Advanced Research Methods part B – Videoclips Factor Analysis
Purpose
Estimate a model which explains variance/covariance between a set of observed variables
(in a population) by a set of (fewer) unobserved factors & weightings.
- Observed variables: data en observed variables verzameld bijvoorbeeld via een
survey
- Je wil de variance/covariance between a set of observed variables begrijpen hoe
interrelateren ze met elkaar
-
Example
What is factor analysis?
- Examines the interrelationships among a large number of variables and then attempts
to explain them in terms of their common underlying dimensions factors
- Interdependence technique: in that an entire set of interdependent relationships is
examined without making the distinction between dependent and independent
variables hoe interrelateren de items met elkaar, het zijn geen voorspellingen
- Define structure among observed variables
- Interrelationships among large number of variables to identity underlying dimensions
(factors)
- Is primarily used for data summarization and reduction
o Data summarization: derives underlying dimensions that, when interpreted
and understood, describe the data in a much smaller number of concepts,
than the original individual variables
o Data reduction: extends the process of data summarization by deriving an
empirical value (factor score) for each dimensions (factor) and then
substituting this value for the original values de data verkleinen om ze in
andere analyses te kunnen gebruiken
- Factor analysis is used in the following circumstances:
o To identity underlying dimensions, or factors, that explain correlations among
a set of variables
o To identity a new, smaller, set of uncorrelated variables to replace original set
of correlated variables in subsequent multivariate analysis (e.g. regressions or
ANCOVA)
1
, Recap: measurement model
E = construct, X31 = items en E21 = measurement errors (systematic biases)
Multi-item measurement
- Increases reliability and validity of measures
- Allows measurement assessment:
o Measurement error je kunt de measurement error verhogen en beoordelen
en op basis hiervan de reliability en validity bepalen
Reliability: de punten liggen bij elkaar
Validity: ze liggen on target
- Two forms of measurement models:
o Formative (emerging): de items samen vormen het construct
o Reflective (latent): de items reflecteren het construct
Reflective measurement models
- Direction of causality if from construct to measure
- Correlated indicators de items correleren met elkaar en de correlatie wordt
gebruikt in de factor analysis
- Takes measurement error into account at the item level
- Validity of items is usually tested with factor analysis
E is construct, E1 is measurement error, lambda is factor loading
Applications
- Assess the validity of construct measurements
- Market segmentation for identifying the underlying variables on which to group the
customers
- Product research: determine the brand attitudes that influence consumers choice
- Price management: identity the characteristics of price-sensitive consumers
!Anything where you would like to assess higher-order dimensions!
2
Purpose
Estimate a model which explains variance/covariance between a set of observed variables
(in a population) by a set of (fewer) unobserved factors & weightings.
- Observed variables: data en observed variables verzameld bijvoorbeeld via een
survey
- Je wil de variance/covariance between a set of observed variables begrijpen hoe
interrelateren ze met elkaar
-
Example
What is factor analysis?
- Examines the interrelationships among a large number of variables and then attempts
to explain them in terms of their common underlying dimensions factors
- Interdependence technique: in that an entire set of interdependent relationships is
examined without making the distinction between dependent and independent
variables hoe interrelateren de items met elkaar, het zijn geen voorspellingen
- Define structure among observed variables
- Interrelationships among large number of variables to identity underlying dimensions
(factors)
- Is primarily used for data summarization and reduction
o Data summarization: derives underlying dimensions that, when interpreted
and understood, describe the data in a much smaller number of concepts,
than the original individual variables
o Data reduction: extends the process of data summarization by deriving an
empirical value (factor score) for each dimensions (factor) and then
substituting this value for the original values de data verkleinen om ze in
andere analyses te kunnen gebruiken
- Factor analysis is used in the following circumstances:
o To identity underlying dimensions, or factors, that explain correlations among
a set of variables
o To identity a new, smaller, set of uncorrelated variables to replace original set
of correlated variables in subsequent multivariate analysis (e.g. regressions or
ANCOVA)
1
, Recap: measurement model
E = construct, X31 = items en E21 = measurement errors (systematic biases)
Multi-item measurement
- Increases reliability and validity of measures
- Allows measurement assessment:
o Measurement error je kunt de measurement error verhogen en beoordelen
en op basis hiervan de reliability en validity bepalen
Reliability: de punten liggen bij elkaar
Validity: ze liggen on target
- Two forms of measurement models:
o Formative (emerging): de items samen vormen het construct
o Reflective (latent): de items reflecteren het construct
Reflective measurement models
- Direction of causality if from construct to measure
- Correlated indicators de items correleren met elkaar en de correlatie wordt
gebruikt in de factor analysis
- Takes measurement error into account at the item level
- Validity of items is usually tested with factor analysis
E is construct, E1 is measurement error, lambda is factor loading
Applications
- Assess the validity of construct measurements
- Market segmentation for identifying the underlying variables on which to group the
customers
- Product research: determine the brand attitudes that influence consumers choice
- Price management: identity the characteristics of price-sensitive consumers
!Anything where you would like to assess higher-order dimensions!
2
