded interesting differences that suggest new lines of inquiry for criminological research. Overall, our study contributed to an understudied area within criminology and should serve as a catalyst for more research.

Specifically, this study found a generally small proportion of respondents reported engaging in hacking in a large international sample (see also Back et al., 2018; Udris, 2016). These levels were lower than that of smaller samples developed in the United States only (e.g., Holt et al., 2012; Marcum et al., 2014). In addition, these findings supported the proposed hypotheses on the relationship between the general theory of crime (Gottfredson & Hirschi, 1990), aspects of social learning theory (Akers, 1998), and juvenile engagement in hacking. Youth with low self-control and deviant peer networks had significantly greater odds of engaging in hacking as found in similar studies of hacking with juvenile and young adults samples (Bossler & Burruss, 2011; Holt et al., 2012; Holt & Kilger, 2008; Marcum et al., 2014; Udris, 2016). Moreover, youth with access to and use of their own computer were more likely to engage in hacking, which aligns with similar research (Holt et al., 2012) as well as the general theory of crime broadly (Gottfredson & Hirschi, 1990). Youth who actively used technology would have more opportunities for online deviant behavior compared with their peers with less access. The results show that youth who engaged in digital piracy were significantly more likely to also hack, consistent with prior research examining patterns of engagement in cyberdeviance broadly (Holt et al., 2012; Marcum et al., 2014).

This study also found support for the hypothesis that associations with deviant peers engaged in delinquency off-line increased one’s own risk of imitating antisocial behavior. Youth who associated with peers who committed property crimes, used drugs, and played computer games more likely to engage in hacking. These findings reinforce the basic postulates of social learning theory, specifically that associations with delinquent peers may increase support for and willingness to engage in deviance on or off-line (see also Holt et al., 2010). The preliminary nature of this result demands future research to consider the relationship between hacking and peer associations generally. It may be that youth interested in hacking have broader deviant peer associations beyond other hackers, as prior studies demonstrate hackers also reported substance use and performed various real-world offenses (Leukfeldt, 2017; Schell & Dodge, 2002). Thus, future qualitative and quantitative assessments must disentangle the role of peers in both off-line and online offending behaviors (McCuddy & Esbensen, 2017).

This study also found mixed support for hypotheses related to the relationships between largely unexamined demographic variables and hacking. For instance, youth from families with higher socioeconomic status, measured by vehicle ownership, were more likely to report engaging in hacking behaviors across models. It is possible that this relationship masked other factors, though car ownership appears to be an appropriate proxy indicator of socioeconomic status. If so, this finding aligns with prior research examining technology usage across income levels, which found that middle to upper socioeconomic status corresponded with less reliance on the schools for Internet access (Daniel, 2005). From an opportunity perspective, the greater availability and accessibility of the Internet and technology to middle and upper socioeconomic groups lead to added opportunities for cybercrime, which provides context to the relationship identified in this study. Future research is needed with improved measures of socioeconomic status to increase our understanding of the impact of technology access to account for juvenile hacking.

By contrast, the hypothesis that youth living in urban areas would be more likely to report hacking behaviors was not supported. The significance of living in a small town was somewhat contrary to the limited prior research that has found that hackers are unlikely to live in rural areas given the sparse technological services available (Holt, 2007; Schell & Dodge, 2002). The relationship observed in this analysis may be function of the notion that living in a rural area with fewer activities and less structured time might increase opportunities for delinquency (Hoeben et al., 2016; Osgood & Anderson, 2004). Although this relationship needs further exploration, it reinforces initial evidence by Marcum and colleagues (2014) who found that over 12% of a sample of youths in rural North Carolina performed simple hacks such as password guessing.

Finally, this study found gender moderated involvement in computer hacking, supporting past research that hacking is a male-dominated activity (Gilboa, 1996; Hutchings & Chua, 2016). The findings also supported the hypothesis that there were different behavioral correlates of hacking, similar to research on traditional delinquency generally (Daigle et al., 2007; Tracy et al., 2009). Owning a computer and use of technology remained key correlates for hacking, though only for boys (see Taylor, 1999). Moreover, associating with peers who engaged in shoplifting carried significantly greater odds for girls to hack compared with boys. In contrast, associating with peers who used drugs carried the highest odds for boys to hack among all other deviant peer activities. Finally, time spent with peers significantly increased the odds of hacking for girls in keeping with some prior research on female delinquency generally (Daigle et al., 2007).

These results suggest there may be gendered pathways for hacking that have not previously been identified in qualitative or quantitative research. Specifically, the significance of opportunity for technological access appears less important for girls given the significance associated with time spent with peers and deviant peer behavior. Further analysis is needed to better explicate the role of gender in hacking, particularly with qualitative data to assess the ways that boys and girls who hack describe their experiences with and motives for hacking (see also Gilboa, 1996; Hutchings & Chua, 2016). Quantitative analyses are also needed to more clearly identify the background structural and individual factors correlated with the onset and persistence of hacking across gender (NCA, 2017).

This study also has direct benefit for policy-makers to combat juvenile hacking, as there are few empirical intervention strategies to date focused on economic cybercrimes (NCA, 2017). The results of this analysis demonstrate early hacking behaviors may be associated with traditional predictors of delinquency, meaning that specialized interventions tailored to cybercrime may not be essential. Instead, integrating information related to simple forms of computer hacking in traditional delinquency prevention programs may prove effective to help curb illegal activities on and off-line. The potential gender differences noted in this analysis also suggest a need to highlight the negative consequences of delinquent peer associations for on and off-line behaviors among females to help minimize the risk of offending. However, substantive empirical research is needed to develop and evaluate any cybercrime prevention programs generally (Holt & Bossler, 2016; Leukfeldt, 2017; NCA, 2017).

Although this study utilized a large juvenile sample to assess computer hacking, several limitations constrain its generalizability. First, there have been continued technological changes that increased Internet access and reduced the cost of computers since the time these data were collected in the mid-2000s. Additional research is needed with more contemporary sample populations to assess any changes in the relationships observed within this data, particularly the third wave of the ISRD. The cross-sectional nature of these data also limits our ability to identify causal relationships between variables. Future studies are needed utilizing longitudinal data sets to better explicate the pathways and trends in hacking behavior among youth (Holt et al., 2012; Marcum et al., 2014; NCA, 2017). In addition, the largely Western population of the ISRD-2 calls to question whether there are differential predictors for hacking in Asian and Oceanic nations generally. Research is needed assessing these issues with similar nationally representative samples.

In addition, this secondary data set did not allow for more nuanced measures as to what constitutes hacking or differences in the technical sophistication of the hacks performed. There were also no measures included in the ISRD-2 for peer hacking behaviors, limiting the potential to assess differential associations with others engaged in similar online offending. Such information would greatly improve our understanding of the dynamics between deviant peer associations, self-control, and an individual’s technical capabilities to hack (Bossler & Burruss, 2011; Holt & Bossler, 2016). In turn, we may increase our understanding of any distinctions that exist between delinquency in online and off-line spaces.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Thomas J. Holt

Author Biographies

Thomas J. Holt is a professor in the School of Criminal Justice at Michigan State University. His research focuses on cybercrime, cyberterrorism, and policing these problems using qualitative and quantitative data sets. He received his PhD from the University of Missouri-St. Louis in 2005.

Jordana N. Navarro is an assistant professor in the Department of Criminal Justice at the Citadel. Her research interests are cybercrime, intimate partner abuse, victimology, and sexual violence. She received her PhD in Sociology from the University of Central Florida in 2013.

Shelly Clevenger is an associate professor in the Department of Criminal Justice Sciences at Illionis State University. Her research focuses on sexual assault, intimate parnter abuse, and cyber-victimization. She received her PhD from Indiana University of Pennsylvania.

References

Akers

(1998). Social learning and social structure: A general theory of crime and deviance. London, England: Routledge.

Bachmann

(2010). The risk propensity and rationality of computer hackers. The International Journal of Cyber Criminology, 4, 643-656.

Back

Soor

LaPrade

(2018). Juvenile hackers: An empirical test of self-control theory and social bonding theory. International Journal of Cybersecurity Intelligence & Cybercrime, 1(1), 40-55.

Baron

S. W.

(2003). Self-control, social consequences, and criminal behavior: Street youth and the general theory of crime. Journal of Research in Crime and Delinquency, 40, 403-425.

Bossler

A. M.

Burruss

G. W.

(2010). The general theory of crime and computer hacking: Low self-control hackers? In Holt

T. J.

Schell

B. H.

(Eds.), Corporate hacking and technology-driven crime: Social dynamics and implications (pp. 38-67). Hershey, PA: IGI Global.

Botchkovar

Marshall

I. H.

Rocque

Posick

(2015). The importance of parenting in the development of self-control in boys and girls: Results from a multinational study of youth. Journal of Criminal Justice, 43, 133-141.

Chapple

C. L.

(2005). Self-control, peer relations, and delinquency. Justice Quarterly, 22(1), 89-106.

Clogg

C. C.

Petkova

Haritou

(1995). Statistical methods for comparing regression coefficients between models. American Journal of Sociology, 100, 1261-1293.

Cohen

Aiken

L. S.

West

S. G.

(1999). The problem of units and the circumstance for POMP. Multivariate Behavioral Research, 34, 315-346.

10.

Daigle

L. E.

Cullen

F. T.

Wright

J. P.

(2007). Gender differences in the predictors of juvenile delinquency: Assessing the generality-specificity debate. Youth Violence and Juvenile Justice, 5, 254-286.

11.

Daniel

A. J.

(2005). An exploration of middle and high school students’ perceptions of deviant behavior when using computers and the Internet. Journal of Technology Studies, 31(2), 70-80.

12.

Furnell

(2002). Cybercrime: Vandalizing the information society. Boston, MA: Addison-Wesley.

13.

Gilboa

(1996). Elites, lamers, narcs, and whores: Exploring the computer underground. In Cherny

Weise

(Eds.), Wired women: Gender and new realities in cyberspace (pp. 98-113). Seattle, WA: Seal Press.

14.

Gottfredson

M. R.

Hirschi

(1990). A general theory of crime. Palo Alto, CA: Stanford University Press.

15.

Grasmick

H. G.

Tittle

C. R.

Bursik

R. J.

Jr. Arneklev

B. J.

(1993). Testing the core empirical implications of Gottfredson and Hirschi’s general theory of crime. Journal of Research in Crime and Delinquency, 30(1), 5-29.

16.

Haynie

D. L.

Osgood

D. W.

(2005). Reconsidering peers and delinquency: How do peers matter? Social Forces, 84, 1109-1130.

17.

Higgins

G. E.

Makin

D. A.

(2004). Does social learning theory condition the effects of low self-control on college students’ software piracy. Journal of Economic Crime Management, 2(2), 1-22.

18.

Hinduja

Ingram

J. R.

(2009). Social learning theory and music piracy: The differential role of online and offline peer influences. Criminal Justice Studies, 22, 405-420.

19.

Hoeben

E. M.

Meldrum

R. C.

Walker

Young

J. T.

(2016). The role of peer delinquency and unstructured socializing in explaining delinquency and substance use: A state-of-the-art review. Journal of Criminal Justice, 47, 108-122.

20.

Holt

T. J.

(2007). Subcultural evolution? Examining the influence of on-and off-line experiences on deviant subcultures. Deviant Behavior, 28, 171-198.

21.

Holt

T. J.

(2009). Lone hacks or group cracks: Examining the social organization of computer hackers. In Schmalleger

Pittaro

(Eds.), Crimes of the Internet (pp. 336-355). Upper Saddle River, NJ: Pearson Prentice Hall.

22.

Holt

T. J.

Bossler

(2016). Cybercrime in progress: Theory and prevention of technology-enabled offenses. New York, NY: Routledge.

23.

Holt

T. J.

Bossler

A. M.

May

D. C

. (2012). Low self-control, deviant peer associations, and juvenile cyberdeviance. American Journal of Criminal Justice, 37, 378-395.

24.

Holt

T. J.

Burruss

G. W.

Bossler

A. M.

(2010). Social learning and cyber-deviance: Examining the importance of a full social learning model in the virtual world. Journal of Crime and Justice, 33(2), 31-61.

25.

Holt

T. J.

Copes

(2010). Transferring subcultural knowledge on-line: Practices and beliefs of persistent digital pirates. Deviant Behavior, 31, 625-654.

26.

Holt

T. J.

Kilger

(2008, April 21-22). Techcrafters and makecrafters: A comparison of two populations of hackers. In WOMBAT Workshop on Information Security Threats Data Collection and Sharing 2008 WISTDCS’08 (pp. 67-78). New York, NY: IEEE.

27.

Holt

T. J.

Schell

B. H.

(2013). Hackers and hacking: A reference handbook. Santa Barbara, CA: ABC-CLIO.

28.

Holt

T. J.

Turner

M. G.

Exum

M. L.

(2014). The impact of self control and neighborhood disorder on bullying victimization. Journal of Criminal Justice, 42, 347-355.

29.

Hutchings

Chua

Y. T.

(2016). Gendering cybercrime. In Holt

T. J.

(Ed.), Cybercrime through an interdisciplinary lens (pp. 181-202). Oxon, UK: Routledge.

30.

Jordan

Taylor

(1998). A sociology of hackers. Sociological Review, 46, 757-780.

31.

Junger-Tas

(2010). The significance of the international self-report delinquency study (ISRD). European Journal on Criminal Policy and Research, 16(2), 71-87.

32.

Junger-Tas

Marshall

I. H.

(2012). Introduction to the International Self-Report Study of Delinquency (ISRD-2). In Junger-Tas

Marshall

I. H.

Enzmann

Killias

Steketee

Gruszczynska

(Eds.), The many faces of youth crime (pp. 3-20). New York, NY: Springer.

33.

Kinkade

P. T.

Bachmann

Smith-Bachmann

(2016). Hacker Woodstock: Observations on an offline cyber culture at the Chaos Communication Camp 2011. In Holt

T. J.

(Ed.), Crime online: Correlates, causes, and context (pp. 29-55). Durham, NC: Carolina Academic Press.

34.

Lanctôt

Guay

(2014). The aftermath of workplace violence among healthcare workers: A systematic literature review of the consequences. Aggression and Violent Behavior, 19, 492-501.

35.

Leukfeldt

E. R.

(Ed.). (2017). The human factor in cybercrime and cybersecurity: Research agenda. The Hague, The Netherlands: Eleven International Publishing.

36.

C. K.

Holt

T. J.

Bossler

A. M.

May

D. C

. (2016). Examining the mediating effects of social learning on the low self-control—Cyberbullying relationship in a youth sample. Deviant Behavior, 37, 126-138.

37.

Maimon

Kamerdze

Cukier

Sobesto

(2013). Daily trends and origin of computer-focused crimes against a large university computer network: An application of the routine-activities and lifestyle perspective. British Journal of Criminology, 53, 319-343.

38.

Marcum

C. D.

(2015). Cyber crime. New York: Wolters Kluwer Law & Business.

39.

Marcum

C. D.

Higgins

G. E.

Nicholson

(2017). I’m watching you: Cyberstalking behaviors of university students in romantic relationships. American Journal of Criminal Justice, 42, 373-388.

40.

Marcum

C. D.

Higgins

G. E.

Ricketts

M. L.

Wolfe

S. E.

(2014). Hacking in high school: Cybercrime perpetration by juveniles. Deviant Behavior, 35, 581-591.

41.

Marshall

I. H.

Enzmann

(2012). Methodology and design of the ISRD-2 study. In Junger-Tas

Marshall

I. H.

Enzmann

Killias

Steketee

Gruszczynska

(Eds.), The many faces of youth crime (pp. 21-65). New York, NY: Springer.

42.

McAfee. (2018). Economic impact of cybercrime- No slowing down. Retrieved from https://www.mcafee.com/enterprise/en-us/forms/gated-form.html?docID=5fee1c652573999d75e4388122bf72f5&tag=ec&eid=18TL_ECGLQ1_CT_WW&elqCampaignId=23163

43.

McCuddy

Esbensen

F. A.

(2017). After the bell and into the night: The link between delinquency and traditional, cyber-, and dual-bullying victimization. Journal of Research in Crime and Delinquency, 54, 409-441.

44.

Meyer

G. R.

(1989). The social organization of the computer underground. DeKalb: Northern Illinois University.

45.

Morris

R. G.

(2010). Computer hacking and techniques of neutralization. In Holt

T. J.

Schell

B. H.

(Eds.), Corporate hacking and technology-driven crime: Social dynamics and implications (pp. 1-17). Hershey, PA: IGI Global.

46.

Muncie

(1999). Youth and crime: A critical introduction. London, England: SAGE.

47.

National Crime Agency. (2017, March 14). New assessment warns industry that cyber criminals are imitating nation state attacks. Retrieved from http://www.nationalcrimeagency.gov.uk/news/1043-new-assessment-warns-industry-that-cyber-criminals-are-imitating-nation-state-attacks

48.

Norton. (2017). Cyber security insights report. Retrieved from https://www.symantec.com/content/dam/symantec/docs/about/2017-ncsir-global-results-en.pdf

49.

Osgood

D. W.

Anderson

A. L.

(2004). Unstructured socializing and rates of delinquency. Criminology, 42, 519-550.

50.

Osgood

D. W.

Wilson

J. K.

O’Malley

P. M.

Bachman

J. G.

Johnston

L. D.

(1996). Routine activities and individual deviant behavior. American Sociological Review, 61, 635-655.

51.

Pratt

T. C.

Cullen

F. T.

(2000). The empirical status of Gottfredson and Hirschi’s general theory of crime: A meta-analysis. Criminology, 38, 931-964.

52.

Pratt

T. C.

Cullen

F. T.

Sellers

C. S.

Thomas Winfree

Jr. Madensen

T. D.

Daigle

L. E.

Gau

J. M.

(2010). The empirical status of social learning theory: A meta-analysis. Justice Quarterly, 27, 765-802.

53.

Rocque

Posick

Zimmerman

G. M.

(2013). Measuring up: Assessing the measurement properties of two self-control scales. Deviant Behavior, 34, 534-556.

54.

Rogers

Smoak

N. D.

Liu

(2006). Self-reported deviant computer behavior: A big-5, moral choice, and manipulative exploitive behavior analysis. Deviant Behavior, 27, 245-268.

55.

Schell

B. H.

Dodge

J. L.

(2002). The hacking of America: Who’s doing it, why and how. Westport, CT: Quorum Books.

56.

Skinner

W. F.

Fream

A. M.

(1997). A social learning theory analysis of computer crime among college students. Journal of Research in Crime and Delinquency, 34, 495-518.

57.

Slatalla

Quittner

(1995). Masters of deception. New York: Harper Collins.

58.

Steinmetz

K. F.

(2015). Craft(y)ness: An enthnographic study of hacking. British Journal of Criminology, 55, 125-145.

59.

Sutherland

E. H.

(1939). Principles of criminology. Philadelphia, PA: J.B. Lippincott Company.

60.

Taylor

(1999). Hackers: Crime in the digital sublime. London, England: Routledge.

61.

Tracy

P. E.

Kempf-Leonard

Abramoske-James

(2009). Gender differences in delinquency and juvenile justice processing: Evidence from national data. Crime & Delinquency, 55, 171-215.

62.

Udris

(2016). Cyber deviance among adolescents and the role of family, school, and neighborhood: A cross-national study. International Journal of Cyber Criminology, 10, 127-146.

63.

Vazsonyi

A. T.

Mikuška

Kelley

E. L.

(2017). It’s time: A meta-analysis on the self-control-deviance link. Journal of Criminal Justice, 48, 48-63.

64.

Wall

D. S.

(2007). Cybercrime: The transformation of crime in the information age. Cambridge: Polity.

65.

Yar

(2005). The novelty of ‘cybercrime’: An assessment in light of routine activity theory. European Journal of Criminology, 2, 407-427.