Synthesis of Research Methods

The Research Problem Reprised

The research problem submitted for the course on “Research Topic, Problem, Purpose, and Questions” revolves around the effectiveness of incarceration and rehabilitation programs for offending youth. Defining the appropriate research questions and how three hypothetical designs might answer them requires that the variables and their interrelationships be defined precisely and methodically.

For the effect or dependent variable, the alternatives presented by professional experience and a review of literature are rehabilitation and recidivism. Recidivism is preferable for being unequivocal and meeting any standard of empirical rigor.

Conceptually, a recidivist is a repeat offender. There are subtle differences in definitions used across legal jurisdictions. In Montana, for instance, the Department of Corrections Advisory Council accepts the Association of State Correctional Administrators (ASCA) method of calculating recidivism based on an offender being remanded to prison, regardless of cause, within three years after release. The phrase, “regardless of cause” occurs because the U.S. Sentencing Commission accommodates three definitions for recidivism: re-conviction, being put on trial without sentencing or other disposition, and a violation of probation or supervision that led to revocation of the privilege (Maxfield, 2005; Montana Dept. of Corrections, 2008).

That recidivism is a vital benchmark of performance for the judicial and correctional systems is shown by the fact that judges consult sentencing tables recommending longer incarceration proportionate to the repeat-offense rate of the offender. Ultimately, every instance of recidivism is testimony that the avowed function of prison as correctional punishment and rehabilitation has failed (Castillo et al., 2004; Grommet, 2009).

What then are the influential variables that impinge on recidivism? In the view of Ashker and Kenny (2008), incarceration is ineffective if it is not punitive enough or inmates do not receive the kind of skills training that would enable them to obtain employment and thus avoid relapsing into crime for economic gain. De la Torre (2007) reinforces the theme of enhancing inmate chances for gainful employment by calling for remedial education while serving time, at the same time that the author trusts in the character-shaping benefit of involvement in sports and church activities (De la Torre, 2007).

Even if the balance of punitive systems, skills training, educational remediation, and character-building rehabilitation were uniform in all penal institutions (or held constant, for analytical purposes), Barnhart (2009) and other researchers argue that variable success at rehabilitation can be traced back to some youthful inmates coming from higher-risk backgrounds. In the context of an effectiveness study, this constitutes an antecedent factor (Trochim and Donnelly, 2008).

This concise overview of the major factors that minimize or aggravate recidivism suggests that it would be ideal if research design could investigate inmate factors, correctional methods and correctional officers themselves.

Quantitative Design

For the purpose of learning the status of the main effect, recidivism, and measuring the strength of the relationship with an inmate and correctional system factors, quantitative methods are the option of choice for a variety of reasons. First, quantitative studies usually encompass larger samples, systematic sampling procedures that permit reliable projections to the population of concern, employ rigorous approaches to minimizing researcher and participant bias, investigate cause-effect relationships or associations between discrete variables, and set methods for enhancing validity. Quantitative approaches also maximize reliability, which means findings can be compared over time or across varying inmate populations (Cozby, 2009).

A quantitative design for the above objectives will comprise three levels of investigation:

1. A tracking study of recidivism rates for one cohort of juvenile inmates in a Connecticut correctional institution. The term “tracking” here is used advisedly because the researcher retains the option for a) embarking on a longitudinal study which will follow the cohort for a period of three years after release (the ASCA cut-off for recidivism benchmarking) on account of having completed the sentence or received probation; or, b) gathering data retrospectively, tracing back recidivism for up to three years before the present confinement. This step in the research design will involve a search on the FBI offender database, backed by extensive checks on state and local police records because not all offenses are reported back to the FBI (Montana Dept. of Corrections, 2008).
2. The antecedent factor will be covered by a custom personal and family history that each juvenile inmate in the cohort will be asked to fill as part of “routine” record-keeping. Covering such risk factors as family socioeconomic status, inmate educational attainment, incidence of gang membership and age at joining, such data will be correlated with recidivism rates to test for a correlation with the predisposing characteristics Barnhart (2009) alludes to.
3. The third major component will be an attitude survey of both inmates and correctional officers. Structured questionnaires will be formulated to quantify which interventions and prison programs have face validity for rehabilitating youthful offenders and helping them avoid recidivism in future. Both inmates and their custodians will be specifically asked for their opinions regarding their overall experience in prison, skills training, remedial education, and job skills training programs.

Qualitative Design

There are at least four stakeholders directly concerned with recidivism outcomes: the offenders themselves, their families, those victimized by the crime, and correctional and judicial officers. Qualitative research approaches are best for examining the attitudes, feelings, resentment, sources of satisfaction or dissatisfaction, and motivations of these stakeholders. Ultimately, the benefit is for the researcher to explore the varying opinions of each stakeholder involved, discover new ideas, and diagnose why recidivism occurs in the first place (Zikmund, 2003).

Given the inhibiting effects of peer pressure, particularly on having to disclose the missteps that lead to committing an offense and being apprehended, focus group discussions are less sound in this context than individual depth interviews (Cozby, 2009).

It is proposed that unstructured depth interviews be conducted with a small sample of thirty juvenile inmates. In the hands of a skilled professional (an outsider since the proponent of this research has administrative authority over the inmates), such a qualitative approach can uncover a rich range of aspirations, attitudes and expectations about leading a reformed life after serving the sentence. Socialization and achievement needs will also be assessed with the semi-projective sentence completion test (Zikmund, 2003) and the self-administered Edwards Personal Preference Schedule paper-and-pencil test (Domino and Domino, 2006).

The victims of the offense and the significant others of inmates will be similarly interviewed about their own expectations and disappointments over the inmate. These two stakeholder groups are important because they comprise the social matrix into which the youthful offender will be released later on.

At the same time, from 15 to 20 correctional judicial personnel will be interviewed about those prison programs that they consider most important in forestalling recidivism and other antisocial behavior in the future.

All in all, this and any other qualitative design will develop rich insights about the attitudes, experiences and risk factors where juvenile offenders are concerned. However, this is traded off against reliability. That is, diagnostic value is high but the findings cannot be projected as representative of even the inmate population in the correctional institution where the research is conducted (Cozby, 2009).

Mixed-Method Design

Recasting the research purpose as an investigation into how prison programs might be improved and made more effective in minimizing recidivism opens up the opportunity for carrying out mixed-methods research.

Trochim and Donnelly (2008) justify mixed methods at the midpoint between naturalistic inquiry and experimental testing. Given that the researcher cannot as yet embark on experimental testing of effective correctional interventions since the potential range is unknown for the population in question, it is proposed that a mixed-methods approach incorporate the qualitative and quantitative designs defined above and that these be supplemented with two techniques.

Recall that the ultimate goal of any prison program is that of reintegrating offenders into society and preventing recidivism, and accomplishing both within budget constraints. It is then helpful to consider developing a cross-institution case study across the three Public Juvenile Detention Centers (Hartford, Bridgeport, New Haven) in the state. Another tool in the mixed-method repertoire will be benchmarking against recidivism statistics for those released from juvenile correctional institutions in the state. Boasting databases that can be queried against uniform definitions of justice variables back to 1995, the Connecticut State Police Crime Analysis Unit (2008) will provide a benchmark against which to determine whether the youth cohort understudy has average, low or high recidivism rates.

Non-Experimental, Quasi-Experimental and Experimental Research Designs

In an experimental study, a researcher aims to investigate:

• Whether an independent variable (IV) does influence the dependent variable (DV) under study.
• Or how much of the IV is required to bring about a desired level of the DV.
• While controlling or holding constant all other IVs and potential intervening or extraneous variables (Zikmund, 2003).

The fundamental goal of experimental design is, logically enough, causal inference: demonstrating that, all other things held constant, an IV does lead to a change in the DV and that alternative explanations have been suitably accounted for (Creswell, 2009).

To illustrate the potential complexity and technical rigor of experimental research design, applying this to the aforementioned, submitted research problem requires considerable thought. One may choose to articulate the general research problem as “what interventions, applied while youthful offenders are in confinement, give the largest and more enduring gains in recidivism?” The choice of recidivism as the unequivocal, readily measurable, and desired end-goal of the correctional system is difficult to implement mainly because too many extraneous variables come into play during the three-year cut-off after the offender has been released. Instead, one should substitute an instrument or professional evaluation of rehabilitation, preferably one for which a benchmark was done at the start of confinement.

On the matter of interventions to be manipulated, one may subject matched groups to such treatments as counseling alone, skills training alone or remedial education alone. (It is also possible, of course, to manipulate these IVs for sub-types or degree of exposure.) If one is found to significantly decrease anti-social attitudes and violence-proneness better than others, the juvenile detention facility may opt to invest more time and money in that rehabilitation intervention.

In contrast, the quasi-experimental designs, otherwise designated field or natural experiments, lack random assignment to conditions. In non-experimental studies, on the other hand, the researcher has little control over the subjects or the research setting and hence, is not in a position to manipulate the IVs. Also termed correlational, passive, naturalistic, and observational (Coryn, 2009), widely-publicized examples of this are those reporting risk factors for lung or colon cancer (e.g., Willet, 2008).

References

1. Barnhart, T. E. (2009). The criminal youth inmate subculture. Corrections Connection.
2. Castillo, R., Sessions, W. K. III, Steer, J. R., Hinojosa, R. H., Horowitz, M. E., O’Neill, M. E., Reilly, E. F. Jr. & Rhodes, D. J. (2004). Measuring recidivism: The criminal history computation of the federal sentencing guidelines. U. S. Sentencing Commission 15-Year Report.
3. Connecticut State Police Crime Analysis Unit (2008). Uniform crime reports: Publications & queriable statistics.
4. Coryn, C. L. S. (2009). What is social science research and why would we want to evaluate it? In C. L. S. Coryn, Evaluating social science research: A handbook for researchers, instructors, and students (pp. 1-14). New York, NY: Guilford.
5. Cozby, P. C. (2009). Methods in behavioral research (10th ed.). Boston: McGraw Hill Higher Education.
6. Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approach (3rd ed.). Thousand Oaks, CA: Sage Publications.
7. Domino, G. & Domino, M. L. (2006) Psychological testing : An introduction (2nd ed.) Cambridge, UK: Cambridge University Press.
8. Gromet, D. M., (2009). Restoration and retribution: People’s negotiation of multiple responses to wrongdoing. (Unpublished Dissertation: Princeton University).
9. Maxfield, L. D. (2005). Measuring recidivism under the federal sentencing guidelines. Federal Sentencing Reporter, 17 (3): 166-170.
10. Montana Dept. of Corrections (2008). Recidivism definition adopted. Correctional Signpost, pp. 1-2.
11. Trochim, W., & Donnelly, J. (2008). The research methods knowledge base (3rd ed.). Mason, OH: Cengage.
12. Willett, W. C. (2008). Epidemiologic studies of diet and cancer. Medical Oncology, 7 (2-3): 93-97.
13. Zikmund, W. (2003). Business research methods (7th ed.). Thousand Oaks, CA: Thomson/South-Western.