You are on page 1of 2

Chapter 8 - Factor Analysis

Factor analysis identifies unobserved variables that explain patterns of correlations within a set of observed variables. It is often used to identify a small number of factors that explain most of the variance embedded in a larger number of variables. Thus, factor analysis is about data reduction. It can also be used to generate hypotheses regarding the composition of factors. Furthermore, factor analysis is often used to screen variables for subsequent analysis (e.g., to identify collinearity prior to performing a linear regression analysis). There are three types of factor analyses, namely exploratory factor analysis, confirmatory factor analysis, and structural equation modeling. The first two techniques are identical from a statistical point of view; however, they are used in different ways. Exploratory factor analysis is used to reveal the number of factors and the variables that belong to specific factors. When we conduct a confirmatory factor analysis, we have clear expectations regarding the factor structure (e.g., because we make use of a previously used survey) and we want to test if the expected structure is indeed present. Structural equation modeling differs from those two techniques, both statistically and practically. It is used to evaluate how well variables relate to factors and what the relationships between the factors are. This technique will be briefly discussed at the end of this chapter. In this chapter, we primarily deal with exploratory factor analysis as it conveys the principles that underlie all factor analytic procedures. Two exploratory factor analytic procedures are commonly used in market research: principal components analysis and principal axis factoring. When the purpose is to summarize information (variance) represented by the variables using a small number of factors, principal components analysis is used. Alternatively, if the aim is to identify underlying factors or dimensions that reflect what communalities variables share, principal axis factoring is used. Principal components and principal axis factoring essentially require the same analysis steps and involve the same interpretation. However, as the concept of principal components analysis is easier to understand, we focus on this approach. After reading this book chapter you should understand

the principles of exploratory and confirmatory factor analysis the difference between principal components and principal axis factoring key terms such as Eigenvalues, communality, factor loadings and factor scores how to determine whether data are suitable for carrying out an exploratory factor analysis how to interpret SPSS factor analysis output the principles of reliability analysis and how to carry it out in SPSS the basic idea behind structuralequation modeling

Exploratory and confirmatory factor analysis ... principal components analysis ... principal axis factoring ... Eigenvalue ... factor loadings ... factor scores ... communality ... antiimage ... Kaiser-Meyer-Olkin (KMO) criterion ... Bartletts test of spericity ... measure of sampling adequacy (MSA) ... Kaiser criterion ... scree plot ... orthogonal and oblique

rotation ... varimax and direct oblimin rotation ... reliability analysis ... Cronbachs Alpha ... structural equation modeling ... LISREL ... PLS path modeling

You might also like