Frequently asked questions in Common multi variate techniques

Many a times, some of my friends asked about frequently asked questions in multivariate techniques at interviews, conferences etc. But, infact these questions do not have any limited scope but i tried a littel in this way here is my prefered questions.

Multi variate techniques:

Multiple regression analysis:

• What is difference b/n  multiple regression and multivariate regression
• How to select inde variables in to the system?
• What are the measures of efficiency?
• What is the specific proc in SAS
• Assumptions underlying and the consequences of their violations.
• Estimation techniques, adv and dis adv

Logistic regression analysis:

• Difference b/n logistic and traditional regression
• Assumptions if any?
• Estimation method?
• Efficiency measures?
• Which domain having major applications
• Odds ratio implementation
• Tests of goodness of fit

Descriminant analysis:

• What is the aim of descriminant analysis
• Methods of constructing descriminant functions
• Fisher discriminant function
• Issue of multi collinearity here
• What is cluster descriminant
• Domain applications

MANOVA:

• Tests of MANOVA
• Structure of model in manova
• Assumptions of MANOVA
• How to use and read in SAS environment
• Difference b/n manova and multiple regression
• PROC MANOVA

Factor analysis:

• What is the difference between PC and PAF (Principle factor analysis?)
• What is a Simple or Clean Factor Structure?
• Types of factor analysis
• Applications in manufacturing
• PROC PRINCOMP, how to improve the performance
• Type of conclusions in FA

Multi dimensional scaling:

• Types of multi dimensional scaling
• How to decide on what dimensions respondents use when evaluating objects
• how many dimensions they may use in a particular situation
• test for the relative importance of each dimension
• how the objects are related perceptually

Correspondence analysis:

• How to use it in market research?
• How to read the parameters?
• Any significant tests
• How to apply in SAS

Conjoint analysis:

• Advantages in market research
• Types of conjoint analysis
• Relation with regression and logistic regression
• Latest developments
• Steps in the design of the studies

Cluster analysis:

• Why it comes under multivariate techniques
• How to choose the variables for the clustering
• What are the types of clustering?
• Measures for efficiency of clustering
• Reports based on clustering

Canonical correlation:

• Why it is important than usual correlation
• What is the complexity involved here
• Application area

Structural equation modeling:

• Why it is so significant
• Applications in SAS
• How to interpret the results