4.4 Teachers’ attitude towards the amount, frequency and situations applied CS
4.4.4 Teachers’ reflection about their amount of CS used in instruction
Four statements about English-only instruction and teachers’ attitude of their amount of CS used in instruction were design in form of a 5-point Likert scale question in order to measure to what extend the participants acknowledge their CS use in teaching grammar. Table 4.8 details the proportion of teachers’
answers from strongly agree (5) to strongly disagree (1), while table 4.8 present the descriptive statistics for each statement.
Figure 4.8: Grammatical structures teachers think are difficult to teach without CS
60
Table 4.8: Teacher’s reflection about their amount of CS used in teaching grammar
Statements
Strongly disagree
(1)
Disagree (2)
Neutral (3)
Agree (4)
Strongly agree
(5) Code-switching should
only be used as a last resort when all other options have
been exhausted.
14% 11.6% 14.0% 34.9% 25.6%
The more English that is used, the better
the results for the learners.
23.3% 41.9% 4.7% 16.3% 14.0%
I want to limit my Vietnamese use in teaching grammar, but I sometimes use
Vietnamese more than I expected when my students have much trouble with English instruction.
7% 27.9% 7.0% 25.6% 32.6%
I can control the amount of code-switching I intended to use when I teach grammar
18.6% 48.8% 4.7% 20.9% 7.0%
As can be seen from the tables, 34.9% of the total teachers agree and 25.6% strongly agree with the idea that ‘CS should only be used as a last resort when all other options have been exhausted’, whereas just over one in ten disapprove. The teachers’ opinion for this statement had a mean of 3.47, which indicates that teachers on average agree with this Likert item (3.41<mean<4.20). Although the majority of participants agreed that CS is beneficial (questions 16) and that L1 and TL assist each other in the grammar classroom (section 4.3), they appeared to think that CS use should be selective
61
and limited rather than becoming habitual and unconscious. Therefore, it seems that teachers prefer to use other techniques to boost learners’ grammar learning process before the choice of CS for the reasons already mentioned in section 4.3. This finding reflects the preference of teachers in the study of Huynh, N.T.
(2018) in the sense that teachers tend to use ways to make their instruction simpler rather than using Vietnamese immediately.
On the other hand, what is striking about the figures in this table is that 41.9% of participants against the second Likert term with the idea that ‘the more English that is used, the better the results for the learners’. The mean score for this statement is 2.56, which means teachers generally disagree with the statement (1.81<mean<2.60). This result is somewhat counterintuitive when teachers mainly agree that CS should be used as the last resort. A possible explanation for this might be that teachers value learners’ comprehension, expectation and consciousness raising about grammar context more than practicing English skills when they teach grammar. This result further supports what is mentioned in section 4.3.1, one of the popular reasons for using CS to
Table 4.9: Descriptive Statistics N Range Mean
Std.
Deviation Variance Code-switching should only be
used as a last resort when all other
options have been exhausted 43 4 3.47 1.369 1.874 The more English that is used, the
better the results for the learners. 43 4 2.56 1.385 1.919 I want to limit my Vietnamese use
in teaching grammar, but I sometimes use Vietnamese more than I expected when my students have much trouble with English instruction.
43 4 3.49 1.387 1.922
I can control the amount of code- switching I intended to use when I teach grammar
43 4 2.49 1.222 1.494
Valid N (listwise) 43
62
teach Beginners is that only-English instruction might make students confused and demotivated.
From the data collected in the pilot study, many teachers reported a common difficulty relating controlling the amount of CS used in grammar instruction. Therefore, there is a need to explore whether teachers in a broader sample have the problem that although they want to limit their Vietnamese use in teaching grammar, they sometimes use Vietnamese more than they expected when their students have much trouble with English instruction. In agreement with researcher’s expectation, only a tiny proportion of the sample strongly disagree and over a quarter disagree with this issue. The mean score is 3.49, which can be inferred that teachers at VUS generally agree on being encountering with this difficulty (3.41<mean<4.02). Similarly, nearly half of the sample disagree that they can control the amount of code-switching they intended to use when they teach grammar. The mean score for the fourth Likert item is 2.49, which shows that teachers in general disagree with the statement. Further explanation from participants shows that teachers need to set a clear goal to control their amount of CS and make sure they do not use CS excessively. If too much CS is used in the classroom, both teachers and learners will be more and more dependant on it, which goes against the goal of creating an English- speaking environment. Controlling the efficient use of CS requires significantly a combination of both teachers’ skills and consciousness.
Apart from table 4.8 and 4.9, in order to measure the consistency of the design and answers in question 17, the reliability test is conducted to ensure this very important aspect of the test’s quality. Because there is one item with the corrected item value less than .3, it is deleted from the scale to increase the test- retest-reliability The result of the reliability test is presented in table 4.10, with Cronbach’s Alpha =.792, which can be suggested that the data for question 17 is reliable enough to interpret the result.
63
Table 4.10: Reliability Statistics
Cronbach's Alpha
N of Items
.792 3
Item-Total Statistics Scale Mean
if Item Deleted
Scale Variance if
Item Deleted
Corrected Item-Total Correlation
Cronbach's Alpha if
Item Deleted The more English
that is used, the better the results for the learners.
17.20 111.200 .857 .478
I want to limit my Vietnamese use in teaching grammar, but I sometimes use Vietnamese more than I expected when my students have much trouble with English instruction.
17.20 179.700 .344 .964
I can control the amount of code- switching I intended to use when I teach grammar
17.20 84.700 .812 .510
Another chi square test is applied to figure out whether there is an association between teachers’ highest relevant qualification to English Language Teaching, teachers’ years of experience and their answer about controlling their CS in question 17. Table 4.11 give information about the relation between teachers’ highest qualification and their answer for the third Likert item. Since more than one-fifth of the expected frequency (80.0%) have expected count lower than 5 when it was computed for a 4x5 crosstabulation table, a Fisher exact test is applied instead of chi-square. As can be seen from the table, a Fisher exact
64
test yielded an exact probability value of .026 which is statistically significant at the .05 level with twelve degrees of freedom. Thus it can be concluded that there is a significant difference between the teachers’ qualification and the answer of whether teachers sometimes use Vietnamese more than expected when students have much trouble with English instruction even though they want to limit their Vietnamese use in teaching grammar. A higher percentage of teachers who have only certificates or Bachelor’s degree agreed and strongly agreed with the fact that although they acknowledge avoiding overusing CS, they use Vietnamese more than expected in the case of classes with much trouble with English instruction. Meanwhile, the majority of teachers who have a Master’s degree related to English Language Teaching reported an opposite opinion. It seems that the higher qualification teachers have, the less likely they come up with this difficulty.
65
Table 4.11: Relationship between teachers’qualification and teacher’s difficulty of using more CS than expected
Teachers sometimes use Vietnamese more than expected when students have much trouble
with English instruction. Total
1 2 3 4 5
Teachers' highest qualification
Bachelor’s Count 0 3 1 7 8 19
Expected
Count 1.3 5.3 1.3 4.9 6.2 19.0
% of Total 0.0% 7.0% 2.3% 16.3% 18.6% 44.2%
Certificate (TESOL, CELTA, TEFL, TKT, etc)
Count 0 0 0 1 3 4
Expected
Count .3 1.1 .3 1.0 1.3 4.0
% of Total 0.0% 0.0% 0.0% 2.3% 7.0% 9.3%
Diploma Count 0 0 0 0 1 1
Expected
Count .1 .3 .1 .3 .3 1.0
% of Total 0.0% 0.0% 0.0% 0.0% 2.3% 2.3%
Master’s Count 3 9 2 3 2 19
Expected
Count 1.3 5.3 1.3 4.9 6.2 19.0
% of Total 7.0% 20.9% 4.7% 7.0% 4.7% 44.2%
Total Count 3 12 3 11 14 43
Expected
Count 3.0 12.0 3.0 11.0 14.0 43.0
% of Total 7.0% 27.9% 7.0% 25.6% 32.6% 100.0%
Chi-Square Tests
66
Value df
Asymp.
Sig. (2- sided)
Exact Sig.
(2-sided)
Pearson Chi-Square 18.420a 12 .104 .152
Likelihood Ratio 21.103 12 .049 .027
Fisher's Exact Test 18.942 .026
N of Valid Cases 43
a. 16 cells (80.0%) have expected count less than 5. The minimum expected count is .07.
The correlation between the fourth Likert item and their years of experience also been investigated using Fisher exact test since more than one-fifth of the expected frequency (90.0%) have expected count lower than 5 when it was computed for a 4x5 crosstabulation table. As can be seen from Table 4.12, a Fisher exact test yielded an exact probability value of .000 which is statistically significant at the .05 level with twelve degrees of freedom. It may be the case therefore that there is a correlation between the teachers’ years of experience and whether they can control the amount of CS they intended to use when they teach grammar. Although a large proportion of teachers generally admit that they are not able to always control their amount of CS in their instruction, teachers who have 10-14 years of teaching experience were likely to agree that they can control their CS more than who have 2-4 years of experience. It seems that the shorter time teachers spend for teaching career, the higher tendency that they report a negative result for their control of CS amount.
67
Table 4.12: Relationship between teachers’ years of experience and whether they can control the amount of code-switching they intended to use
Teachers can control the amount of
code-switching they intended to use Total
1 2 3 4 5
Years of experience
10_14 Count 0 1 0 4 2 7
Expected
Count 1.3 3.4 .3 1.5 .5 7.0
% of Total 0.0% 2.3% 0.0% 9.3% 4.7% 16.3%
15+ Count 0 0 2 0 0 2
Expected
Count .4 1.0 .1 .4 .1 2.0
% of Total 0.0% 0.0% 4.7% 0.0% 0.0% 4.7%
2_4 Count 6 6 0 2 0 14
Expected
Count 2.6 6.8 .7 2.9 1.0 14.0
% of Total 14.0% 14.0% 0.0% 4.7% 0.0% 32.6%
5_9 Count 2 14 0 3 1 20
Expected
Count 3.7 9.8 .9 4.2 1.4 20.0
% of Total 4.7% 32.6% 0.0% 7.0% 2.3% 46.5%
Total Count 8 21 2 9 3 43
Expected
Count 8.0 21.0 2.0 9.0 3.0 43.0
% of Total
18.6% 48.8% 4.7% 20.9
% 7.0% 100.0
%
68
Chi-Square Tests
Value Df
Asymp. Sig.
(2-sided)
Exact Sig.
(2-sided) Pearson Chi-
Square 63.864a 12 .000 .000
Likelihood Ratio 35.338 12 .000 .000
Fisher's Exact
Test 28.191 .000
N of Valid Cases 43
a. 18 cells (90.0%) have expected count less than 5. The minimum expected count is .09.
So far it is found that Vietnamese teachers at the center were aware of the pedagogical use of CS rather than using it unconsciously and habitually. They highly support the use of CS in order to meet leaners’ needs and expectations in line with CS’s pedagogical effects on leaners’ outcomes. However, they acknowledge a need to control the amount of CS so that they do not overuse it.
An interesting finding was found on their difficulty when teachers try to limit using CS excessively versus the teaching reality when they encounter classes with low level of English comprehension. As a result, there is a need to investigate some solutions which help teachers harmonize ensuring learners’
grammar outcomes by using CS into ensuring an English-speaking environment.
A deeper investigation on the suggestions will be presented in the following part.