Analyze and critique of another student's discussion on uni-variate and multivariate statistics

DIRECTIONS:

After reading the following student’s discussion, you are to offering your analysis of their plan to make the univariate study into a multivariate study. Be sure to support your critique with specific reasons and examples.

Top of Form

U01D1 – Comparing Univariate and Multivariate Statistics

In the book Applied multivariate statistics for the social sciences: Analyses with SAS and IBM’s SPSS Pituch and Stevens (2015) defined multivariate statistics as the method by which multiple – two or more – dependent variables. Definition: Conversely, univariate statistics are comprised of statistics approaches appropriate for analyzing a single dependent variable (Warner, 2013).

Similarities and Differences

Given that multivariate statistics is based on the existence of multiple dependent variables, many similar functions exist between the two approaches. Univariate and multivariate approaches share assumptions of central tendency/normality, independent observations and equality of population variance/homogeneity. In addition, multivariate approaches include assumptions of the absence of multivariate outliers, linearity of data, absence of multicollinearity and the assumption of equity of covariance matrices (Pituch & Stevens, 2015). Similarities also exist in the discussions of error terms, and significance tests conducted. While letter terms for each function might differ slightly, the concepts used in univariate and multivariate approaches are similar. Difference begin to emerge as calculations and concepts used in multivariate statistics take on a matrixed approach to calculation, while approaches used in univariate statistics remain somewhat two dimensional in nature (Pituch & Stevens, 2015; Warner, 2013).

Examples of Univariate and Multivariate Use

For the purposes of this discussion, a univariate study might include the application of testing a particular approach to improving employee satisfaction in work environment. A univariate approach might consider the differences in employee performance between two groups of participants. One group would act as the control group having no changes in how they were trained for a particular task. The second group, or experimental group might be involved in a pilot study of a new training approach. This example would provide potential differences between two employee groups using one dependent and one independent variable, with a control in place. A multivariate approach would differ in that the multivariate approach could include more than two groups of employees, all being tested with differing approaches to see if any one group might differ significantly from other groups in terms of employee satisfaction while at the same time testing for changes in productivity and absenteeism. In this example, more than two dependent variables would be used to see if changes in training affected multiple variables.

References

Pituch, K. A., & Stevens, J. P. (2015). Applied multivariate statistics for the social sciences: Analyses with SAS and IBM’s SPSS (6th ed.). New York, NY: Routledge.

Warner, R. M. (2013). Applied statistics: From bivariate through multivariate techniques (2nd ed.). Thousand Oaks, CA: Sage. Retrieved from http://online.vitalsource.com/books/9781452268705