Definition
Principal component analysis (PCA) condenses many correlated variables into a few new uncorrelated components that capture as much of the original variability as possible. The first component explains the most variance, the second the most of what remains, and so on. It serves dimensionality reduction, visualization of multivariate data and removal of multicollinearity.
Multivariate methods
Choose the number of components by explained variance (e.g. 80 %), the Kaiser criterion (eigenvalue > 1) or the scree plot. Loadings tell which original variables make up a component and let you name it — say "size" and "shape" for product dimensions.
Standardize the data before PCA, otherwise variables with large units dominate. PCA is a descriptive method without distributional assumptions; component scores can be reused in regression, clustering or control charts for multivariate SPC.
In Statistica
Statistics → Multivariate Exploratory Techniques → Principal Components & Classification Analysis returns eigenvalues, the scree plot, loadings, scores and a biplot; PCA with classification can display supplementary variables and cases. For industrial data Statistica also offers PCA-based multivariate SPC (Hotelling T²).
Related terms
- Factor analysisFactor analysis looks for a small number of hidden (latent) factors that explain the correlations among many measured…
- Cluster analysisCluster analysis divides observations into groups (clusters) so that units within a cluster are as similar as possible…
- CorrelationCorrelation measures the strength and direction of a linear relationship between two quantities with a number from −1 to 1.
- Linear regressionLinear regression describes how a dependent variable (yield, price, consumption) depends on one or more explanatory…
Knowledgebase guides
FAQ
- How many components should I keep?
- As many as explain a reasonable share of variance and can be interpreted — usually 2–5. The scree plot shows the elbow beyond which components add only noise.
- How does PCA differ from factor analysis?
- PCA is a mathematical transformation maximizing variance; factor analysis models hidden causes (factors) and separates common variance from measurement error.
Try it on your own data
Statistica free for 30 days
Full version, no credit card. Or get a pricing estimate in a minute.
Multivariate methods
Updated: September 2026.