Survey research, in which each question represents a measured variable, relies heavily on factor analysis techniques, including confirmatory factor analysis.
In contrast to exploratory factor analysis, which attempts to describe and identify the factors that influence a set of measures or responses, the confirmatory approach tests specified factors.
Confirmatory factor analysis can test the level of correlation among a set of factors, test the validity of a specific set of factors and compare how well different sets of factors fit the same set of data.
Because it tests whether a defined set of factors (known as a factor model) fit a set of data, confirmatory factor analysis is similar to other statistical procedures that test hypotheses.
A confirmatory analysis involves selecting the factors to test, obtaining correlations among the variables and assessing how well the factors fit the data.
The Chi-square statistic allows analysts to compare different factor models in confirmatory factor analysis.