1) How do teachers create structure for engagement in foreign language classes?
7.3 Results
Table 7.1. Descriptive statistics for each separate condition.
ALT Absent No JTE HRT Leads (n=199)
ALT Present No JTE HRT support (n=199)
ALT Absent JTE Leads HRT support (n=137)
ALT Present JTE Support HRT support (n=137) Length of
explanation
Mean= 2.88 / SD= .78 Mean= 2.74 / SD= .77 Mean= 2.88 / SD= .81 Mean= 2.78 / SD= .72 Length of
explanation
Skew = -.39 Skew = -.26 Skew = -.52 Skew = -.46
Length of
explanation Kurtosis = 2.91 Kurtosis = 2.78 Kurtosis = 2.95 Kurtosis = 3.23 Length of
explanation
95% CI= 2.77 / 2.99 95% CI= 2.63 / 2.85 95% CI= 2.75 / 3.02 95% CI= 2.66 / 2.90 Clarity of
explanation
Mean= 3.56 / SD= .69 Mean= 3.40/SD= .78 Mean= 3.31/SD= .86 Mean= 3.31 / SD= .87 Clarity of
explanation
Skew =-1.24 Skew = -1.14 Skew = -1.14 Skew = -.97
Clarity of
explanation Kurtosis = 3.84 Kurtosis = 3.56 Kurtosis = 3.58 Kurtosis = 2.87 Clarity of
explanation
95% CI= 3.47 / 3.66 95% CI= 3.29 / 3.51 95% CI= 3.17 / 3.46 95% CI= 3.16 / 3.45 Pacing
Mean= 3.30 / SD= .80 Mean= 3.24 / SD= .81 Mean= 3.31 / SD= .80 Mean= 3.30 / SD= .69
Pacing Skew = -.94 Skew = -.88 Skew = -1.04 Skew = -.87
Pacing
Kurtosis = 3.43 Kurtosis = 3.20 Kurtosis = 3.57 Kurtosis = 4.04 Pacing
95% CI= 3.18 / 3.41 95% CI= 3.13 / 3.36 95% CI= 3.17 / 3.44 95% CI= 3.18 / 3.42 Classmates
engagement
Mean= 3.41 / SD= .74 Mean= 3.39 / SD= .80 Mean= 3.31 / SD= .81 Mean= 3.30 / SD= .77 Classmates
engagement
Skew = -1.10 Skew = -1.23 Skew = -1.05 Skew = -.95
Classmates
engagement Kurtosis = 3.67 Kurtosis = 3.94 Kurtosis = 3.31 Kurtosis = 3.52 Classmates
engagement
95% CI= 3.30 / 3.51 95% CI= 3.28 / 3.50 95% CI= 3.17 / 3.44 95% CI= 3.17 / 3.43 Emotional
support
Mean= 3.60 / SD= .65 Mean= 3.53 / SD= .71 Mean= 3.51 / SD= .78 Mean= 3.60 / SD=.69 Emotional
support
Skew = -1.77 Skew = -1.61 Skew = -1.73 Skew = -1.97 Emotional
support Kurtosis = 6.16 Kurtosis = 5.45 Kurtosis = 5.62 Kurtosis = 7.00 Emotional
support
95% CI= 3.51 / 3.69 95% CI= 3.43 / 3.63 95% CI= 3.38 / 3.64 95% CI= 3.48 / 3.71 Clear lesson
goals
Mean= 3.34 / SD= .82 Mean= 3.29 / SD= .75 Mean= 3.35 / SD= .76 Mean= 3.29 / SD=.74 Clear lesson
goals
Skew = -.98 Skew = -.88 Skew = -1.18 Skew = -.74
Clear lesson
goals Kurtosis = 3.52 Kurtosis = 3.49 Kurtosis = 4.27 Kurtosis = 2.93 Clear lesson
goals
95% CI= 3.23 / 3.46 95% CI= 3.18 / 3.39 95% CI= 3.22 / 3.48 95% CI= 3.17 / 3.42 Teachers’
affect when speaking Eng-lish
Mean= 3.62 / SD= .65 Mean= 3.68 / SD= .62 Mean= 3.46 / SD= .78 Mean= 3.59 / SD=.70 Teachers’
affect when speaking Eng-lish
Skew = -1.74 Skew = -2.16 Skew = -1.28 Skew = -1.80
Teachers’
affect when speaking
Eng-lish Kurtosis = 5.61 Kurtosis = 7.84 Kurtosis = 3.80 Kurtosis = 6.04 Teachers’
affect when speaking
Eng-lish 95% CI= 3.53 / 3.71 95% CI= 3.60 / 3.77 95% CI= 3.33 / 3.59 95% CI= 3.47 / 3.71 Teachers’
amount of English out-put
Mean= 2.97 / SD= .83 Mean= 3.50 / SD= .73 Mean= 3.24 / SD= .90 Mean= 3.62 / SD=.68 Teachers’
amount of English out-put
Skew = -1.04 Skew = -1.47 Skew = -1.10 Skew = -1.8377
Teachers’
amount of English
out-put Kurtosis = 3.32 Kurtosis = 4.83 Kurtosis = 3.39 Kurtosis = 5.86 Teachers’
amount of English
out-put 95% CI= 2.85 / 3.09 95% CI= 3.40 / 3.60 95% CI= 3.09 / 3.39 95% CI= 3.51 / 3.74 Spoken
Eng-lish output by Students
Mean= 2.78 / SD= .89 Mean= 2.88 / SD= .85 Mean= 2.92 / SD= .85 Mean= 2.80 / SD= .99 Spoken
Eng-lish output by Students
Skew = -.23 Skew = -.35 Skew = -.50 Skew = -.39
Spoken Eng-lish output by
Students Kurtosis = 2.26 Kurtosis = 2.38 Kurtosis = 2.74 Kurtosis = 2.12 Spoken
Eng-lish output by Students
95% CI= 2.65 / 2.90 95% CI= 2.74 / 3.02 95% CI= 2.80 / 3.04 95% CI= 2.63 / 2.96
At the univariate level, within-subjects repeated-measures factorial ANOVA tests found significant differences in effects on the condition of the frequency of teachers’ output for both ALTs, F(1, 671) = 59.63, p < .00, partial η2 = .15, and JTEs, F(1, 671) = 9.72, p < . 01, partial η2 = .03, but no significant interaction effects between the two. The R2 was calculated at .58, indicating that this model accounts for roughly 58% of the variance. Further
ANOVAs revealed between-subjects differences on teachers’ affect, F(1, 671) = 4.65, p = .03, partial η2 = .02, as well as within-subjects differences between when ALTs were present and not, F(1, 671) = 4.17, p = .04, partial η2 = .01. This model was shown to account for nearly 60% of the variance, R2 = .59. No significant differences were shown between students’
output for any of the conditions. The confidence interval for these conditions is also available in Table 1, indicating not only lack of statistical difference, but also strong overlap between groups.
7.3.2 Research Question 2) e. What differences in speaking output do students report in classes taught by native and non-native teachers?
A within-subjects ANOVA test found no significant differences in student self-reported output between classes taught by HRTs, ALTs, JTEs, or any combination of the above.
Results show no meaningful difference between individuals at schools with and without JTEs, F(1, 334) = .03, p = .87. Likewise classes taught by ALTs showed a similar lack of statistical signficance, F(1, 334) = .21, p = .65. Classes taught by ALTs and JTEs neared a significant effect, F(1, 334) = 3.40, p = .07, but based on the sample size, a non-significant result does not warrant further inspection. The mean, standard deviation, and confidence interval values displayed in Table 2 confirms the lack of difference in all conditions. While all results here were non-significant, similar to the above results, the model accounted for roughly 60% of the variance, R2 = .60.
7.3.3 Research Question 2) f. What effects do perceptions of each type of teachers’
spoken output have on students' reported speaking output?
A simultaneous multiple regression was used to test the relationship between the classroom environment and students’ spoken output. Results were run individually on the 4 grouping conditions. The resulting standardized beta coefficients for each group and variable can be
seen in Table 7.2. The findings consistently display a pattern of influence from teachers’
classroom environment and output frequency across the three contexts where the homeroom teacher is most involved in the lesson execution. The model R2 in each indicates that these variables explain between 7 and 16 percent of the variance associated with students’ spoken engagement.
The most consistent predictor of student output was the frequency of the teachers’
output, though it did not demonstrate a significant effect in classes led by all three teachers.
Likewise, clarity of lesson goal showed a significant relationship in schools without a JTE, but not in schools employing JTEs. None of the other hypothesized factors showed any influence on self-reported output, statistical or otherwise. In classes taught by HRTs alone, affect when speaking English significantly predicted output, though not in any other condition. Classes taught by all three teachers showed the weakest influence on students’
spoken output. The similarity of the beta values suggests similar effects across contexts.
Table 7.2. Self-reported spoken output against each of the predicted classroom variables.
Variable HRT alone (n = 199)
Standardized β HRT/ALT (n = 199)
Standardized β HRT/JTE (n = 137)
Standardized β HRT/JTE/ALT (n = 137) Standardized β
Explanation length .11 .06 -.10 -.11
Explanation clarity -.10 .03 -.11 .13
Pace of activities .01 .07 .04 .15
Peers’ engagement .01 -.05 .06 .06
Autonomy-Support .02 .03 .10 .09
Clear lesson goal .20** .26*** .13 .15
Teacher affect when
speaking English .19* .023 .06 .01
Frequency of Eng-lish output by
teacher .22** .21** .23** .06
R2 .192 .179 .127 .162
Adjusted R2 .158 .145 .072 .111
*p < .05, **p < .01, ***p < .001