Research studies indicate that one of the daunting tasks that many research students usually faces while conducting their quantitative research is identifying the appropriate statistical approach to use while analyzing their study (Mertler & Vannatta, 2010). Fortunately, multivariate has solved the problem by providing the capacity to examining two or more variable and their in-depth relationship. Multivariate analysis models can be classified into two main groups- logit or discriminant. Discriminant mainly addresses the structure of the data extensively while the logit is slightly demanding in terms of assumptions (Roberts, 2008).
In qualitative research, multivariate analysis may be appropriate at different circumstances. These circumstances include; when identifying critical design drivers and correlations across hierarchical drivers, addressing concepts in changing scenarios, analysis of alternatives. The application of multivariate analysis in these circumstances provides various advantages. These are; validating the scale of the index, helping to select a set of variables from a relatively larger set and finally reducing the higher number of variables to relatively smaller data that may be easier to model. In order to fully utilize the multivariate analysis in a qualitative analysis, certain statistical conditions must be met. First, the co-variance/variance matrices of the predators should be equitable, second, the normally distributed predators must mitigate against the use of independent variables. Violation of these two principles is a great offence while carrying out a quantitative analysis (Roberts, 2008).
In conclusion, when investigating the relationship between two or more quantitative variables, correlation or regression, usually an appropriate test is required. These tests may include path analysis (Mertler & Vannatta, 2010). Path analysis uses multiple applications of multiple regressions to estimate both direct and indirect causal relations. As stipulated, before any valid research is conducted, it is important for the researcher to hypothesize the model based on the previous theory and research. Thereafter, the model is graphically represented in the form of a path diagram to estimate the strength and the relationship of the model (Mertler & Vannatta, 2010).
2+ IV (quantitative)
1+ DV (quantitative)
Roberts, K. (2008). The Analysis of Organizational Performance Utilizing Multivariate Analysis. Economics, Management and Financial Markets, 3(2), 13-30. Retrieved from http://search.proquest.com/docview/1032550833?accountid=1611
Mertler, C. A., & Vannatta, R. A. (2010). Advanced and multivariate statistical methods: Practical application and interpretation. Glendale, CA: Pyrczak.