Multiple Regression in Practice

Answer

Quantitative Article Critique- Multiple Regression as a Practical Tool for Teacher Preparation Program Evaluation

This paper is a critical evaluation of the article by Williams (2012) entitled, Multiple regression as a practical tool for teacher preparation program evaluation. The aim of the study was based on the rationale that with regards to the No Child Left Behind mandates, a number of accountability demands and budget cuts aimed at improving institutional programs required “…practical, quantitative evaluation methods…” (Williams, 2012, p. 1) for utilization educators in various educational institutions. These institutions were in great need of practical and quantitative evaluation techniques which could be internally employed to appropriately examine teacher based preparation measures and its relation to effective coursework. Williams (2012) utilized multiple regression techniques as a tool for linking teacher certification outcomes and coursework. The experiment was repeated under uniform investigational conditions as the participants were assigned different treatment and control groups. The approach, therefore, established an evidence of causation between the variables by showing existence of associations between the variables (teacher certification outcomes and coursework) (Williams, 2012). The technique contains the capacity to sort out the magnitude and existence of causal effects of one or more independent variables upon a dependent variable of interest at a given time.

The research question in the study by Williams (2012) sought to establish a correlation between teacher preparation programs at the course and student level in a large public and a small private university and the resultant performance. The null hypothesis for the study outlined that the quality of teacher preparation programs exhibited a direct impact on the quality of student course at college levels. The variable were measured by establishing a causal relationship between the three metric variables, one continuous dependent variable and two independent variables. The rationale behind the use of multiple regression analysis was that the independent variables have a direct effect on the stability of the dependent variable. This existing correlation among the various variables was assessed using the coefficient of determination (Williams, 2012); Frankfort & Leon, 2018).

The variables were examined using two different case studies whereby one particularly examined data generated from a smaller unit of privatized university and the other examined information generated from a wider public university. The statistical considerations that were focused upon included missing or confounding variables, power, grade inflation, bivariate correlations, statistical assumptions as well as beta weights. Randomization was employed to facilitate assigning of participants to various groups to eliminate selection bias and any confounding bias while ensuring the associate groups are comparable regardless of any other factors other than the ones under investigation. This was an experimental study with greater internal validity.

Using multiple regression was the best approach because several educational institutions like colleges are currently seeking better practical procedures to facilitate evaluation of teacher preparation programs and coursework to promptly identify points of strength and weaknesses. There was a growing urge to employ a useful quantitative model that can directly link the tutor preparation coursework to the results on teacher certification assessment programs. Multiple regressions was the most appropriate tool for examination of utility and generalizability in evaluating the preparation programs employed by teachers at student and course levels in small private colleges and public institutions within north central Texas. The state exam certification results were employed as the primary measure for the student’s level of success. Moreover, multiple regressions was particularly used to facilitate study examination of the utility of model as employed in answering specified questions with regards to the importance of introducing additional program in the study performance. The approach provided answers to questions on whether additional specific courses were essential, the specific courses that best prepare teachers for their profession as well as the courses that needed to be restructured. Williams (2012) displayed the study data. A >0.5 effect was, however, noted, which was a large difference effect. The null hypothesis was strongly opposed because all the fore mentioned factors had a direct impact on the analysis and evaluation of variables related to teacher preparation programs and coursework for examining education performance. The results from the various groups were comparable. The effect was estimated using an evaluation design where the effect size is estimated by taking the difference between the mean of treatment group minus the mean of the control group and then divided by the standard deviation of either of the two groups.

The results stood alone and indicated that multiple regression contains the potentiality to provide a highly meaningful information regarding program evaluation during the examination of teacher preparation programs in situations that offer fewer course sections, especially at the private university level. The variance relating to multiple course sections currently nested in individualized courses was found to have effects on multiple regression outcomes for public university analysis. Procedural methods of hierarchical linear modeling (HLM) and growth mixture modeling (GMM) were regarded to contain more validity and appropriateness during evaluation of teacher preparation programs, especially for the case of larger university institutions, where all nested variables were considered to be more prevalent.