Data acquired from both students and parents were factor analyzed to determine whether the items grouped into meaningful clusters or scales. Items that cluster together can be organized into scales for the purpose of contrasting different groups of respondents. The technique allows the user to gain greater understanding of diverse items on a questionnaire than is possible when items are treated separately and groups are compared on each item. The researchers used principal components with varimax rotation with listwise deletion for missing data. The results of the factor analyses will be presented first, followed by group comparisons on the different scales that emerged.
Factor Analyses
Student Questionnaire.
Two criteria were employed for the selection of factors. First, an eigenvalue of 1.00 or more was used to determine the number of factors that best fit the questionnaire data. A loading of .40 for retention of an item was the second criterion. The first analysis related to the "Integrative/Instrumental Motivation for Studying Asian Languages" scale, given in Part 1 of the student questionnaire. The results revealed that all eight items loaded onto a single factor (see Table 1). As can be noted in Table 1, the factor loadings of items were all over .50. The eigenvalue for the single factor was 2.97 and the Cronbach alpha was 0.83, which indicates that the scale was reliable for the student population. The percent of total variance explained by this single factor was 37.12 %. It is important to note that the result of the factor analysis in this study was not consistent with the original scale (Gardner 1985), which contained items for assessing two types of orientation toward language study, instrumental and integrative orientation. The findings suggest that the students in the present study made no distinction between instrumental and integrative motives for learning a foreign language.
Factor analysis of the 14 items relating to reasons for studying Asian languages in Part 2 of the student questionnaire revealed that a three-factor solution was best for these items. The eigenvalues for the three factors were 3.04, 1.72, and 1.02. Two items were dropped from the final analysis because they did not load at .40 or above on any factor. Those two items were "I want to use the language for practical situations such as ordering at a restaurant" and "I need to complete a foreign language requirement to get into college." Table 2 shows the results. Five items that loaded on Factor 1 constituted a subscale of motivation based on students' ethnic heritage. Factor 2 consisted of three items that are school related and pertained to what students had heard about the language class, teacher, or peers in the language class. Factor 3 was defined by four items that describe personal reasons for studying an Asian language. The Cronbach alphas were also respectable: .78 for Factor 1 ("Ethnic Heritage-Related Motivation"), .77 for Factor 2 ("School-Related Motivation"), and .69 for Factor 3 ("Personal Interests-Related Motivation"). The total variance explained by the three factors was 48.2%.
The results of the factor analysis on the "Parental Involvement in Foreign Language Study" scale (Part 3 of the student questionnaire) showed that all 9 items loaded on a single factor (see Table 3). The eigenvalue for this factor was 2.97 and the Cronbach alpha was .89. Each item's factor loading was above .50 and the percent of total variance explained by the single factor for students was 37.12%.
Parent Questionnaire.
Factor analysis for parents' responses to the questionnaire about attitudes toward foreign language learning (Part 1) revealed that the nine items all loaded on a single factor (see Table 4). The eigenvalue for this single factor was 3.54 and the Cronbach alpha was .86, which suggested that this scale was reliable. Each item's factor loading was above .40. The variance explained by this scale was 39.36%.
All nine items in the second part of the parent questionnaire regarding parental involvement in their child's foreign language study also loaded onto a single factor (see Table 3). The eigenvalue for the parent data was 4.45. The Cronbach alpha was .90. Each item's factor loading was above .50. The factor accounted for 49.42% of the total variance.
Student Group Comparisons
Japanese and Chinese Programs.
Students' responses to the various parts of the questionnaire were analyzed with respect to gender (males vs. females), grade level (elementary vs. high school), language program type (Japanese vs. Chinese), and ethnic heritage background. Analyses were first computed with only the Japanese and Chinese programs because there was no elementary Korean program in our study. For the "Instrumental/Integrative Motivation" scale, a statistically significant difference between male and female students in their motivation to learn the languages (F [1, 481] = 12.35, p < .0001) emerged. Female students (M = 4.65) had higher instrumental/integrative motivation to learn Asian languages than did male students (M = 4.46). This pattern was consistent for students in both elementary and high schools and in both Japanese and Chinese language classes (see Table 5).
The next set of analyses related to the three subscales of motives or reasons for learning these Asian languages. Regarding the first subscale, "Ethnic Heritage-Related Motivation," we detected significant differences by grade level and by language program type. Elementary school students (M = 3.54) had significantly higher scores on the heritage variable than did high school students (M = 1.74, F [1, 468] = 84.15, p < .0001). This suggests that elementary students had enrolled in these Asian language programs because of heritage-related factors. Extending this interpretation, we also noted that students in the Chinese programs (M = 3.04) had higher motivation scores than did students in the Japanese programs (M = 2.00). The difference was statistically significant (F [1, 468] = 5.06, p = 0.03), suggesting that students enrolled in Chinese language classes were more likely to have enrolled for reasons of ethnic heritage than were students enrolled in Japanese language classes. This finding was further supported by the percentage of ethnic heritage students in the programs. Whereas 68.9% of elementary school students were ethnically related to the language that they were studying, this was true of only 30.8% of students in the high school programs. On the other hand, 61.2 % of the students in the Chinese language programs reported being of Chinese ethnic descent, whereas only 26.8 % of the students in the Japanese language programs were of Japanese heritage. Table 6 presents the mean scores of the heritage motivation scale by students' ethnic heritage background in each language program. Analysis of variance showed that in both the Chinese and Japanese programs, students who were ethnically related to the language that they were studying had a significantly higher score on the "Ethnic Heritage-Related Motivation" scale than did students who were of a different ethnic heritage (F [2, 108] = 41.69, p < .0001) for the Chinese programs and (F [2, 366] = 175.00, p < .0001) for the Japanese programs.
For the second subscale, "School-Related Motivation" for taking an Asian language, a significant effect of grade level was found (F [1, 474] = 9.52, p = .002). Elementary school students expressed higher school-related motivation (M = 2.98) than did their high school counterparts (M = 2.26). A more detailed examination revealed a significant interaction for grade level by language program type (F [1, 474] = 5.92, p = .02). The differences in motivation between the elementary and high school levels was due to the responses of students in the Japanese language programs. Scheffˇ tests indicated that elementary school students of Japanese (M = 3.29) had significantly higher motivation scores than their high school student counterparts (M = 2.23, p < .0001). In the Chinese programs, however, there was no significant difference in motivation between elementary and high school respondents. This finding suggests that elementary school students in this study probably enrolled in Japanese classes because their parents knew that the school had a good Japanese program or good Japanese teachers.
On the third subscale, "Personal Interests-Related Motivation," there was a significant gender difference (F [1, 471] = 11.18, p < .0001). Female students (M = 4.18) scored higher on the scale indicating personal interests-related motivation for learning an Asian language than male students (M = 3.90). This measure also revealed a significant difference for language program type (F [1, 471] = 7.87, p = .005). The Japanese program students (M = 4.12) had higher scores on the "Personal Interest-Related Motivation" scale than students in the Chinese programs (M = 3.72).
In the questionnaire portion entitled "Perception of Parental Involvement," students were also asked to reflect upon their parents' level of involvement in their foreign language study. We found a significant difference in students' perceptions of their parents' involvement by gender, language program type, and grade level. As depicted in Figure 1, female students' perceptions of their parents' involvement (M = 3.52) in both the Japanese and Chinese language programs was significantly higher than that of male students (M = 3.14, F [1, 480] = 8.26, p = .004). Moreover, students in the Japanese programs (M = 3.51) and in the elementary schools (M = 3.84) perceived significantly higher parental involvement than students in the Chinese programs (M = 2.68) or in the high schools (M = 3.10, F [1, 480] = 72.66, p < .0001 and F [1, 480] = 59.02, p < .0001, respectively).
Home | Organization | Map | Events and Discussions | Resources | Standards and Frameworks
