| Popis: |
The described RQ-mode factor-analysis program, written in Pascal, is based on a principal-component analysis of the correlation matrix (R-mode). It makes use of the Jacobi-threshold method to determine the eigenvalues and eigenvectors of the correlation matrix. Other types of factor analysis are implemented easily by modification or replacement of the standardization or similarity procedures. The Eckart—Young theorem, which unites the concepts of R- and Q-mode factor analysis by the theory of basic structure, is applied to calculate the Q-mode solution from the R-mode solution. The Kaiser-Varimax rotational scheme is used to optimize the R-mode solution. The Q-mode factor loadings and R-mode factor scores are written to two separate (disk) files for further use. The program, intended mainly for use on microcomputers, makes limited use of computer memory. Although the number of variables used in the analysis is limited by the size of the internal memory (in the test configuration of 64 kbyte 37 is the maximum), the number of samples is restricted only by the capacity of the background memory (e.g., the floppy or hard disk). The program was tested successfully on a hypothetical data set. The results also were checked with standard statistical computer packages. |
