Respond to at least two of your colleagues’ postings
Post an analysis of the difference between causation and correlation within the context of your DBA doctoral research study. In your analysis, do the following:
· Assess the implications for professional practice when a researcher implies causation after using correlation (e.g., bivariate correlation) analyses.
· Explain why the results of bivariate correlation analyses are considered weak in terms of internal validity.
· Explain how would you extend or modify a research design to examine a true cause-and-effect relationship.
Guillermo
Implications for professional practice when a researcher implies causation after using correlation analyses.
Correlation is a statistical technique to describe the effect between two or more phenomena when they occur together. However, the fact that phenomena occur together, does not mean that one causes the other (Cowls & Schroeder, 2015). To establish causation, the researcher must ensure that the data used to calculate correlation were obtained experimentally, taking care to control or eliminate any other reasonable explanations (confounding threats) that may challenge the validity of the experiment (Bailey, Duncan, Watts, Clements, & Sarama, 2018). When researcher implies causality, practitioners may invest considerable time and effort to manipulate the independent variable without achieving the expected outcomes. In addition to the financial cost, the validity of the research, the theory, and the author may come into question (Bleske-Rechek, Morrison, & Heidtke, 2015).
Strength of bivariate correlation analyses.
Cowls & Schroeder (2015) explain that data processing is fast and inexpensive, a very large number of correlations can be calculated quickly for very large samples. However, correlation is essential, but not enough to make causal inferences with reasonable confidence (Bailey, et al., 2018). As mentioned earlier, correlation only measures the change between variables. The strength of the correlation indicates a certain alignment but does not provide enough information to understand what drives the changes. There are other, more advanced statistical models, like linear regression, that are more appropriate to substantiate causality.
Research Design to Establish Cause and Effect relationship.
Establishing cause and effect relationships demands design and implementation rigor. In addition to providing a clear hypothesis and adequate sample size, the rubric demands the use of a minimum of two predictor variables to avoid the confounding or mediator effects of a third variable (Walden University, n.d. b). The use of at least two predictor variables is the minimum requirement. The choice of statistical method can help ensure that appropriate assumptions are met and that potential confounding variables are eliminated to improve the validity of the results.
References
Bailey, D. H., Duncan, G. J., Watts, T., Clements, D. H., & Sarama, J. (2018). Risky business: Correlation and causation in longitudinal studies of skill development. American Psychologist, 73, 81-94. doi: 10.1037/amp0000146
Bleske-Rechek, A., Morrison, K. M., & Heidtke, L. D. (2015). Causal inference from descriptions of experimental and non-experimental research: Public understanding of correlation-versus-causation. Journal of General Psychology, 142(1), 48–70. doi:10.1080/00221309.2014.977216
Cowls, J., & Schroeder, R. (2015). Causation, correlation, and big data in social science research. Policy & Internet, 7(4), 447-472. doi:10.1002/poi3.100
Walden University. (n.d.b). DBA doctoral study rubric and research handbook. Retrieved from http://academicguides.waldenu.edu/researchcenter/osra/dba
Luke
Assess the implications for professional practice when a researcher implies causation after using correlation (e.g., bivariate correlation) analyses.
Correlation occurs when a change in one variable will result in another variable to change (Coogan, 2015). Causation, on the other hand, is present when a force exists that explains the correlation from one variable acing on another variable (Coogan, 2015). Therefore, causation relies on correlation. If a researcher implies causation after correlation, their study may lack reliability and validity.
Explain why the results of bivariate correlation analyses are considered weak in terms of internal validity.
The bivariate correlation called the Pearson product-moment correlation coefficient assesses the linear relationship between quantitative variables in a sample (Green & Salkind, 2017). To pinpoint potential influential outliers in a data set, often scatterplots are used when conducting the bivariate correlation test (Green & Salkind, 2017). Based on the above definitions of correlation and causation, I believe correlation analyses are considered weak in terms of internal validity due to the fact that they do not explain the cause of the results, only that there is or is not a correlation.
Explain how would you extend or modify a research design to examine a true cause-and-effect relationship.
Managers might think there is a casual link when interpreting data, but in reality a cause-effect relationship does not exist (Porporato, Tsasis, & Maria Marin Vinuesa, 2017). One potential way to examine a true cause-and-effect relationship it to use the balanced scorecard concept. The balanced scorecard provides a framework where the primary objective is to turn strategic organizational goals into a balanced perspective throughout numerous organizational performance dimensions (Porporato, Tsasis, & Maria Marin Vinuesa, 2017). From what I learned, the balanced scorecard is an effective way to examine a true cause-and-effect relationship.
References
Coogan, L. L. (2015). Teaching across Courses: Using the Concept of Related Markets from Economics to Explain Statistics’ Causation and Correlation. B>Quest, 1–10. Retrieved from http://www.westga.edu/~bquest/
Green, S. B., & Salkind, N. J. (2017). Using SPSS for Windows and Macintosh: Analyzing and understanding data (8th ed.). Upper Saddle River, NJ: Pearson.
Porporato, M., Tsasis, P, & Maria Marin Vinuesa, L. (2017). Do hospital balanced scorecard measures reflect cause-effect relationships? International Journal of Productivity and Performance Management, (3), 338.
doi:/10.1108/IJPPM-02-2015-0029