1. Would the interaction between the input rate and English proficiency level affect the omission of words and segments in student interpreters’ output?
2. Would input rates affect the omission of words and segments in student interpreters’
output?
Four corresponding hypotheses were formulated as follows:
Hypothesis 1 English proficiency would affect the effects of input rate on the omission of words in student interpreters’ output.
Hypothesis 2 Input rates would affect the omission of words in student interpreters’ output.
Hypothesis 3 English proficiency would affect the effects of input rate on the omission of segments in student interpreters’ output.
Hypothesis 4 Input rates would affect the omission of segments in student interpreters’
output.
The means and standard deviations of OW and OS based on the observed data were calculated and presented in Table 4.1 and 4.2. Table 4.1 shows that the group means of OW were varied in the range from 2.00 to 7.67, and the group standard deviations of OW were varied in the range from 1.15 to 3.61. Table 4.2 shows that the group means of OS were
85
varied in the range from 12.63 to 43.67, and the group standard deviations of OS were varied in the range from 0.63 to 12.47.
Table 4.1 Means and Standard Deviations of Omissions of Words Arranged by Input Rate, English Proficiency, and Speech
Rate 100 130 160 Subtotal
English Proficiency
M SD n M SD n M SD n M SD n
Speech A
Low
7.67 1.53 3 5.00 3.61 3 4.40 1.67 5 5.45 2.50 11
High4.42 1.72 7 3.67 1.08 6 4.75 1.94 4 4.24 1.54 17
Subtotal5.40 2.22 10 4.11 2.10 9 4.55 1.68 9 4.71 2.02 28
Speech BLow
4.80 3.56 5 2.33 2.31 3 2.00 1.73 3 3.36 2.94 11
High2.50 1.47 4 3.50 1.89 7 2.08 1.56 6 2.76 1.71 17
Subtotal3.78 2.94 9 3.15 1.97 10 2.06 1.51 9 3.00 2.24 28
Speech CLow
3.67 1.15 3 3.80 1.92 5 4.67 1.53 3 4.00 1.55 11
High6.33 2.60 6 5.63 1.65 4 4.14 1.95 7 5.26 2.25 17
Subtotal5.44 2.52 9 4.61 1.95 9 4.30 1.77 10 4.77 2.07 28
Total4.89 2.59 28 3.93 2.03 28 3.66 1.96 28 4.16 2.25 84
Note. The means refer to the numbers of OW.86
Table 4.2 Means and Standard Deviations of Omissions of Segments Arranged by Input Rate, English Proficiency, and Speech
Rate 100 130 160 Subtotal
English Proficiency
M SD n M SD n M SD n M SD n
Speech A
Low
20.67 3.06 3 26.67 10.21 3 40.80 8.41 5 31.45 11.68 11
High14.64 7.16 7 22.75 8.69 6 30.25 4.19 4 21.18 9.29 17
Subtotal16.45 6.69 10 24.06 8.78 9 36.11 8.54 9 25.21 11.30 28
Speech BLow
21.60 7.37 5 33.33 6.11 3 43.67 4.04 3 30.81 11.24 11
High12.63 4.03 4 23.21 7.14 7 38.92 8.55 6 26.26 12.47 17
Subtotal17.61 7.46 9 26.25 8.14 10 40.50 7.44 9 28.05 12.00 28
Speech CLow
24.00 7.55 3 33.60 9.96 5 42.67 4.04 3 33.45 10.33 11
High19.17 8.47 6 24.88 0.63 4 32.21 9.20 7 25.88 9.43 17
Subtotal20.78 8.06 9 29.72 8.42 9 35.35 9.25 10 28.86 10.32 28
Total18.21 7.35 28 26.66 8.45 28 37.25 8.47 28 27.38 11.21 84
Note. The means refer to the numbers of OS.The results of linear mixed model analyses for OW (See Table 4.3) and OS (See Table 4.4) fail to indicate that there was a significant interactive effect involving the
hypotheses-related factors, namely English proficiency by input rate and English proficiency by speech by input rate. This suggests that English proficiency did not affect the effects of
87
input rate on OW and OS. Therefore, Hypothesis 1 and Hypothesis 3 are not supported by the empirical evidence in this research.
The results in Table 4.3 and 4.4 indicate that the main effect of input rate was statistically significant on OW, F (2, 66) = 3.157, p < .05, as well as on OS, F (2, 66) = 169.885, p < .01.
Table 4.3 Three-Factor ANOVA Summary Table for Omissions of Words
Source
F df1 df2 p
speech 10.295** 2 66 .000
input rate 3.157* 2 66 .049
English proficiency .068 1 66 .796
English proficiency by speech
3.690* 2 66 .030
speech by input rate .276 4 66 .892
English proficiency by input rate
1.130 2 66 .329
English proficiency by speech by input rate
2.187 4 66 .080
Note. * p < .05; ** p < .01
88
Table 4.4 Three-Factor ANOVA Summary Table for Omissions of Segments
Source
F df1 df2 p
speech 6.305** 2 66 .003
input rate 169.885** 2 66 .000
English proficiency 7.765** 1 66 .007
English proficiency by speech
.199 2 66 .820
speech by input rate 2.188 4 66 .080
English proficiency by input rate
.442 2 66 .644
English proficiency by speech by input rate
.393 4 66 .813
Note. ** p < .01
The estimated marginal means (EMM) based on the linear mixed model were calculated and pairwise comparisons were conducted and adjusted by the Bonferroni procedure, as shown in Table 4.5, 4.6 and 4.7.
Table 4.5 Estimated Marginal Means of Omissions of Words and Segments among Three Input Rates
Input rate 100 130 160
Estimated marginal means (EMM)
OW 4.899 3.988 3.674
OS 18.784 27.407 38.086
89
Table 4.6 Pairwise Comparisons of Estimated Marginal Means of Omissions of Words Input rate
Pairwise contrasts
Contrast Estimate
Std. Error
t df p
100-130 .912 .507 1.800 66 .153
100-160 1.225 .507 2.419 66 .055
130-160 .314 .507 .619 66 .538
Note. The Bonferroni-adjusted significance level is .05.
Table 4.7 Pairwise Comparisons of Estimated Marginal Means of Omissions of Segments Input rate
Pairwise contrasts
Contrast Estimate
Std. Error
t df p
100-130 -8.623** 1.049 -8.219 66 1.073E-11
100-160 -19.302** 1.049 -18.398 66 .000
130-160 -10.679** 1.049 -10.179 66 7.550E-15
Note. The Bonferroni-adjusted significance level is .05. ** p < .01
The results in Table 4.5 indicate that the EMM of OW was 4.899 for 100 wpm, 3.988 for 130 wpm, and 3.674 for 160 wpm. The post hoc results show that none of the pairwise
comparisons of EMM of OW was significant (See Table 4.6). Although the omnibus comparison among three input rates on OW was significant, the detailed pairwise
comparisons were not significant. The results of EMM suggest that OW was higher at 100 wpm than at 130 wpm and 160 wpm, contrary to the expectation of the present study that faster input rates would lead to more omissions of words. Therefore, Hypothesis 2 is not supported.
90
The EMM of OS was 18.784 for 100 wpm, 27.407 for 130 wpm, and 38.086 for 160 wpm (See Table 4.5). The results in Table 4.7 show that the pairwise comparison of EMM of OS was significant for 100-130, 100-160 and 130-160. The EMM difference of OS between 100 wpm and 130 wpm was -8.623, which was significant at .01 level, t (66) = -8.219. The EMM difference between 100 wpm and 160 wpm was -19.302, which was significant at .01 level, t (66) = -18.393. The EMM difference between 130 wpm and 160 wpm was -10.679, also significant at .01 level, t (66) = -10.179. The post hoc test suggests that OS was higher at 160 wpm than at 130 wpm and at 100 wpm, consistent with the expectation that omissions of segments would increase as the input rate became faster. Therefore, Hypothesis 4 is
supported by this research.
As for the factors unrelated to the hypotheses, Table 4.3 shows that the interaction between speech and English proficiency on OW was significant, F (2, 66) = 3.690, p < .05.
However, the post hoc tests indicate that none of the differences of OW between two English proficiency groups in speech A, B, and C was significant. The results in Table 4.3 also shows that there was a significant main effect of speech on OW, F (2, 66) = 10.295, p < .01. The EMM of OW is 4.985, 2.869, and 4.706 for speech A, B, and C, respectively. Post hoc pairwise comparisons adjusted by the Bonferroni procedure revealed that the difference of OW between speech A and B was significant, t (66) = 4.176, p < .01, and the difference of OW between speech B and C was also significant, t (66) = -3.625, p < .01, showing that OW was significantly higher in speech A and C than in speech B.
For OS, the results in Table 4.4 indicate that the main effect of speech was significant, F (2, 66) = 6.305, p < .01 and the main effect of English proficiency was also significant, F (2, 66) = 7,765, p < .01. The EMM of OS is 25.96, 28.89, and 29.42 for speech A, B, and C.
There was a significant difference of OS between speech A and B, t (66) = -2.793, p < .05, as well as between speech A and C, t (66) = -3.296, p < .01. This suggests that OS in speech A was significantly lower than in speech B and C.
91
In terms of the differences of OS between two English proficiency groups, the EMM of OS is 31.889 for the low group, significantly higher than 24.295 for the high group, t (66) = 2.787, p < .01.
4.2 Substitutions of words and segments (SW & SS)