Female Copycat Crime: An Exploratory Analysis

Abstract

Although much has been written concerning female delinquency and criminality, little has been forwarded about female copycat crime. Utilizing four data sets from surveys collected over the past decade, the characteristics of females who reported involvement in copycat crimes, the crimes they attempted, and how they compared with male copycat offenders are described. Two research questions are considered: “How much does being female predict copycat crime?” and “How do female copycat crime offenders differ from noncopycat females?” Results indicate that females are less likely to have a copycat crime history but there was also evidence of the copycat crime gap between females and males narrowing. In general, copycat criminality was more common in younger and incarcerated females, although playing or not playing video games did not differentiate copycat crime females from noncopycat crime females. The quantity of media consumed also did not predict copycat crime in either gender. A typical copycat female offender committed their crime, often involving a property crime, while a teenager in a spontaneous group act. The relationship between females and criminogenic media is likely different compared with that of males, and the debunking of crime media content for younger females needs to be pursued.

Introduction

It is accepted that females commit less crime than males (Chesney-Lind Pasko 2013; Smith 2014). However, a narrowing of the gap between female and male criminality has also been long noted (Adler et al. 1975). Debate over the driver of this narrowing revolves around three competing explanations: females becoming more criminal, males becoming less criminal, and net-widening policies causing more female arrests (Estrada et al. 2016; Steffensmeier et al. 2005, 2006). Excluding homicide [see, for example, Fox and Fridel (2017)] research on the gender crime gap has not extensively looked at specific types of offenses. A unique crime type where the gender gap has not been examined is copycat crime.

What makes copycat crime unique is its genesis in the yoking of two crimes through a media crime model (Surette 2013). As generally conceptualized, the components of copycat crimes are “generator crimes” (a precursor crime to subsequent copycat crimes); “criminogenic models”; “copycat criminals”; and “copycat crimes.” Although the generator and copycat crimes may be geographically and temporally distanced, they share a unique criminogenic relationship with the first crime serving as the source of the later crime. In sum, the copycat crime concept implies the operation of a powerful dynamic regarding the criminogenic influence of media on offenders (Surette 2016). Research on copycat crime has established reasonable grounds that such crimes occur at a significant rate and are the genesis of crime careers for a minority but substantial portion of offenders (Helfgott 2015; Surette 2017; Surette and Maze 2015). An estimate of the prevalence of adult male copycat crime by Surette (2014) was that one in four offenders have attempted a copycat crime in their lifetimes.

The recent ascent of social media has heightened copycat crime concerns (Kien et al. 2019) and copycat crime lies at the crux of a criminogenic media influence (O'Toole et al. 2014). The current long-standing debate over the impact of violent video games on player aggressive behaviors highlights the importance of exploring media and copycat crime.¹ As knowledge about criminogenic media effects remains sparse, a study of females and copycat crime is valuable for understanding the nature of the crime gender gap and the dynamics of female crime.

Materials and Methods

Data for this study were obtained from four surveys collected between 2008 and 2015. The surveys included both male and female respondents and queried each respondent on a comparable set of questions regarding copycat crime. All four surveys were voluntary, anonymous, utilized informed consent, and were approved through University Institutional Review Boards. The Florida surveys were administered to male and female residents at group homes and inmates at a large county jail facility. Approval for surveying the group homes was obtained through court-appointed guardians. Respondents completed informed consent forms before survey administration, which stated that participation was voluntary, anonymous, and could be withdrawn at any point. In the two jail surveys, high-risk inmates were not included but general populations inmates and those in drug and alcohol treatment, education, and religious programs as well as trustees, inmate institutional workers, and work release inmates were sampled. Access was arranged through jail administrators and correctional officers were present at all survey administrations. The three Florida survey instruments and protocols were approved through the University of Central Florida Institutional Review Board. The Trinidad survey was part of a country-wide crime victimization survey that utilized a national multistage cluster, proportionate sample of private households. Data were collected through face-to-face administration of a questionnaire by interviewers recruited from the Trinidadian Central Statistical Office and the student population of The University of the West Indies. Selected respondents were provided a brief description of the survey, assured that their responses were confidential and anonymous, and informed of their right to terminate the interview at any time. Protocol approval was obtained from the Ethics Committee of the Faculty of Medical Sciences, The University of the West Indies.

Providing baseline data on copycat crime prevalence, all respondents were asked if they had ever considered committing a copycat crime and if they had ever attempted one. Wording for “considered a copycat crime” was: “The media includes television, radio, movies, videos, video games, music, books, magazines, newspapers and the Internet. Can you remember ever having learned something in the media that was against the law and thought about trying to do something similar?” Wording for “attempted a copycat crime” was: “Can you remember having tried to do something against the law that you learned about in the media?” Asking individuals about their copycat crime histories suffers the usual weaknesses of self-report measures. However, self-report has been found to be a valid method, particularly in terms of measuring serious and rare forms of criminality (Farrington 1988; Farrington et al. 1996; Junger-Tas and Marshall 1999; Kivivuori 2011; Klein 1988; Krohn et al. 2010; Steinberg and Monahan 2007; Thornberry and Krohn 2000). The reliability of self-reported measures is improved when a criminal behavior is measured through “ever” questions and for more rarely committed acts such as copycat crime (Bruinsma 1989; Junger-Tas and Marshall 1999). Thus, the copycat crimes recalled by respondents would be more serious copycat events, reducing the inclusion of trivial imitations and increasing the probability of differentiating true copycat from noncopycat offenders (Junger-Tas and Marshall 1999). The nature of copycat crime as a relatively uncommon but unique activity made it a good candidate for self-report methods (Table 1).

Table 1.

Copycat Surveys with Female Respondents

Survey	No. of Females	Female age range	Sample method	Research design	Source
Central Florida 2008, n = 134	71 Nonincarcerated high-risk youth	9–19	Nonrandom	Written questionnaire	Opportunity sample of youth in community service center programs and area group homes
Orange County Florida 2010, n = 487	134 County jail inmates	19–72	Nonrandom	Written questionnaire	Opportunity sample of general population and work release inmates
Orange County Florida 2012, n = 220	60 County jail inmates	18–50	Nonrandom	Written questionnaire	Opportunity sample of general jail population and work release inmates
Trinidad and Tobago 2015, n = 601	310 Adult general nonincarcerated residents	18–96	Random	Face-to-face interviews	A household respondent selected using a Last Birthday method followed by administration of a structured questionnaire

The surveys

The oldest data set was a 2008 nonrandom survey of 140 nondelinquent central Florida youth in foster care and group homes. Six respondents did not report their gender leaving 133 subjects composed of 53.4% females for analysis. The second survey was a 2010 nonrandom, voluntary, anonymous survey of Orange County Florida corrections facilities and included both incarcerated general population inmates and work release sentenced offenders; high-risk and disciplinary lock-down inmates were not included. This survey queried 574 respondents for a working sample of 487 of which 27.5% were females. Despite being nonrandom, the sample was found to demographically compare well with the 2010 total jail population (Surette 2013). The third data set mirrored the second survey data as a nonrandom, voluntary survey of the incarcerated adult corrections population in Orange County Florida. This effort surveyed 249 respondents for a working sample of 220 respondents of which 27.3% were female. The last survey was a random sample of nonincarcerated general population adults in the nation of Trinidad and Tobago. Gender was collected on 601 respondents of whom 51.6% were female and the sample was deemed a valid representation of the Trinidadian population (Surette and Chadee 2019).

Collectively the four efforts surveyed 1564 individuals, of whom 123 (7.8%) did not identify their gender. The final working data set thus included 575 females (40%) and 866 males. Over the three US-based surveys, there were 287 whites, 349 blacks, and 127 Hispanics; in the Trinidadian survey, there were 204 African Trinidadians, 235 Indonesian Trinidadians, and 168 mixed race respondents. Although the four surveys varied regarding goals and methodologies, all respondents were similarly asked about their having considered or attempted a copycat crime in their lifetimes and when a copycat crime was reported; details about their copycat crimes were collected. In addition, basic demographic and media consumption data were collected. Demographically, females were not different from males in terms of their average age (36.6 males, 36.9 females), nor in their average daily consumption of television (females 4.2 h, males 4.0 h per day). They did differ significantly in terms of their Internet and video game hours per day, females reported spending 3.6 h per day surfing the Internet, males spent 3.0 h per day (t-test = 2.14, sig. 0.033), female respondents averaged 1.65 h per day playing video games, males 2.97 h (t-test = 4.36, sig. 0.000). Females were also more likely to be white, especially among the incarcerated. As expected and reflecting the divergent populations the respondents were drawn from, the Trinidad general population respondents were older than the two adult incarcerated samples. Collectively, the four data sets provided a wide range of subject types allowing an examination of self-reported copycat crime among females and across female/male comparisons.

The analysis strategy was to first examine copycat crime prevalence by gender and next conduct a set of related group comparisons and describe the details of female copycat crime. A set of bivariate correlations were culled for variables for logistic regression models to address two research questions: “Does being female predict copycat crime consideration or attempt?” and “What factors distinguish female copycats from female noncopycat offenders?” The second research question was examined using logistic models run only on female respondents. Logistic regression analysis procedures used were as recommended by Mertler and Vannatta (2005) and Norusis (2008).²

Results

What does these data reveal about female copycat offenders? First, consistent with Surette (2014), Table 2 reflects that the prevalence of copycat crime was significantly less among females. The gender crime gap extends to copycat crime, both as considered and attempted acts. Specifically, 11.5% (67 of 583) females reported that they had considered a copycat crime in their lives, less than the 21.6% of males (192 of 887). Similarly, 6.2% (36 of 583) females reported that they had attempted a copycat crime in their lives, also significantly less than the 17.9% of males (159 of 887) who reported a copycat crime attempt. Although these proportion differences were statistically significant, the effect-size measure (Hedges g) difference between females and males indicates that although not trivial, the effect of gender on copycat crime consideration (0.319) and attempt (0.349) falls between small and medium in impact.³

Table 2.

Difference Between Considered and Attempted Copycat Crime Percentage by Gender and Survey

	Considered %							Attempted %
	n	F	M	Dif.	g	F	M	Dif.	G
Combined surveys	1441	11.5	21.6	10.1^*	0.319	6.2	17.9	11.7^*	0.349
Central Florida youth 2008	134	11.1	17.7	6.6		6.9	9.7	2.8
Orange County Jail 2010	487	21.0	25.5	4.5		9.4	22.8	13.4^*	0.344
Orange County Jail 2012	220	28.6	34.7	6.1		22.2	28.8	6.6
Trinidad 2015	601	3.9	10.0	6.1^*	0.267	1.3	7.2	5.9^*	0.298

Proportional t-test difference between female and male proportions significant at 0.05, Hedges g effect size.

When prevalence was examined by survey, Table 2 shows within survey variance with a marked difference between jailed versus nonincarcerated respondents. Regarding copycat crime consideration, females were significantly different from males, a difference driven by the 2015 Trinidadian general adult survey respondents. The effect size of this difference was small however. For both jail inmate surveys and at-risk youth, contemplation of a copycat crime was consistently lower in female than among male respondents in the other three surveys, although the differences were not statistically significant. Concerning copycat crime attempts, a similar consistent pattern of lower proportions of females having attempted a copycat crime across all four surveys was found and was statistically significantly different in the 2010 jail inmate group and in the 2015 Trinidad adult population survey, the effect-size measures for both fell between small and medium. Suggesting a gender gap narrowing, jailed females were significantly less likely than males to report a copycat attempt in 2010; however, they were not significantly different in 2012. The two jail surveys reflect that as far as copycat crime attempts were concerned, females closed the gap with males from 2010 to 2012, females increasing 13% (9.4–22.2) compared with an increase of 6% for males (22.8–28.8).

Older compared with younger females

Dichotomizing the respondents into groups of adults (those older than 25 years) and young adults and juveniles (those 25 years or younger), the younger female respondents were mildly more likely to have considered copycat crimes (18.2% vs. 9.0%, effect size g = 0.289) and attempted (12.1% vs. 3.9%, effect size g = 0.342) than older females. Comparative values for the young male respondents were 27.4% considered and 21.8% attempted compared with 19% older males considered and 16.2% attempted a copycat crime. Older and younger males were significantly different regarding copycat crime consideration (sig. = 0.005) but the effect (g = 0.205) was small.

Incarcerated compared with nonincarcerated females

Incarcerated female respondents were substantially more likely to have considered copycat crimes (23.4% vs. 5.2%, effect g = 0.592) and attempted (13.4% vs. 2.4%, g = 0.469) than nonincarcerated females. Comparative values for the incarcerated male respondents were 28.5% considered and 24.7% attempted compared with 11.3% of the nonincarcerated males who had considered and 7.6% who had attempted a copycat crime. The effect size of incarceration for males was moderate and similar to females for both considered copycat crime (male effect g = 0.467) and attempted a copycat crime (male effect g = 0.456). Similar to younger compared with adult females, female inmates were as likely to have considered copycat crime and less likely to have attempted one (g = 0.456) than comparable males.

Video gamers compared with nongamer females

Three of the surveys (the Trinidad survey did not) asked respondents whether they played video games and how many hours they played during a typical day. In these surveys, females who played video games were not significantly more likely to have considered a copycat crime (19.3% vs. 16.9%) than those who did not. Concerning attempted copycat crime, female gamers also did not significantly differ from female nongamers (12.3% vs. 7.8%).⁴ Comparative values for the male respondents were 26.3% of male gamers who had considered copycat crime compared with 23.2% of nongamers; for attempted copycat crime, 22.7% of male gamers had attempted a copycat crime and 20.2% of nongamer males had attempted one. Among males, neither difference was significant.

The lack of video game play significance for females regarding copycat crime consideration and attempt extended to the number of hours played, whereas male game play hours differed significantly between those who had considered a copycat crime and males who had not. Females who had considered a copycat crime averaged 1.22 h of game play daily, whereas females who had not considered copycat crime averaged 1.74 daily game hours. Concerning copycat crime attempts and daily video game hours, females who had attempted a copycat crime averaged 3.06 h per day, those who had not averaged 1.46 h. While approaching a statistically significant relationship between game hours and copycat crime attempts, this hourly difference failed to meet a standard 0.05 confidence level (t = 2.015, sig. = 0.053). Overall, female gamers and nongamers were as likely to have considered and attempted copycat crimes and generally as likely as comparative game playing and nongame playing males. The exception was for nongame players where 7.6% of female nongamers compared with 20.1% of male nongamers reported having attempted a copycat crime (t = 3.496, sig. = 0.001, g = 0.353). Relevant for the ongoing debate over video games, game play and copycat crime were not significantly different across gender and within gender female gamers and female nongamers were as statistically likely to have considered (18.2% vs. 16.7%) or attempted (16.1% vs. 7.6%) copycat crime.⁵

The characteristics of female copycat crime

Thirty-six females and 159 males described attempted copycat crimes. Summarized in Table 3, the females tended to be younger and fell within a narrower age range than male copycats. Nineteen of the 36 females were younger than 17 years when they committed a copycat crime. Most of the female copycat crimes were spur of the moment decisions committed in groups. Two-thirds of copycat females and males committed spontaneous crimes and were subsequently punished. Indicating that they felt that the media played a noncrucial role in their crimes, females more than males felt that they would have committed their crime without exposure to the media generator crime suggesting a crime molding rudder effect rather than a crime causing trigger effect for media. The crimes female copycats described were similar to males in terms of being violent, property, or drug related. The sole crime divergence was that male copycats were more likely to have committed a sex crime. The copycat crimes described by females included serious offenses such as aggravated battery, carjacking, and bank robbery but mostly consisted of property crimes like theft and shoplifting, and drug offenses. Overall supporting a narrowing of the gender gap, except for females being more likely to have committed their copycat crime in a group and less likely to report sex crimes, there were no substantial differences between female and male copycat crime characteristics.

Table 3.

Copycat Crimes—Female and Male Copycat Comparisons

	Females (n = 36)	Males (n = 159)	g
Mean age	16.8	18.1
Age range	8–30	6–52
Preplanned	38.2%	35%
Caught/punished	66.7%	66.5%
Group^*	60%	32.1%	0.589
Commit anyway	65.8%	53.3%
Violent	29.6%	28.0%
Property	59.3%	56.3%
Drugs	11.1%	9.4%

χ² = 9.61, significant at 0.002.

In sum, female copycat offenders were less prevalent than male copycat criminals. Females were half as likely to have considered committing a copycat crime and one-third as likely to have attempted one. Female copycats also tended to be younger (none were older than 30 years) than male copycats and their crimes were more often committed in groups. Similar to males, female copycats were spontaneous and unsuccessful about two-thirds of the time. The quantity of media consumption and video game play were not important for either gender. The copycat crime gender prevalence gap was closing, however, for incarcerated females where the copycat crime attempt rate increased at twice the rate of male inmates. Based upon the bivariate correlations reported in Table 4 between considered and attempted copycat crime, a set of predictor variables for a set of logistic regressions associated with the study's two research questions were conducted.

Table 4.

Selected Bivariate Pearson Correlations

	Copycat attempt	Female	Age	White	African American	Trinidad	Jailed	TV hours	Internet hours
Considered	0.472^**	−0.130^**	−0.164^**	0.150^**	0.041	−0.216^**	0.228^**	0.009	0.011
Attempt		−0.169^**	−0.144^**	0.071^**	0.077^**	−0.206^**	0.224^**	0.019	0.048
Female			0.009	0.007	−0.158^**	0.203^**	−0.252^**	0.020	0.069^*
Age				−0.090^**	−0.260^**	−0.440^**	−0.180^**	−0.161^**	−0.224^**
White					−0.286^**	−0.414^**	0.414^**	0.008	−0.160^**
United States African						−0.472^**	0.320^**	0.317^**	0.032
Trinidad							−0.833^**	−0.333^**	0.140^**
Jailed								0.211^**	−0.169^**
TV hours									0.145^**

Pairwise n's range from 1454 to 1564.

Significant at 0.05.

Significant at 0.01.

Research question 1: How predictive is being female for copycat crime?

The logistic model for considered copycat crime had four predictor variables—female, age, white, and incarcerated; for the attempted copycat crime model, African American (1 = yes, 0 = no) was added.⁶ The logistic regression reported in Table 5, shows that being female significantly reduced the likelihood of copycat crime consideration. Being in jail made it three times as likely to have considered a copycat crime, being white increased consideration probability, and being older decreased the probability. Age was impactful although its odds/ratio was close to 1 because of the effect of increasing one year in age having a slight effect but over the entire age range was highly significant. In sum, being female and older significantly reduced the likelihood of copycat crime consideration, being incarcerated and white increased it. However, although the model was significant overall, the moderate R² suggest that the model was missing important additional predictor variables.

Table 5.

Research Question 1 Logistic Regression Considered Copycat Crime, Block Entry Female, Age, White, Jailed (N = 1420)

Predictor	Logit	SE	Wald	P	Exp (b)
Female	−0.488	0.166	8.66	0.003	0.614
Age	−0.033	0.006	31.70	0.000	0.967
White	0.457	0.198	7.22	0.007	1.579
Jailed	1.110	0.190	39.88	0.000	3.036

Model χ² = 140.7; −2-log likelihood = 1177.85; sig. = 0.000; Nagelkerke R² = 0.156.

Considered copycat crime = 249, recommended for analysis = 40.

SE, standard error.

A logistic regression block model was also run on attempted a copycat crime on all respondents employing the predictors of female, age, white, African American, and jailed. As given in Table 6, as with copycat crime consideration, being female continued to significantly reduce the likelihood of having attempted a copycat crime. In addition, being older substantially reduced the likelihood of having attempted a copycat crime, whereas being incarcerated increased copycat crime attempt probability four and a half times.

Table 6.

Research Question 1: Logistic Regression Attempted Copycat Crime Block Entry Female, Age, White, African American, Jailed (N = 1420)

Predictor	Logit	SE	Wald	p	Exp (b)
Female	−0.884	0.207	18.27	0.000	0.413
Age	−0.033	0.007	21.71	0.000	0.968
White	−0.133	0.224	0.35	0.552	0.875
African American	−0.293	0.213	0.213	0.169	0.746
Jailed	1.546	0.225	47.37	0.000	4.693

Model χ² = 138.06; −2-log likelihood = 956.99; model sig. = 0.000; Nagelkerke R² = 0.172.

Attempted copycat crime = 184, minimum needed for analysis = 50.

Research question 2: What factors distinguish female copycats from female noncopycat offenders?

Table 7 provides the logistic regression results for females who had considered or not considered a copycat crime employing the predictors of age, incarcerated, and white.⁷ As reported, incarcerated and younger females were significantly more likely to have considered a copycat crime.

Table 7.

Research Question 2: Logistic Regression Considered Copycat Crime Block Entry Age, White, Jailed (N = 567)

Predictor	Logit	SE	Wald	p	Exp (b)
Age	−0.071	0.015	23.71	0.000	0.931
White	0.558	0.330	2.86	0.091	1.747
Jailed	1.671	0.348	22.98	0.000	5.315

Model χ² = 77.5; −2-log likelihood = 334.46; sig. = 0.000; Nagelkerke R² = 0.247.

Considered copycat crime = 67; recommended for analysis = 30.

Limited to 27 females who had attempted a copycat crime, the logistic regression results in Table 8 was restricted to three predictor variables of age, incarcerated, and daily number of Internet hours as suggested from the bivariate correlations in Table 4. The results for this limited model show that being incarcerated was the strongest predictor of female copycat crime history (raising its likelihood sixfold), followed by age. The number of hours spent surfing the Internet dropped from significance.

Table 8.

Research Question 2: Logistic Regression: Attempted Copycat Crime Block Entry Age, Jailed, Internet Hours (N = 396)

Predictor	Logit	SE	Wald	P	Exp (b)
Age	−0.074	0.026	8.48	0.004	0.928
Jailed	1.851	0.477	15.07	0.000	6.363
Internet hours	0.064	0.037	2.99	0.084	1.066

Model χ² = 25.4; −2-log likelihood = 176.9; sig. = 0.000; Nagelkerke R² = 0.155.

Attempted copycat crime = 27, recommended for analysis = 30.

Overall, the logistic regression's chi-square and 2-log likelihood values indicated that the selected variable models were significant and statistically reliable for distinguishing between copycat crime and noncopycat crime respondents. The Nagelkerke's R² values indicate that the models explained from 15% to 25% of group membership and correctly classified from 80% to 90% of the cases. Irrespective, they did not discriminate well between copycats and noncopycats in the multivariate analyses, relying on bulk placement of cases into the noncopycat crime groups. Regression coefficients statistics did indicate that being female and being incarcerated consistently predicted self-reported copycat crime, whereas being younger and incarcerated differentiated females who had considered copycat crime and could be speculated to play important roles in copycat crime attempts. However, much of female copycat crime remained unexplained and likely owing to factors not included.

Discussion

This study found that by 2012 incarcerated females had significantly closed the gap with male inmates regarding copycat crime prevalence, mirroring within the realm of copycat crime the larger societal narrowing crime gap (Adler et al. 1975; Estrada et al. 2016; Fox and Fridel 2017; Smith 2014; Steffensmeier et al. 2005, 2006). Females were consistently less likely than males to commit copycat crime but by a declining degree. Comparison between female versus male copycat crime trends with other specific crime types and longitudinal studies are needed. It is likely that female copycat crime dynamics will emerge differently (herein, for example, it was much more likely to have been a group act) and female interaction with criminogenic media to substantially differ. The results also indicate that simple exposure to media does not drive copycat crime. Playing video games and the number of hours spent watching television or surfing the Internet did not emerge as important factors for copycat crime consideration or attempts. Speculation is that what copycat offenders bring to their media consumption is more important than the quantity of the consumption or the content of the media. The media as a crime catalyst or rudder is suggested. Thus, copycat crime could result from offenders extracting crime instructions and by creating composite copycat crime models that combine real-world and media models. Copycat offenders may consume very little media but likely consume what media they do in a manner different from noncopycats.

Limitations

This study was limited to a small number of female copycat offenders drawn from unique populations and its generalizability is unknown. In addition, a restricted set of predictor variables were available and subsequently the logistic models displayed weak to moderate explanatory power and were likely underspecified. The self-report data would also have been bolstered by direct behavioral or official co-measures to validate the respondents copycat crime reports. Finally, the most recent survey used was collected in 2015 and the nature of the media and consumer media interactions have evolved substantially with the continued emergence of social media.

Conclusion

Future gender related research questions include: Do females see the media as credible, useful, and accurate regarding crime? How important is the demographic matching of generator crime models with potential female copycats? What role do preexisting criminal proclivities and criminal efficacy play for females considering a copycat crime? Does the quality of exposure differ by gender and across media types? What role does social media play in female copycat crime generation? Each question needs to be explored by gender and specifically within female copycats.

The primary policy implication from this study is that the debunking of criminogenic media content for females, particularly for younger and incarcerated ones, needs to be pursued. The refinement of counseling interventions should be adapted as additional research knowledge of female copycat crime emerges (Dill et al. 2011). Not indicated in this study, if media exposure is a direct cause of female copycat crime, policies aimed at controlling access and content would be sensible (Surette 2013). If the media are catalysts for female crime, a focus on at-risk females to reduce the likelihood of offending rather than focusing on exposure to generator crimes makes sense. The development of interactive, widely available social media has made such efforts more pressing.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Funding Information

No funding was received for this article.

References

Adler

, Adler

, Levins

. (1975). Sisters in Crime. (McGraw-Hill, New York, NY).

Bruinsma

. (1989). Scaling and reliability problems in self-reported property crime. In Cross-National Research in Self-Reported Crime and Delinquency. M Klein, ed. (Springer, Dordrecht, The Netherlands) pp. 131–151.

Chesney-Lind

, Pasko

. (2013). The Female Offender. (Sage, Thousand Oaks, CA).

Dill

, Redding

, Smith

, et al. (2011). Recurrent issues in efforts to prevent homicidal youth violence in schools: Expert opinions. New Dir Youth Dev. 129, 113–128.

Estrada

, Bäckman

, Nilsson

. (2016). The darker side of equality?. The declining gender gap in crime: Historical trends and an enhanced analysis of staggered birth cohorts. Br J Criminol. 56, 1272–1290.

Farrington

. (1988). Self-reported and official offending from adolescence to adulthood. In Cross-National Research in Self-Reported Crime and Delinquency. M Klein, ed. (Kluwer Academic Publishers, Dordrecht, The Netherlands) pp. 399–423.

Farrington

, Loeber

, Stouthamer-Loeber

, et al. (1996). Self-reported delinquency and a combined delinquency seriousness scale based on boys, mothers, and teachers: Concurrent and predictive validity for African Americans and Caucasians. Criminology. 34, 493–517.

Ferguson

, Copenhaver

, Markey

. (2020) Reexamining the findings of the American Psychological Association's 2015 task force on violent media: A meta-analysis. Perspect Psychol Sci. [Epub ahead of print]; DOI: 10.1177/1745691620927666.

Fox

, Fridel

. (2017). Gender differences in patterns and trends in US homicide, 1976–2015. Violence Gender. 4, 37–43.

10.

Gallar

, Ferguson

. (2020). Violent video games do not contribute to societal violence and crime. In Oxford Handbook of Digital Technologies and Mental Health. M Potenza, K Faust, D Faust, eds. (Oxford University Press, Oxford, United Kingdom) pp. 189–201.

11.

Groves

, Prot

, Anderson

. (2020). Violent media use and violent outcomes. In Oxford Handbook of Digital Technologies and Mental Health. M Potenza, K Faust, D Faust, eds. (Oxford University Press, Oxford, United Kingdom) pp. 202–211.

12.

Helfgott

. (2015). Criminal behavior and the copycat effect: Literature review and theoretical framework for empirical investigation. Aggress Violent Behav. 22, 46–64.

13.

Junger-Tas

, Marshall

. (1999). The self-report methodology in crime research. Crime Justice. 25, 291–367.

14.

Kien

, Begay

, Lee

, Stefanidis

. (2019). Social media during the school shooting contagion period. Violence Gender. 6, 201–210.

15.

Kivivuori

. (2011). Discovery of Hidden Crime: Self-Report Delinquency Surveys in Criminal Policy Context. (Oxford University Press, Oxford, United Kingdom).

16.

Klein

. (1988). Cross-National Research in Self-Reported Crime and Delinquency. (Kluwer Academic Publishers, Dordrecht, The Netherlands).

17.

Krohn

, Thornberry

, Gibson

, Baldwin

. (2010). The development and impact of self-report measures of crime and delinquency. J Quant Criminol. 26, 509–525.

18.

Mertler

, Vannatta

. (2005). Advanced and Multivariate Statistical Methods. (Pyrczak Publishing, Glendale, CA).

19.

Norusis

. (2008). SPSS Statistical Procedures Companion. (Prentice Hall, Upper Saddle River, NJ.)

20.

O'Toole

, Beresin

, Berman

, et al. (2014). Celebrities through violence: The copycat effect and the influence of violence in social media on mass killers. Violence Gender. 1, 107–116.

21.

Smith

. (2014). Long-term trends in female and male involvement in crime. In The Oxford Handbook of Gender, Sex, and Crime. R Gartner, B. McCarhty, eds. (Oxford University Press, Oxford, United Kingdom) pp. 139–157.

22.

Steffensmeier

, Schwartz

, Zhong

, Ackerman

. (2005). An assessment of recent trends in girl's violence using diverse longitudinal sources: Is the gender gap closing?. Criminology. 43, 355–406.

23.

Steffensmeier

, Zhong

, Ackerman

, et al. (2006). Gender gap trends for violent crimes, 1980 to 2003: A UCR-NCVS comparison. Femin Criminol. 1, 72–98.

24.

Steinberg

, Monahan

. (2007). Age differences in resistance to peer influence. Dev Psychol. 43, 1531–1543.

25.

Surette

. (2013). Cause or catalyst: The interaction of real world and media crime models. Am J Crim Justice. 38, 392–409.

26.

Surette

. (2014). Estimating the prevalence of copycat crime: A research note. Crim Justice Policy Rev. 25, 703–718.

27.

Surette

. (2016). Measuring copycat crime. Crime Media Cult. 12, 37–64.

28.

Surette

. (2017). Copycat crime. In Oxford Research Encyclopedia of Criminology: Crime, Media, and Popular Culture. (Oxford University Press, Oxford, UK). [Epub ahead of print]; DOI: 10.1093/acrefore/9780190264079.013.33.

29.

Surette

, Chadee

. (2019). Copycat crime among non-incarcerated adults. Curr Issues Crim Justice [Epub ahead of print]; DOI: 10.1080/10345329.2019.1640058.

30.

Surette

, Maze

. (2015). Video game play and copycat crime: An exploratory analysis of an inmate population. Psychol Pop Media Cult. 4, 360–374.

31.

Thornberry

, Krohn

. (2000). The self-report method for measuring delinquency and crime. In Measurement and Analysis of Crime and Justice. D Duffee, ed. (National Institute of Justice, Washington, DC) pp. 33–83.