Research Design and Methods Analysis in Criminal Justice
Research Design, Methodology, and Sampling
The study focused on the analysis of the results of researches on delinquency and other behavioral problems among youth. It utilized the data and other pertinent information from the studies conducted among youth through Denver survey in Colorado, Pittsburgh study in Pennsylvania, and Rochester development study in New York. The data were chosen based on the first two years (1987-1988) of these studies (Huizinga, Loeber, Thornberry, and Cothern, 2000).
The research studies conducted among youth in Denver, Pittsburgh, and Rochester in 1987 to 1988 enlisted 4, 000 subjects with seven to fifteen years of age (Huizinga, Loeber, Thornberry, and Cothern, 2000). These three independent studies have utilized longitudinal research designs to ensure that life span development and behavioral changes among youth will be properly taken into account. Regular-interval interviews were made to these children and one of their parents privately. However, subject selection varied in the conducted studies. In Denver, 806 males and 721 females or 1, 527 subjects were selected from 20, 000 randomly chosen households while the 1, 517 boys in Pittsburgh were randomly chosen from first, fourth, and seventh grades in the public schools (Huizinga, Loeber, Thornberry, and Cothern, 2000). By conducting an interview with each boy, teachers, and parents, the 30% of the subjects with the most disruptive behaviour were selected. Then, the 1,000 seventh and eight grades students where randomly chosen from the public schools (Huizinga, Loeber, Thornberry, and Cothern, 2000). Even though these three studies have own sampling technique and subjects with different socio-demographic characteristics, they do share significant similarities. Denver, Pittsburgh, and Rochester studies are all longitudinal researches that conducted investigation in parallel with the subjects’ stages of psycho-social developments at regular interval pattern. In each research, interviews with the subjects, their parents, and school officials were personally done. Also, the individual police, school, and court records of each subject were analyzed to come up with the most reliable and valid data concerning behavioral problems and delinquency (Huizinga, Loeber, Thornberry, and Cothern, 2000). Additionally, within the investigation periods there was an average of 90% retention among the selected youth in their respective places. Thus, changes on the developmental and behavioral attributes of each individual were documented (Huizinga, Loeber, Thornberry, and Cothern, 2000). Moreover, collaboration among Denver, Pittsburgh, and Rochester researchers was done by using the same research instruments for different variables, hence, comparison and contrast among findings is possible (Huizinga, Loeber, Thornberry, and Cothern, 2000).
Design and Method Analysis
Several studies have dealt with delinquency and other behavioral problems in different settings and perspectives. Thus, those researches were conducted for different purposes and with samples having different socio-demographic attributes. It is therefore essential to reflect on those findings to gain a wider perspective on the issues, thus, meta-analysis was employed in making an extensive assessment on the researches conducted among youth in Denver, Pittsburgh, and Rochester. By doing so, the implementation of a logically sound program on prevention and intervention with delinquency and other disruptive behaviors among youth is hoped to be properly addressed.
Meta-analysis is a method of research which involves intensive examination and discussion of the available literature for a particular issue (Guzzo, Jackson, and Katzell, 1987). Its procedures are purposively designed for the examination and integration of various quantitative researches of similar nature. By synthesis of the available data and information, one may find the gap among findings and the weaknesses on the methodological parts of various studies (Guzzo, Jackson, and Katzell, 1987). Hence, this may provide a good background for the researcher leading for a plausible and valid method for a certain issue that was not addressed on previous studies.
Based on the bulletin report, from the collaborative efforts of Denver, Pittsburgh, and Rochester research teams resulted to an intensive findings on behavioral problems. The three teams, even though dealt with different populations, have employed identical research instruments for data gathering (Huizinga, Loeber, Thornberry, and Cothern, 2000). They have utilized longitudinal technique in the interview, data verification, and development and behavioral changes observations. Thus, any crucial changes on the characteristics of their subjects were given emphasis. Through an intensive examination of the results of these studies Huizinga, Loeber, and Cothern (2000) were able to find a possible ground on their notion concerning the inter-relations among delinquency and other behavioral problems. However, limitations of the three studies hindered for the meta-analysis to create a logical ground for this notion.
The meta-analysis only utilized the data for the first two years of the three studies which has a ten-year duration. Several more years, subjects’ development and behavioral changes are possible. Since, Pittsburgh study has no female subject any generalization that may be drawn on the account of the investigated variables is not valid and reliable. Also, some data denoting the possible inter-relations among delinquency and other behavioral problems are unreliable due to poor sampling. The meta-analysis therefore must be revised to include the newest data on behavioral problems with a better sampling technique. Hence, the over all findings of the meta-analysis were affected due to the inadequate data provided by the three studies.
Meta-analysis is appropriately utilized for the synthesis of similar studies in the natural sciences but has several limitations when applied in social and behavioral sciences. The difficulty in controlling environmental conditions, variation on the purposes of the study, and the employed sampling technique in the various studies, even though dealing with similar issue, may badly affect the result of meta-analysis. Wolf (1986) identified several disadvantages of this method: generalization is not possibly drawn from the results of several studies that, although dealt on the same issue, utilized different instrument, tested different variables, and followed different methodology for these create a great difference on the outcomes; it is illogical to make a plausible interpretation from the integration of the results of several studies for it entails convergence of good and poor research methodologies; results of several studies although may harmoniously conform to one another but create bias for they came from dissimilar studies; and more often non-significant findings in published research are omitted to avoid disgrace on the study. Therefore, the researches that will be included in meta-analysis must be cautiously chosen.
Through meta-analysis, the variables that persistently came out in several studies can be figured out and subjected for further studies (Guzzo, Jackson, and Katzell, 1987). Guzzo, Jackson, and Katzell (1987), suggested several ways to avoid the possible imperfections of meta-analysis: ensure that most if not all, studies done on similar issues be included in the exploration of literature; intensive examination on the nature of methodology of the study; and verification on the validity and reliability of research instruments, data gathering and statistical treatment. These suggestions were similar to Wolf’s (1986) ideas in resolving the issues against meta-analysis. He suggested that only studies with logically designed methodology should be included in the meta-analysis and validation of data should be done through examination of the original copy of the research manuscript.
Based on the above discussion on the features and possible biases of meta-analysis, the bulletin report of Denver, Pittsburgh, and Rochester studies must be modified. Not only present data should be included in the revision, rather, the subjects and individuals who are the sources of all information in the three studies should be taken into account. Primarily, effective sampling technique should be employed in the areas of study to both male and female populations. Then, other socio-demographic factors such as parents' educational attainment, nature of work, and income should be explored. Nonetheless, the availability of program or support on the prevention of delinquency and other disruptive behavior in the community and educational institution should also be examined.
Guzzo, R.A., Jackson, S.E., and Katzell, R.A. (1987). Meta-Analysis Analysis. Research in Organizational Behavior, 9: 407-442.
Huizinga D. , Loeber, R. , Thornberry, T.P. , and Cothern, L. (2000). Co-occurrence of Delinquency and Other Problem Behaviors. Office of Juvenile Justice and Delinquency Prevention, Juvenile Justice Bulletin. Retrieved November 17, 2008, from http://www.ncjrs.gov/pdffiles1/ojjdp/182211.pdf
Wolf, F.M. (1986). Meta-Analysis: Quantitative Methods for Research Synthesis. Michigan: Sage Publication.