Factor analysis

Factor analysis looks for groups of questions that are answered similarly. Each group probably measures something in common. The loading (between −1 and 1) shows how strongly a question belongs to its group; from about .30 it counts as belonging. The default is principal component analysis as in SPSS; principal axis factoring with promax looks for common traits that may be correlated.

Example: Ten questions about learning form three groups: planning, motivation, distraction.

Caution: What a group is called is decided by people, based on the content. A negative loading means the question runs opposite to the others.