Both the correlation and regression analysis are parametric techniques that test for linear association between two or more variables. Correlation analysis tries to quantity the degree to which two or more variables are related and tells the direction of the relationship. Contrary, the regression analysis aims to find the best line that predicts a dependent variable from a predictor variable. One of the research studies that adopts the use of regression technique is by Kanyongo, Certo, and Launcelot, 2006, that predicted reading achievement using home environment factors. Tentatively, Kanyongo et al. (2006) used Grade 6 students from Zimbabwe to carry out the research. The study found that Social Economic Status (SES) factors, characterized by possession of things like TVs, refrigerators, meals, piped water, electricity, were good predictors of the reading performance among the Grade 6 students.
In this scenario, the authors used regression to test if there was a linear association between environmental factors and the reading performance, and if there was, what was the magnitude of the relationship. First, the authors had a goal to assess the environmental factors that influenced the reading performance among the selected participants. Therefore, this tells that the key objective was to test if the selected variables were linearly associated with the reading performance. Second, the authors aimed to assess the best predictor variables of the reading performance and this tells that among the environmental factors selected, there is a chance that not all of them can predict the reading performance (Christensen et al. 2011). Therefore, with the use of regression, this research goals were achievable.
Delegate your assignment to our experts and they will do the rest.
Regression analysis is suitable in this type of study, hence the best choice. First, it has the statistical power to achieve the research goals based on a number of factors. First, the on top of identifying the linear relationship of the study variables chosen, it has the power to identify which factors significantly predict the dependent variable. Still in this, the basic assumptions of using the regression analysis were met. The distributions of the data scores collected were normally distributed, there was no auto-correlation, and they were assumed to have a linear relationship backed with sufficient secondary research – key elements which are basic assumptions for the regression analysis (Vogt, 2006). Tentatively, the type of environment that students learn in, affect their performance – however, it is hard to tell which ones can significantly predict the reading performance for the selected Grade 6 students. Further, the nature of the data and their respective scales of measurement met the basic requirement for testing regression. The dependent variable was measured in a ratio scale while the SES scores were measured on an ordinal scale indicating that the data was suitable for this type of test (Vogt, 2006).
The authors displayed partial data for the study findings. Noticeably, the data for only the independent SES variables chose and the regression weights were presented. Tables for the regression output including the ANOVA table and coefficients were left out. It is easy to understand on what the independent variable analysis communicates but for the association between the reading scores (the dependent variable) and these predictors, it is not clear since no table is provided (Vogt, 2006). With this in place, the study has a shortcoming since it is hard to understand what the data in the tables indicate from a mere look. Such an aspect has impeded the ability of these findings to stand alone when it comes to communicating the study outcome. Personally, I cannot interpret the results, especially on the authors reached the conclusions. Even the model summary output and the F-statistic is not provided and it is hard to understand where the results came from and this is a major weakness identified in the study despite the correct choice of regression as the analysis technique.
References
Christensen, L. B., Johnson, B., Turner, L. A., & Christensen, L. B. (2011). Research methods, design, and analysis. Allyn & Bacon Boston, MA.
Vogt, W. P. (2006). Quantitative research methods for professionals in education and other fields. Columbus, OH: Allyn & Bacon .
Kanyongo, G. Y., Certo, J., & Launcelot, B. I. (2006). Using Regression Analysis to Establish the Relationship between Home Environment and Reading Achievement: A Case of Zimbabwe. International Education Journal , 7 (4), 632-641.
