Multivariate Data Analysis 7th Edition Chapter1

...

Q5. Why is the knowledge of measurement scales important to an understanding of multivariate data analysis?

Because first we need to avoid mistakenly using metric data as nonmetric data or using nonmetric data as metric data. Secondly, the measurement scale is also critical because it can help decide which multivariate data analysis technique is the most applicable.

Q6. What are the differences between statistical and practical significance? Is one a prerequisite for the other?

The differences between them are very obvious. Statistical significance ignores which kind of results they might be Is it good or is it bad. Practical significance helps to further tell people that the result is a good one and it is applicable in the real world.

Q7. What are the implications of low statistical power? How can the power be improved if it is deemed too low?

If the statistical power is too low, it means that we cannot find difference between the two objects we are analyzing.

We can make the statistical power increase by lower β. To lower β, we can increase the sample size, increase α, or increase the effect size.

Q8. Detailed the model-building approach to the multivariate analysis, focusing on the major issues at each step.

Textbook mentions a six-step model-building approach for multivariate analysis.

Step1: Define the research problem, objectives, and multivariate technique to be used

We need to focus that a concept in the concept model we build, rather than a specific variable is defined both in dependence and interdependence situations.

Step2: Develop the analysis plan

Step3: Evaluate the assumptions underlying the multivariate technique

The researcher must make sure that both statistical and conceptual assumptions are met before any model estimation is made.

Step4: Estimate the multivariate model and assess overall model fit

After deciding the overall model, the researcher must also determine if the results are undulyaffected by any single or a small set of observations that indicate the result at last unstable or not grenerilized.

Step5: Interpret the variates

Step6: Validate the multivariate model

Validate the model by diagnostic analysis before accepting the results.