Respond to at least two colleagues by explaining how that colleague might rule out one of the confounding variables that they identified.
Colleague 1: Debby
Being able to look at the different designs and choosing the right design for the information necessary to give an accurate accounting is imperative. Looking at the variables and outcomes wanting to be measured is also an important part of choosing a statistical design. The outcome of the design should be able to tell whether the goals of the client have been met (Dudley, 2014). In the study Social Work Research: Chi Square (Plummer, Makris, & Brocksen, 2014b), the outcome of the client was the outcome data measured.
The intervention provided by the organization was to rehabilitate recently paroled prison inmates and get these clients ready for full-time employment (Plummer, et. al., 2014b). The design was to use a quasi-experimental research design and the program started with thirty recently paroled clients, the intervention group (Plummer, et. al., 2014b). There was also another thirty recently paroled individuals that were waiting to enter the rehabilitation program, the comparison group (Plummer, et. al., 2014b). The parole officers of each individual within both the intervention group as well as the comparison group were provided surveys regarding the employment and demographics of the individual (Plummer, et. al., 2014b). The independent variable (rehabilitation program group) and the dependent variable (employment outcome), were measured using the Pearson chi-square and compared to the comparison group.
This study found the difference in the two groups were highly significant with a p value of .003 which is beyond the usual alpha-level of .05 which is used by researchers to determine the significance of the design used (Plummer, et. al., 2014b). This type of findings would give the organization reason to believe that the rehabilitation program could be effective when working with these clients in being able to obtain full-time employment (Plummer, et. al, 2014b).
The validity of the rehabilitation program may be compromised by the two groups selected for the study. For example, there was no random selection when choosing the groups. Also, gaining employment may or may not prove that these individuals can maintain employment and for how long. This type of study would need a random selection of the groups as well as follow-up for a specific amount of time in order to follow how these groups were able to maintain the full-time employment. The individuals that did not find employment and the individuals that found part-time employment would also need to be followed to measure whether full-time employment was achieved after a period of time. This type of study gives a basic measurement but in order to truly get a valid conclusion regarding the rehabilitation program and the ability to gain and maintain full-time employment, there would need to be further evaluation.