PCA & dimension reduction

What underlying dimensions sit behind this battery?

Run principal component analysis on numeric and opinion-scale columns to explore latent structure and reduce dimensionality beyond one-at-a-time crosstabs and banners. AddMaple extracts components that capture the greatest variance across the battery.

See which questions belong together.

Review component loadings to interpret what each dimension represents and identify the questions that cluster together. Optional Varimax rotation can produce a simpler, more interpretable loading structure while keeping rotated components orthogonal.

Carry the structure into the wider study.

Use the components as inputs for segmentation or as a clearer way to tell the research story. The result stays beside the banners and charts it helps explain, rather than becoming an export from a separate factor-analysis workflow.

Use PCA as a way to organise evidence.

PCA does not decide what a construct means or prove that a dimension causes an outcome. It gives the researcher a compact view of co-movement in a battery, which can then be checked against the questionnaire wording, loadings, variance explained, and the purpose of the study before being used in a segment or narrative.