2021
2022
MARKET RESEARCH
METHODS – THEORY SPSS
PROF. VAN KENHOVE / L. DE KERPEL
SAM STROO
, PART 1: FACTOR ANALYSIS
THEORY RECAP
o Factor analysis is an interdependent technique: we do not make a distinction between what is the
dependent and what is my independent variable
o Look for patterns in our data: end goal is data reduction while preserving as much info as possible
o Discover latent dimensionality: you need to dig deeper to find it
o We are looking at which variables are highly correlated variables (that measure more or less the same
constructs) and group them together to form factors (end goal)
o In FA we group variables by their correlations, such that variables in a group (= factor) have high
correlations with each other, to explain as much variance as possible
TYPES OF FACTOR ANALYSIS
Exploratory: we don’t know how many factors we will get in the end/ confirmatory does start from
assumptions maybe from previous research
ð R factor analysis: groups a set of variables into latent dimensions/factors based on similar response
patterns (better: PCA)
ð Q factor analysis: groups large numbers of respondents into distinct different groups based on similar
response patterns in the variables in the analysis (better: cluster analysis)
WHAT IS VARIANCE? Total variance consists of common variance, unique variance & error variance
- Shared variance between 2 variables = squared correlation E.g. r = 0.50 means that a variable shares
25% of its variance with the other variable
- Total variance of a variable:
1) Common variance: variance in a variable that is shared with all other variables in the analysis (~
communality)
2) Specific variance or unique variance: variance associated with only one specific variable
3) Error variance: variance due to noise
CFA (common factor analysis, only considers common variance with the goal to uncover latent dimensions) ≠
PCA (data reduction, with the goal of minimum number of factors with max variance, considers total variance)
SAM STROO 1
2022
MARKET RESEARCH
METHODS – THEORY SPSS
PROF. VAN KENHOVE / L. DE KERPEL
SAM STROO
, PART 1: FACTOR ANALYSIS
THEORY RECAP
o Factor analysis is an interdependent technique: we do not make a distinction between what is the
dependent and what is my independent variable
o Look for patterns in our data: end goal is data reduction while preserving as much info as possible
o Discover latent dimensionality: you need to dig deeper to find it
o We are looking at which variables are highly correlated variables (that measure more or less the same
constructs) and group them together to form factors (end goal)
o In FA we group variables by their correlations, such that variables in a group (= factor) have high
correlations with each other, to explain as much variance as possible
TYPES OF FACTOR ANALYSIS
Exploratory: we don’t know how many factors we will get in the end/ confirmatory does start from
assumptions maybe from previous research
ð R factor analysis: groups a set of variables into latent dimensions/factors based on similar response
patterns (better: PCA)
ð Q factor analysis: groups large numbers of respondents into distinct different groups based on similar
response patterns in the variables in the analysis (better: cluster analysis)
WHAT IS VARIANCE? Total variance consists of common variance, unique variance & error variance
- Shared variance between 2 variables = squared correlation E.g. r = 0.50 means that a variable shares
25% of its variance with the other variable
- Total variance of a variable:
1) Common variance: variance in a variable that is shared with all other variables in the analysis (~
communality)
2) Specific variance or unique variance: variance associated with only one specific variable
3) Error variance: variance due to noise
CFA (common factor analysis, only considers common variance with the goal to uncover latent dimensions) ≠
PCA (data reduction, with the goal of minimum number of factors with max variance, considers total variance)
SAM STROO 1
