Question

Which of the following statements are not true with regard to multivariate methods of data analysis?
A. Structural equation model with latent variables is exploratory type of multivariate analysis
B. Discriminant analysis may be exploratory or confirmatory in nature
C. Logit choice model is a multivariate method of data analysis
D. ANOVA is a special case of canonical correlation with discrete Y variable
E. Canonical correlation explains co-variation between two sets of multiple variables
Choose the correct answer from the options given below:

• A. A and E only
• B. A and D only
• C. A, B and C only
• D. C, D and E only

Hint:

Remember that Structural Equation Modeling (SEM) is always a confirmatory technique because it requires the researcher to specify the path model before running the analysis.

Answer 1

The Correct Option is B

Step 1: Understanding the Concept:
Multivariate analysis involves statistical methods used to analyze datasets that contain multiple variables measured on each experimental unit.

Step 2: Detailed Explanation:
Let us evaluate each of the statements to identify which ones are not true:
- Statement A is not true: Structural Equation Modeling (SEM) with latent variables is a confirmatory multivariate technique, not an exploratory one.
It is used to test pre-specified theoretical models and hypotheses about relationships among observed and latent variables.
- Statement B is true: Discriminant analysis can be used to describe differences between groups (exploratory) or to classify new observations into pre-defined groups (confirmatory).
- Statement C is true: The logit choice model (logistic regression) is a multivariate technique used to analyze relationships where the dependent variable is categorical.
- Statement D is not true: Analysis of Variance (ANOVA) is a univariate or multivariate technique that compares means across groups.
It is a special case of the General Linear Model, but it is not a special case of canonical correlation with a discrete dependent (Y) variable.
Canonical correlation is designed to analyze relationships between two sets of continuous variables.
- Statement E is true: Canonical correlation analysis is used to identify and quantify the association between two sets of multiple variables.
Therefore, statements A and D are not true.

Step 3: Final Answer:
The statements that are not true are represented by option (B).