RESULTS
Reliabilities of the Instruments
In the pilot study, the reliability test based on 74 valid subject data revealed that each subscale in the motivation and the metacognitive vocabulary strategy questionnaires reached internal consistency (Cronbach’s alpha) over .70. Table 4 demonstrates values of Cronbach’s alpha for each subscale in the questionnaires, indicating that the instruments were reliable. No modification was made on the questionnairres.
Table 4
Reliability Results in the Pilot Study
Similarly, in the formal study, the reliability test results based on the responses of 224 participants suggested that each subscale in the motivation and the metacognitive vocabulary strategy questionnaires reached satisfactory internal consistency. All the Cronbach’s alpha values were above .70, establishing a reliable ground for the
Subscale Numbers of items Cronbach's alpha
Motivational intensity 11 0.863
Intrinsic motivation 11 0.863
Extrinsic motivation 15 0.886
Amotivation 3 0.847
Requirement motivation 5 0.937
Selective attention 7 0.833
Self-initiation 5 0.885
Consolidation 5 0.736
57
following analyses and interpretations. Table 5 shows the alpha value of each subscale in these questionnaires in the phase of formal study.
Table 5
Reliability Results in the Formal Study
Subscale Numbers of items Cronbach's alpha
Motivational intensity 11 0.882
Intrinsic motivation 11 0.935
Extrinsic motivation 15 0.905
Amotivation 3 0.780
Requirement motivation 5 0.943
Selective attention 7 0.717
Self-initiation 5 0.885
Consolidation 5 0.749
Participants’ Overall Performance on Instruments
In the formal study, the overall English learning motivation (including intensity and types), use of metacognitive vocabulary learning strategies, and vocabulary achievement of the 224 participants were revealed through descriptive statistics (see Table 6).
Given the 4-point scale and a total of 11 items concerning motivational intensity in the motivational questionnaire, the sum threshold scores ranging from strongly disagree to strongly agree were 11, 22, 33, and 44. Since the calculated mean of all the participants’ motivational intensity was 25.90, the students’ strength of motivation to learn English roughly fell on the median 27.50, with a slight skew to the disagree end. With regard to motivational types, the sum threshold scores for intrinsic motivation were 11, 22, 33, and 44; for extrinsic motivation, they were 15, 30, 45, and 60; amotivation, 3, 6, 9, and 12; requirement motivation, 5, 10, 15, and 20.
58
Respectively, the mean of intrinsic motivation was 26.9, indicating the participants held almost neutral intrinsic motivation toward English learning. The mean of extrinsic motivation, 40.35, showed that the participants were almost strongly moved by external forces to learn English. In contrast, they were hardly driven by amotivation since the mean was 4.84, which meant that on average, this sample of students knew well why they were learning English. In terms of requirement motivation, the mean 10.76 revealed that the degree of the participants’ reliance on school assignments to learn English was close to average.
Table 6
Means on the Instruments by All the Participants
Categories Means
Motivational intensity 25.90
Intrinsic motivation 26.90
Extrinsic motivation 40.35
Amotivation 4.84
Requirement motivation 10.76
Metacognitive vocabulary learning strategy 43.09
Vocabulary achievement 40.27
Likewise, given the 4-point scale and a total of 17 items on metacognitive vocabulary learning strategy questionnaire, the sum threshold scores were 17, 34, 51, and 68, specifying strongly disagree, disagree, agree, and strongly agree. The mean of all the participants’ strategy use, 43.09, indicated they used these metacognitive strategies at average frequency.
In terms of the participants’ performance on the 60-item vocabulary test, the mean of 40.27 was higher than the passing score, 36, given the highest score of 60
59
and 1 point for each correct answer. This result suggests that the difficulty level of the vocabulary test might be appropriate for this sample of students.
Effects of SES on Motivation, Metacognitive Strategy Use, and Vocabulary Acquisition
In the formal study, 224 participants were drawn from 6 intact first-grade classes of different departments. They were grouped into high, mid, and low socio-economic status following Hallinghead’s (1957) procedure before further statistical analysis on the effects of SES was made. There were 53 high-SES, 88 mid-SES, and 83 low-SES subjects. Respectively, they accounted for 24%, 39%, and 37% of all participants.
Table 7 demonstrates the distribution of the participants by socio-economic status.
Table 7
Distribution of Participants Based on SES
Group Numbers of participants Percentage
High SES 53 24%
Mid SES 88 39%
Low SES 83 37%
The three groups’ responses to each dimension in the motivational questionnaire – motivational intensity, intrinsic motivation, extrinsic motivation, amotivation, and requirement motivation – are described in Table 8. The descriptive statistics showed that there seemed to be a positive correlation between motivational intensity and SES status, with students of higher SES demonstrating higher learning motivation. Specifically, the means of low-SES, mid-SES, and high-SES groups on motivational intensity were 25.54, 26.05, and 26.21 respectively. In extrinsic motivation, the similar tendency was also found. The respective means of the three SES groups concerning extrinsic motivation were 39.33, 40.95, and 40.96.
60
Table 8
Descriptive Statistics of the Motivation Questionnaire by SES
Number Mean SD
61
The participants’ responses to the metacognitive vocabulary strategy questionnaire and their performance on the vocabulary test are demonstrated in Table 9. The descriptive statistics showed that in metacognitive vocabulary strategy, the means of low-SES, mid-SES, and high-SES groups were 42.21, 43.98, 42.96 respectively. As for vocabulary achievement, the means were 40.14, 40.50, 40.08, without a conceivable tendency.
Table 9
Descriptive Statistics of the Metacognitive Strategy Questionnaire and the Vocabulary Test by SES
62
To answer the first research question – whether students of different SES differ significantly in English learning motivation (including motivational intensity and motivation types), use of metacognitive vocabulary learning strategies, and vocabulary learning achievement – MANOVA and one-way ANOVA were performed.
To be specific, connection between SES and motivational intensity and four motivation types was examined through MANOVA, while SES’ association with strategy and vocabulary test was respectively measured by one-way ANOVA.
The results showed that students of different socio-economic background did not possess significantly different level of motivational intensity or different types of motivation (F = 0.691, p =.733). In other words, those from higher SES were not significantly more motivated to learn English than lower-SES peers. Neither were they characterized by different types of learning motivation toward English, compared with lower-SES counterparts. Table 10 presents the MANOVA result, showing the insignificant connection between SES and motivation with respect to both intensity and types.
Table 10
MANOVA Results
Effect Value F Hypothesis df Error df p
SES group Wilks' Lambda 0.967 0.691 10.000 408.000 0.733
Likewise, results from one-way ANOVA (Table 11) demonstrated that students of different SES did not use metacognitive vocabulary strategies at significantly different frequency (F = .88, p = .416). That is, SES was not associated with the frequency students exert their metacognition in the process of English lexical learning.
63
Table 11
One-way ANOVA of Metacognitive Vocabulary Strategy by SES Groups
SS df MS F p
Metacognitive vocabulary
strategy
Between groups 130.916 2 65.458 0.88 0.416 Within groups 16145.266 217 74.402
Total 16276.182 219
Neither did SES have significant effect on learners’ vocabulary achievement.
Table 12 demonstrates that the three SES groups did not perform differently on the vocabulary test (F = .068, p = .934). Again, higher-SES students did not necessarily possess wider lexis than lower-SES students did.
Table 12
One-way ANOVA of Vocabulary Achievement by SES Groups
SS df MS F p
vocabulary achievement
Between groups 7.965 2 3.983 0.068 0.934
Within groups 12917.963 221 58.452 Total 12925.929 223
In sum, the results revealed that the relation between SES and motivation (including intensity and types), metacognitive vocabulary learning strategy, and vocabulary achievement was not evident. Students of different socio-economic status did not significantly differ in learning motivation, use of metacognitive vocabulary strategy, and vocabulary achievement.
Explanatory Power of SES, Motivation, and Metacognitive Strategy in Vocabulary Acquisition
Hierarchical regression was executed in order to answer the second research question: What is the explanatory power of learning motivation and metacognitive
64
vocabulary learning strategy for learners’ vocabulary achievement, with SES controlled? The finding indicated that SES’ sole accountability for vocabulary achievement was nearly null (R2 = .000; p = .989), whereas the inclusion of motivation and strategy into the analysis led to a significant increase in accountability of vocabulary achievement (R2 = .288, R2 Change = 28.8%, p = .000). In other words, motivation (including intensity and types) and strategy explained up to 28.8% of learners’ vocabulary achievement, which was a significant change since SES itself was incapable of predicting vocabulary achievement. Table 13 reveals the change in accountability of vocabulary achievement before (Model 1) and after (Model 2) the inclusion of motivation and strategy as independent variables.
Table 13
Model Summary of Hierarchical Regression Analysis
Model R R Square
Change Statistics R Square
Change
F Change df1 df2 p
1 .001 0.000 0.000 0.000 1 205 0.989
2 .536 0.288 0.288 13.396 6 199 0.000
Despite the significance of motivation and strategy together in explaining vocabulary achievement, a closer inspection of the standardized regression coefficients (Beta) unveiled that among all the predictors – motivational intensity, intrinsic motivation, extrinsic motivation, amotivation, requirement motivation, and metacognitive vocabulary learning strategy – only motivational intensity exerted a significant effect on vocabulary achievement, while the other variables did not.
To be specific, as Model 1 in the Table 14 reveals, SES was hardly associated with vocabulary achievement (Beta = -0.001, p = 0.989), which was consistent with the results obtained for the first research question. Nevertheless, when motivation and
65
strategy were also taken into account in Model 2, motivational intensity outperformed the other variables and was found to be the only variable that significantly explained learners’ lexical learning (Beta = .309, p = 0.012). Moreover, the positive coefficient value indicated that when learners’ SES was controlled, the higher the students’
motivational intensity was, the better they performed on the vocabulary test.
Table 14
In conclusion, SES barely had any relation with vocabulary achievement. Since SES was not related to vocabulary learning, controlling SES did not make any difference in measuring the accountability of motivation and strategy for vocabulary learning. More importantly, the results from hierarchical regression revealed that with SES controlled, the sole and only variable that significantly explained vocabulary
66
achievement was motivational intensity. Motivational types and metacognitive strategy use did not play a significant role in explaining vocabulary acquisition.
67