Practitioners in all fields and disciplines use research evidence to justify their decisions. Therefore, the quality of research findings plays a huge role in how various practical approaches, models, and decisions are implemented. Criminal justice is no exception to this rule. The growing body of qualitative and quantitative research provides a good basis for improving the quality of criminal justice operations and results. The purpose of the given work is to evaluate a recent research article from the standpoint of its methods, designs, and practical implications. The focus of the given analysis is the study by Thomas J. Holt, Adam M. Bossler and David C. May (2012), titled “Low self-control, deviant peer associations, and juvenile cyberdeviance”, published in American Journal of Criminal Justice.
The research problem Holt et al. (2012) explored in their study is cybercrime. Numerous theories were proposed to explain the factors of and motivations behind cybercrime. Earlier researchers used mostly two different theoretical frameworks to explain cybercrime – social learning theory and the general theory of crime (Holt et al., 2012). These theories were originally designed to explain traditional street crimes, but they have proved to be valuable, as applied to online offenses. The main shortcoming of the earlier studies is that most of them were based on university samples.
Consequently, the generalizability of their results beyond universities is quite poor. Thus, the purpose of the study was to address the abovementioned issues “by examining the effects of low self-control and deviant peer associations on multiple forms of cybercrime offending in a population of middle and high school students” (Holt et al., 2012, p. 379). The researchers expected that their findings would add to the existing literature on social learning, low self-control, and their capacity to explain various forms of cybercrime.
The study was designed as correlational research, since the researchers sought to define and explain correlations among several different variables. According to Jackson (2011), in correlational studies, researchers seek to define whether two variables correlate. To put it simply, researchers try to define how two different variables are related. One of the most interesting aspects of the study is the way the independent and dependent variables are operationalized. Cyberdeviance was the main dependent variable, operationalized as a variety of illegal and unethical online behaviors, including the use of transfer of pirated media, intentional use or transfer or a pirated copy of commercial copyrighted software, viewing illicit sexual content, posting threatening or mean messages, or accessing somebody else’s account without permission (Holt et al., 2012). Independent variables included: low self-control and deviant peer association. Low self-control was measured with the help of Grasmick et al. scale, whereas deviant peer association was intended to clarify whether cybercrime could be treated as a collective phenomenon (Holt et al., 2012). Control variables included: non-school hours, computer skills, computer location, and demographics (Holt et al., 2012).
Both inductive and deductive logic find their application in the study by Holt et al. (2012). Inductive logic is used, when researchers observe an event, explore its facets, and make empirical generalizations or even design a theory (Dantzker & Hunter, 2011). Holt et al. (2012) use inductive logic to generalize their findings and conclude that peer offending is a more important factor of cyberdeviance among youngsters than low self-control. By contrast, deductive logic has its roots in theory. As Dantzker and Hunter (2011) put it, deductive logic begins with a theory, guiding the researcher towards hypothesis development and testing. This is exactly what happens in the first sections of the article, where Holt et al. (2012) describe previous studies, theories of self-control and crime, identify the existing research gaps, and develop a research question to be answered in their study.
The research study is quantitative. First, its purpose was to explain a correlation between an independent and a dependent variable. Second, it was based on statistics, calculations, and descriptive numbers. Third, structured scales and research instruments were used to gather and process the primary data. Fourth, generalizations were made based on the study results (Holt et al., 2012). Fifth, the researchers did not try to reconsider the relevance of the quantitative results through the prism of their own experiences, as it usually occurs in qualitative research. The latter is used to gain a better insight into the nature of a specific phenomenon. Quantitative designs are more predictive in their numerical methods and forms, but they must be developed with caution and care to avoid subjectivity and bias.
The study was based on quantitative design. The sample was made of the middle and high school students residing in a large metropolitan area in Kentucky (Holt et al., 2012). The total number of respondents was 518, with a total of 435 cases actually analyzed (Holt et al., 2012). 25 percent of the sample was represented by high school students, and 35 percent came from the middle school (Holt et al., 2012). 50 percent of the participants were female, and 79 percent were White (Holt et al., 2012). The researchers do not provide any information as to the rates of return. Linear regression models were used to analyze the statistical data. The results confirmed that self-control could predict various forms of cyberdeviance. At the same time, peer offending was found to have stronger effects on online offending than low self-control (Holt et al., 2012).
In addition, the researchers questioned the relationship between cyberdeviance and offline offense: they did not find any clear correlation between the two variables (Holt et al., 2012). The article does not include any section for conclusions or recommendations. However, the study itself could have been done differently. At the beginning of the article, Holt et al. (2012) suggest that earlier samples included mostly university students, thus making their results less generalizable to other populations. Holt et al. (2012) could have designed a different sample, made of non-students, in order to close the existing gap in empirical literature and offer a new insight into the problem of cybercrime.
