Suter (1976) studied the correlations between English pronunciation accuracy scores and a battery of 20 variables for 61 nonnative speakers of English. Although he found that 1 2 of the 20 were significantly correlated with pronunciation accuracy, he did not attempt to assess the relative importance of the 1 2 in accounting for the criterion's variance. By limiting his analysis to a consideration of zero order correlations it was not possible to determine which of the significant predictors were true predictors of pronunciation accuracy in and of themselves and which were only significant predictors by virtue of their substantial correlations with other, more efficient predictors of pronunciation accuracy. In the present paper we apply more advanced techniques to the analysis of the predictors and conclude that only 4 variables, one of them a composite of 2 of Suter's predictors, are useful in accounting for the variability of subjects' pronunciation accuracy scores. Language LearningVol. 30, No. 2 better determine which of the many predictor variables he examined are the substantive predictors of pronunciation accuracy and which of his predictors do not really contribute to the explanation of pronunciation accuracy. Since Suter's study (1976) comprised the most tightly controlled and comprehensive analysis to date of such a wide range of variables felt to influence learner's pronunciation accuracy, we felt the data worthy of reanalysis with more powerful multivariate techniques. Since these techniques are not widely known in the field of adult second language research, the reanalysis of Suter's data might serve as a good example of the usefulness of such more advanced techniques in the study of adult second language learning.Analysis Suter (1976) presented the details of the data acquisition process, which will not be repeated here. We will, however, briefly recap the list of variables he considered. Suter's dependent variable, or criterion, consisted of the average rating by fourteen judges of sixty-one nonnative speakers of English. The judges rated the subjects on a subjective scale of English pronunciation accuracy. Judges' ratings of the subjects were highly correlated. The predictor variables included: 6. Number of months of residency with native speakers of English (MRNS) 7. Number of years of formal classroom training in English (YFCT) 8. Number of months of intensive formal classroom training in English (MICT) 9. Number of weeks of formal classroom training focused specifically on English pronunciation (WFTP) 10. Proportion of a subject's teachers who were themselves native speakers of English (PTNS) 11. The subject's native language ( L l ) 12. The number of languages the subject can converse in (NLng) 13. The sex of the subject (Sex) 14. The subject's economic motivation (EcMo) 15. The subject's social prestige motivation (SPMo) Learning Vol. 30, No. 2 Language

Previous studies have sought to establish the significance of various parameters in the determination of the patternings of Russian formant frequency trajectories in vowel-consonant-vowell (VCV) syllables. In the present study, 9600 measurements of first and second formant frequency were made on Russian VCV nonsense words. Variance for speakers was controlled through regression. Stepwise multiple regression was employed to determine the relative contributions of six predictors to the explanation of the patternings of first and second formant frequency in the test words. The six predictors included the height of the first and second vowels, the fron/back dimensionality of the first and second vowels, the place of articulation of the consonant, and the palatalization or nonpalatalization of the consonant. Results of the regression analyses are summarized. The regression coefficients for the useful predictors of the various dependent variables comprise a model of formant frequency patterns in Russian VCV utterances.

No abstract