Factor analysis is a strategy that relies on large sample sizes to produce accurate (stable) results. If sample sizes are very, very large, then the results will likely be accurate. The first major size limitation is that when the sample has less than ten scores per variable, computational problems arise. This forms a bare minimum sample size. For really accurate results, some researchers recommend upwards of 500 subjects. When determining the quality of factor analysis results, always look for large sample sizes.
Key Terms
latent construct, factor analysis, factor score, higher-order factor, eigenvalue, factor loading, rotation
Last Modified: 02/14/2019