First Advisor

Kimberley Brown

Date of Publication


Document Type


Degree Name

Master of Arts (M.A.) in Teaching English to Speakers of Other Languages


Teaching English as a Second Language


Oral communication, Anxiety, English language -- Study and teaching (Higher) -- Japanese speakers, Japanese students -- Education (Higher) -- United States



Physical Description

1 online resource (vii, 176 p.)


This study attempted to locate some possible sources of oral Foreign Language Anxiety (FLA) among Japanese students in the United States. This study proposed that the following three factors were possible sources of FLA: 1) the subjects' traitlike anxiety, which is carried by individuals across all communication-bound contexts; 2) the subjects' self-perceived oral proficiency levels in English and 3) the subjects' gapsize (i.e., the distance between their self-perceived and their self-expected oral proficiency levels in English) . This research examined whether the above three independent variables and the dependent FLA variable were significantly correlated, and if so, which one had the strongest correlation with the FLA variable. Also, whether the subjects' biographical variables had a significant effect on their FLA levels was investigated. All the variables were quantified through a questionnaire. The subjects' FLA levels and traitlike anxiety levels were measured by a 10-item, Personal Report of Communication Apprehension inventory (PRCA, Mccroskey, 1978). The subjects' self-perceived oral proficiency levels were measured by asking the subjects to rate their self-perceived oral proficiency level from 1 (poor) to 5 (fluent). The gapsize was quantified by asking the subjects to rate it on a scale from 1 (minimal) to 5 (maximal). The statistical methodology used in obtaining the PRCA scores in this study differed from McCroskey's in its interpretation of Likert type scales. The scales were treated as interval data in McCroskey's study, while, in this study, they were interpreted as ordinal data. After hierarchically ordering the subjects' answers, non-parametric tests were performed on them. Overall, each of the three variables and the FLA variable were found to be significantly correlated at p < .01. The traitlike anxiety variable, the proficiency variable and the gapsize variable correlated at .46, -.45 and -.33, respectively. The participants' demographic variables (age, gender, status at school or year(s) of residence in English speaking places) did not have a significant effect on their FLA levels. A discussion of the results was provided, with references to previous studies.


If you are the rightful copyright holder of this dissertation or thesis and wish to have it removed from the Open Access Collection, please submit a request to and include clear identification of the work, preferably with URL

Persistent Identifier