Application of an Alternative Method of Canfirmatory Factor Analysis (Based on the Matherials of Phonemic Speech Perception Research of American Schoolchildren)

The research of phonemic speech perception of American school children revealed shortcomings of traditional confirmatory factor analysis. An alternative version of this method was developed. The new version of confirmatory factor analysis suggests examination of factor model free (spare) parameters interrelation. It allows finding parameters directly providing for single optimal solution. The essence of this method is to build redefined equations set, which is resolved directly, and testing adequacy of obtained set with statistical fit indices. The adequacy of obtained model is determined by chi-square statistic, which allows checking the hypothesis about presentability of sampling variances and co-variances of latent variables that constitute the right-hand part of the overdetermined equations set, and by the components of variances and co-variances of latent variables contained in the analyzing model. The developed alternative method of confirmative factor analysis also allows assessing statistical value of model's components. The results of examination of free parameters by alternative version of factor analysis didn't show significant difference from results of traditional factor analysis. These findings are encouraging and future applications of the new approach are discussed.

1. Korolyuk V. S., Portenko N. I., Skorohod A. V., Turbin A. F. Spravochnik po teorii veroyatnostei i matematicheskoi statistike. M., 1985.
2. Kuravskii L. S., Kornienko P. A. Primenenie neironnyh setei dlya identifikacii lokusov kolichestvennyh priznakov v psihogenetike // Neirokomp'yutery: razrabotka i primenenie. 2007. № 4.
3. Louli D., Maksvell A. Faktornyi analiz kak statisticheskii metod. M., 1967.
4. Bishop Y. M. M., Fienberg S. E., Holland P. W. Discrete multivariate analysis: Theory and practice. Cambridge, 1975. [Link to the new edition]
5. Bollen K. A. Structural equations with latent variables. N. Y., 1989.
6. Branum-Martin L., Carlson C. D., Carlo M., Fletcher J. M., Francis D. J., Mehta P. D., Ortiz A. Bilingual phonological awareness: Multilevel construct validation among Spanish-speaking kindergarteners in transitional bilingual education classrooms // Journal of Educational Psychology. 2006. Vol. 98. № 1.
7. Goldstein H. Multilevel statistical models (3rd ed.). L., 2003.
8. Loehlin J. C. Latent variable models: An introduction to factor, path, and structural analysis. Hillsdale, N. J., 1987.
9. Neale M. C., Cardon L. R. Methodology for genetic studies of twins and families. Dordrecht, the Netherlands: Kluwer Academic Publishers, 1992.
10. Schatschneider C., Francis D. J., Foorman B. R., Fletcher J. M., Mehta P. The dimensionality of phonological awareness: An application of item response theory // Journal of Educational Psychology. 1999. Vol. 91.