Concept Architecture
Concept
Theoretically, Factor Rotation is a mathematical transformation applied in factor analysis to improve the interpretability of the extracted factor solution while preserving its statistical fit. It is founded on linear algebra and matrix transformation theory. The concept exists because an initial factor solution often distributes loadings across multiple factors, making substantive interpretation difficult. Rotation seeks a simpler factor structure in which variables load strongly on relatively few factors.
Mathematically, factor rotation transforms the original loading matrix through an orthogonal or oblique transformation. Orthogonal rotations preserve factor independence, whereas oblique rotations allow factors to correlate. The transformation leaves the reproduced covariance matrix unchanged but redistributes factor loadings to produce a more interpretable solution according to a specified optimisation criterion, such as Varimax, Quartimax, Equamax, Oblimin or Promax.
In practice, factor rotation is performed after factor extraction during exploratory factor analysis. The rotated loading matrix is examined to identify variables associated with each latent construct, assess communalities and interpret the underlying factor structure. In health economics, rotated factor solutions are widely used during the development and validation of patient-reported outcome measures, health-related quality-of-life instruments and other latent construct questionnaires.
Purpose
Used to improve the interpretability of factor analysis by transforming factor loadings into a simpler structure while preserving the explanatory information contained within the extracted factors.
Mathematical Formulae
Primary Formula
Rotated loading matrix:
L* = LT
where:
- L = original loading matrix
- T = rotation matrix
- L* = rotated loading matrix
For orthogonal rotation:
T?T = I
Supporting Formulae
Reproduced covariance matrix:
� = LL? + ?
After rotation:
� = LL? + ?
Varimax optimisation criterion:
V = ?? [(1/p)?? ???? ? {(1/p)?? ???�}�]
where:
- p = number of variables
- ??? = rotated factor loading
Related Mathematical Methods
- Exploratory Factor Analysis
- Confirmatory Factor Analysis
- Varimax Rotation
- Quartimax Rotation
- Equamax Rotation
- Oblimin Rotation
- Promax Rotation
- Factor Loading
- Eigenvalue Analysis
Example
An exploratory factor analysis of a health-related quality-of-life questionnaire initially produces two factors with diffuse loadings.
Before rotation:
| Variable | Factor 1 | Factor 2 |
|---|---|---|
| Mobility | 0.58 | 0.46 |
| Pain | 0.61 | 0.42 |
| Anxiety | 0.44 | 0.56 |
Following Varimax rotation:
| Variable | Factor 1 | Factor 2 |
|---|---|---|
| Mobility | 0.83 | 0.12 |
| Pain | 0.81 | 0.18 |
| Anxiety | 0.16 | 0.79 |
The rotated solution provides a clearer interpretation, with Factor 1 representing physical health and Factor 2 representing psychological health.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| MMULT | =MMULT(B2:C10,F2:G3) | Multiplies the loading matrix by the rotation matrix to obtain rotated loadings. |
| TRANSPOSE | =TRANSPOSE(F2:G3) | Computes the transpose of the rotation matrix. |
| MUNIT | =MUNIT(2) | Verifies orthogonal rotation against the identity matrix. |
| SUMSQ | =SUMSQ(B2:C2) | Confirms communality is preserved after rotation. |
VBA (Optional)
Automate orthogonal and oblique factor rotations, producing rotated loading matrices, communalities and factor interpretation reports.
Sources
- Harman HH. Modern Factor Analysis.
- Kaiser HF. The Varimax Criterion for Analytic Rotation in Factor Analysis. Psychometrika. 1958.
- Fabrigar LR, Wegener DT. Exploratory Factor Analysis.
- Hair JF, Black WC, Babin BJ, Anderson RE. Multivariate Data Analysis.
- Brown TA. Confirmatory Factor Analysis for Applied Research.
Related Concepts (3)
Library
Publications
1
Bayesian Methods in Health Economics — Gianluca Baio, 1st Edition ed., 2012 (Chapman & Hall / CRC Press)
An overview of Bayesian statistical methods for the analysis of health economic data, covering economic evaluation concepts, statistical cost-effectiveness analysis, Bayesian computation and MCMC, and applied health economic evaluation.
BookView source →
Frequently Asked Questions (6)
What is factor rotation?
A technique transforming an initial factor solution into a more interpretable structure, typically by maximising the contrast between high and low loadings.
Source: Kaiser 1958
What does factor rotation do to make a solution interpretable?
Factor rotation reworks an initial factor solution into a form that is easier to interpret, without changing how much variance the factors explain in total. It does this by rearranging the loadings so that each variable loads strongly on one factor and weakly on the others, sharpening the contrast that lets each factor be read as a distinct dimension. Orthogonal rotation keeps the factors uncorrelated, while oblique rotation lets them correlate. Clarifying which items belong to which factor is its purpose. Kline (2015) describes this technique.
Source: Kline 2015
Why is factor rotation used?
Factor rotation is used because the initial factor solution, while explaining the correlations, often has loadings spread across factors in a way that is hard to interpret, and rotation seeks a simpler structure in which each variable loads strongly on one factor and weakly on others. This clarifies what each factor represents. So factor rotation is used to improve interpretability, transforming the factors into a configuration where the pattern of loadings is clearer, without altering the overall variance explained, which is why rotation is a routine step in factor analysis, turning a mathematically valid but opaque solution into one that can be meaningfully interpreted.
Source: Kaiser 1958
What are orthogonal and oblique rotation?
Orthogonal rotation keeps the factors uncorrelated with one another, as in the varimax method, while oblique rotation allows the factors to correlate, which is appropriate when the underlying constructs are expected to be related. Orthogonal rotation gives simpler, independent factors, and oblique rotation more realistic ones when factors are correlated. So orthogonal and oblique rotation differ in whether the rotated factors are allowed to correlate, with the choice depending on whether the constructs are thought to be independent or related, and oblique rotation is often preferred in practice because many real constructs are correlated, though it complicates interpretation slightly.
Source: Kaiser 1958
How does factor rotation aid interpretation?
Factor rotation aids interpretation by producing a simple structure, in which each variable loads highly on one factor and near zero on the others, so that the factors are more clearly defined by distinct sets of variables. This makes it easier to identify what each factor represents. So factor rotation aids interpretation by sharpening the pattern of loadings toward simple structure, which reduces ambiguity about which variables belong to which factor and helps the analyst name and understand the factors, turning a diffuse initial solution into one where the factors correspond to interpretable groups of variables.
Source: Kaiser 1958
Does factor rotation change the variance explained?
Factor rotation does not change the total variance explained by the retained factors; it redistributes the explained variance among them to achieve a clearer structure, so the overall fit and the communalities of the variables remain the same while the loadings are rearranged. So factor rotation preserves the total variance accounted for by the factors, altering only how that variance is apportioned across them to improve interpretability, which is why rotation can be applied freely to seek a clearer solution without sacrificing the amount of variation the factors collectively explain, changing the appearance of the solution but not its explanatory power.
Source: Kaiser 1958
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 16 Dec 2025
Content version: 1.0.0
Canonical Identity
- Persistent URI
- https://healtheconomics.wiki/concept/factor-rotation
- Term code
- HE-ES-SA-063
Stable URI · Machine-readable · Resolvable · CC BY 4.0