Multiple Factor Analysis
Also known as: MFA, MFA multiple
Multiple Factor Analysis (MFA) is a dimension reduction technique developed by Escofier and Pagès (1985) for analyzing multiple groups of variables measured on the same observations. MFA balances the influence of each variable group to provide a unified view of how observations relate across multiple perspectives.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Apply MFA when analyzing multiple related datasets on the same objects (e.g., survey responses plus behavioral data plus physiological measurements), when you want equal weight for each data domain regardless of inherent variance, or when integrating diverse data sources. Ideal for sensory evaluation, consumer research, and multi-method assessment.
Strengths & limitations
- Balances variable groups: prevents high-variance groups from dominating the analysis
- Integrated view: reveals overall structure while respecting each data dimension
- Flexible grouping: groups can have different numbers of variables and different types
- Interpretability: provides both group-level and global-level insights
- Visualization: component plots reveal patterns across and within groups
- Assumes linear relationships: non-linear patterns are not captured
- Group specification: requires meaningful a priori grouping of variables; poor grouping biases results
- Sample size: needs sufficient observations relative to total variables
- Complexity: interpretation requires understanding multiple tables simultaneously
Frequently asked
How do I decide how to group variables?
Groups should be theoretically meaningful or practically distinct. For example: psychological scales form one group, demographic variables another. Groups need not be equal size, but very small groups may be unstable.
What does 'balancing' variable groups mean?
MFA normalizes each group by dividing by the sum of its eigenvalues from group-level PCA. This ensures no group dominates due to higher inherent variance, giving equal weight to each domain.
Can MFA handle categorical variables?
Standard MFA assumes continuous variables. For categorical data, use Multiple Correspondence Analysis (MCA) or its variants. Mixed data can be handled by pre-processing (dummy-coding categorical variables).
How many components should I extract?
Use scree plot or cumulative variance explained. Typically extract 2-4 components that explain 70-80% of variance. More components become hard to interpret.
How do I interpret partial axes in MFA?
Partial axes show how each variable group contributes to the global dimension. Plotting both global and partial points reveals whether all groups align or if some diverge on a dimension.
Sources
- Escofier, B., & Pagès, J. (1985). Analyses factorielles simples et multiples : Objectifs, méthodes et interprétation. Dunod. ISBN: 9782040116835
- Pagès, J. (2004). Multiple Factor Analysis by Example Using R. Chapman and Hall/CRC. ISBN: 9781482234700
- Abdi, H., & Valentin, D. (2013). Multiple Factor Analysis. John Wiley & Sons. link ↗
How to cite this page
ScholarGate. (2026, June 3). Multiple Factor Analysis. ScholarGate. https://scholargate.app/en/psychometrics/multiple-factor-analysis
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Exploratory Structural Equation ModelingPsychometrics↔ compare
- Fuzzy ANOVAPsychometrics↔ compare
- Latent Transition AnalysisPsychometrics↔ compare
- Partial Least Squares Structural Equation ModelingPsychometrics↔ compare
- Redundancy AnalysisPsychometrics↔ compare