This chapter reports the results of the grammar tests and the attitude exploration questionnaire to answer the research questions proposed in the study. The data presentation is laid out in two main sections equivalent to the two research questions posted. The results of the findings are discussed afterwards.
4.1 Results
4.1.1 Students’ grammar competence
In this section the students’ grammar scores in the pretest and posttest were presented separately. Subsequently, a paired sample t-test was calculated to find out whether the treatment was statistically significant and afterwards the tests of the normal distribution of the two sets of scores was also examined to assure that the result of this test was rigorous.
4.1.1.1 Distribution of the test scores
As shown in figure 4.1 the distribution of the pretest scores in the two groups has the same band ranging from 4.5 to 6.5. Specifically, mark 5 is reserved for 28% in both groups. As for mark 4.5, 5.5, 6 and 6.5 there are slight differences with 12%, 32%, 20% and 8% in the control group compared to 8%, 28%, 24% and 12% in the experimental group, correspondingly.
59
———
Pretest scores (control)
84.5 m5 m5.5 m6
6.5
Pretest scores (experimental)
s45 gs m5.5 m6
B65
Figure 4.1: Distribution of grammar scores of the pretests in the control group and the experimental group
As presented in figure 4.2 the distribution of the posttest scores in the two groups has different bands with 4.5-to-6.5 range in comparison to 5.5-to-7.5 one. As for band 5.5, there is a significant difference with 32% in the control group compared to 8% in the experimental group. In terms of band 6, there is a slight difference with 28% in the control group compared to 20% in the experimental group. However, as for band 6.5, the value in the experimental group, 32% is more than double to the value in the control group, 12%. Band 7 and 7.5 which are only in the experimental group correspondingly accounts for 28% and 12%.
60
Posttest scores (control)
m45 N5
M 5.5 m6 M65
Posttest scores (experimental)
m5.5
=6 6.5 N7 N75
Figure 4.2: Distribution of grammar scores of the posttests in the control group and the experimental group
As indicated in figure 4.1 and 4.2 the distribution of the scores of the two groups is distinctive; hence, comparing the mean scores of the groups is necessary to see whether the difference is statistically significant.
4.1.1.2 Comparison of the mean scores
As for the control group the calculation of the means in the pretest and posttest is presented in table 4.1.
61
Control group
95% confidence interval
Pretest Posttest of the difference
X SD xX SD Lower Upper t df P
5.42 57 5.58 57 -.39 .07 -1.39 24 | .175
P>.05
Table 4.1: Grammar means scores and a paired-sample t-test for the control group’s
As shown in table 4.1, on average, the students in the control. group achieved higher scores in the posttest (¥=5.58) than scores in the pretest (¥=5.42). The mean increase was .16, with 95% confidence interval of the difference between the means of .07 to .39. However, according to Pallant (2007) the mean difference is not statistically significant with t(24)=1.39, p= .175 > .05, indicating that there is no significant increase in the achievement scores from the pretest to the posttest; therefore, the Hp could not be rejected. Inferentially, learning grammar with the approach often used at
pretest and posttest
the center is not effective in improving the students’ grammar competence.
As for experimental group the calculation of the means in the pretest and posttest is presented in table 4.2.
62
Experimental group 95% confidence interval Pretest Posttest of the difference
X SD X SD Lower Upper t df P
5.52 | .58 | 6.58 | .57 -1.32 -.79 -835 | 24 | .000
P<.05
Table 4.2: Grammar means scores and a paired-sample t-test for the experimental group’s pretest and posttest
As shown in table 4.2, on average, the students in the experimental group achieved higher scores in the posttest (X¥=6.58) than scores in the pretest (X¥=5.52). The mean increase was 1.06, with 95% confidence interval of the difference between the means of .079 to 1.32. According to Pallant (2007), the mean difference is statistically significant with t(24)=8.35, p= .000 < .05, indicating that there is a significant increase in the achievement scores from the pretest to the posttest; therefore, it’s safe to reject Hp, Inferentially, learning grammar with dictogloss task is effective in improving the students’ grammar competence.
As indicated in table 4.1 and table 4.2 the posttest grammar means score of the experimental group (X=6.58) is higher than that of the control group (X=5.58). In order to know whether the difference in the means score is statistically significant an
independent sample t-test is run and its result is presented in the following table:
63
Fontrol 5.58 X Posttest SD 57 25 N
Experimental 6.58 57 25
Independent sample t-test
Levene’s test for 95% confidence
equality of variances interval of difference
F Sig. t Df P Lower Upper
Equal variances .000 1.000 -6.186 48 .000... -1.32 ~-.67 assumed.
Equal variances -6.186 48 .000 -1.32 -.67
not assumed
P<.05
Table 4.3: Grammar means scores and an independent sample t-test for both groups’
posttest
As shown in table 4.3 according to Levene’s test (Sig. = 1.000 > .05) there is no difference in the groups’ variances; therefore, t-value in Equal variances assumed line is employed. On average, in terms of the posttest, the students in the experimental group achieved higher scores (X=6.58) than the scores of the control group’s students (X=5.58). The mean increase was 1.00, with 95% confidence interval of the difference between the means of .067 to 1.32. According to Pallant (2007), the mean difference is Statistically significant with t(48)=6.186, p= .000 < .05, indicating that there is a significant difference in the posttest achievement scores of the two groups of students;
therefore, it’s safe to reject Hp Indeed, in terms of posttest scores the students in the experimental group outperformed the students in the control group. Inferentially, learning grammar with dictogloss task is effective in improving the students’ grammar
64
competence. With this result the conclusion is that dictogloss task better assists the students in enhancing the grammar competence than the approach often employed at
the center does.
4.1.1.3 Test of normal distribution
In order to ensure the results of the t-tests are rigorous the two tests including Shapiro- Wilk (S-W) and Q-Q plot are conducted to meet the assumption about normal distribution.
In terms of pretest scores the calculation of S-W test in the two groups is presented in
the following table:
Test of normality Pretest scores
Control group Experimental group
Mean (X) 5.42 5.52
Std. Deviation (S.D.) 57 58
Minimum 4.5 4.5
Maximum 6.5 6.5
Statistic (D) .924 921
Df 25 25
Sig. (p) .063 054
Table 4.4: Test of normality in both groups’ pretest scores
As shown in table 4.4 the pretest scores of both control group (D(25)= .924, p= .063>
.05) and experimental group (D(25)= .921, p= .054> .05) are normally distributed.
65
To confirm the statistics above the Q-Q plot is carried out in the following chart:
Normal Q-Q Plot of Pretest scores (Control) Normal Q-Q Plot of Pretest ecores (Experimental)
+ ° +
Expected Normal
e ù
Expected Normal a +
+ +
45 T 3 † Observed Value $5 T 65 T 65 T 45 7 so † Observed Value 55 7 §0 T 65
Figure 4.3: Normal distribution of the pretest scores in the control group and experimental group
In the chart the expected values form a straight diagonal line whereas the observed values are individual plots. It is concluded that the data is normally distributed as the individual plots distribute along the line.
In terms of posttest scores the calculation of S-W test in the two groups is presented in the following table:
66
Test of normality Posttest scares
Control group Experimental group
Mean (X) 5.58 6.58
Std. Deviation (S.D.) ` `
Minimum 4.5 5.5
Maximum 6.5 7.5
Statistic (D) 924 924
Df 25 25
Sig. (Pp) .063 .063
Table 4.5: Test of normality in both groups’ posttest scores
From the statistics shown in the table the posttest scores in both control group (D(25)=
924, p= .063> .05) and experimental group (D(25)= .924, p= .063> .05) are normally distributed.
To confirm the statistics above the Q-Q plot is carried out in the following chart:
67
Expected Normal
Normal Q-Q Plot of Posttest scores (Control) Normal Q-Q Plot of Posttset scores (Experimental)
3 Expected Normal
9
+ 1
Observed Value 60 ‘Observed Vaiue 65 19 7s T
Figure 4.4: Normal distribution of the posttest scores in the control group and experimental group
In chart 4.4 the expected values are a straight diagonal line whereas the observed values are plotted as individual points. The data is normally distributed as the observed values fall along the line.
In short, thanks to the statistics analyzed and interpreted it is concluded that learning with dictogloss task truly helps improve students’ grammar competence with the high significance.
4.1.2 Students’ attitudes towards learning grammar with dictogloss
As mentioned in chapter 3 the questionnaire consists of 14 items, each of which has a five-point Likert response scale with strongly agree being scored as 1 and strongly disagree being scored as 5. In the questionnaire 14 items belong to 3 main dimensions:
68
(1) Item 1 to item 5 to measure the students’ overall satisfaction with dictogloss
task. .
(2) Item 6 to item 7 to measure the students’ attitudes towards the dictogloss
texts.
(3) Item 8 to item 14 to measure the students’ attitudes towards dictogloss procedures.
L1.2.1 Students? overall satisfaction with dictogloss task
he chart below clearly shows the students’ positive attitudes towards the 5 items as a uch higher level of positive attitudes towards dictogloss task is expressed in mparison to very few negative attitudes of disagreement or strong disagreement.
| oncretely, 72% agreed that dictogloss task is enjoyable in general; 80% revealed that hẹy want teachers to spend more time on dictogloss task and less time on doing homework in the textbook; 68% felt that they like to learn other grammatical points with dictogloss besides ones taught in this course; 76% thought learning with dictogloss task is lively and 64% agreed that learning with dictogloss helps them increase their motivation in using grammar structures.
69
BSA BA ON
@D
mSD
K ltem 2
BSA mSA
BA BA
ON ON
Item 3 Í Item 4
l# SA BSA
BA mA
%N SN
Item 5
Figure 4.5: Students’ evaluation about overall satisfaction with dictogloss task
70
The table below indicates the agreement of the students towards the 5 statements as all the means scores are less than 3 (X1=1.92, X2=1.76, ¥3=2.00, ¥4=1.76, ¥=2.60).
Moreover, the standard deviations in the 5 items are not high, which means there are not diverse distributions in the way they response to the statements.
Item x SD
Item 1 1.92 81
Item 2 1.76 77
Item 3 2.00 81
Item 4 1.76 83
Item 5 2.60 1.29
Table 4.6: Mean scores and Standard deviation of items 1-5
In short the students’ evaluation for the first section is generally positive. Although there are some items to which the students show their uncertainty, they are not significant. The students clearly show their favorable attitudes towards dictogloss task.
4.1.2.2 Students’ attitudes towards the dictogloss texts
The chart below shows that the majority of the students (56%) agreed with the text length while a number of them (68%) disagreed with the fact that the text content is interesting. This indicates favorable attitudes towards the text length but negative attitudes towards text content. One thing that should be considered is a few (24% for item 6 and 20% for item 7) expressed their uncertainty towards the statements.
71
Item 7
@ SA BA
BA ON
%N BD @D
RSD BSD
Figure 4.6: Students’ evaluation about the dictogloss texts The table below shows the agreement of the students to item 6 (X¥=2.48<3) and the disagreement to item 7 (¥=3.88>3). Moreover, the standard deviation in item 6 (SD=1.38) is high, which means there are diverse distributions in the way the students response to the statement; conversely, the standard deviation in item 7 (SD=1.09) illustrates the less diverse distributions and thus it contributes to assure the negative
attitudes towards item 7.
Item Xx SD
Item 6 2.48 1.38
Item 7 3.88 1.09
Table 4.7: Mean scores and Standard deviation of items 6-7
In short the students’ evaluation for the two items in section 2 is in contrast as the majority of the students show their favorable attitudes towards the text length but negative attitudes towards text content.
72
4.1.2.3 Students’ attitudes towards dictogloss procedures
As mentioned in chapter 2 dictogloss task consists of 4 stages including preparation, dictation, text reconstruction and error correction and analysis. As for the first stage the responses are varied as 56% either strongly agreed or agreed with the time length; 16%
showed their uncertainty and 28% either disagreed or strongly disagreed with the time length. As for the second stage the majority of the students (88%) indicate their agreement towards the time length. About the third stage, a number of the students (60%) revealed their positive attitudes towards the time length; however, in terms of using English for the whole discussion their responses show the negation as 76%
disagreed with the statement. Most of the students (56%) agreed that this stage helps them recognize the grammatical points they don’t know or don’t know clearly. As for the last stage most of the students (80%) either disagreed or strongly disagreed with the time length; however, they showed the positive attitudes towards the fact that this stage helps them find out their false in using grammatical structures.
73
#8 SA
m@D msD oN
Z Se >ằ
mSA ON
BA mD
ON mSD
ON BD mSD
Figure 4.7: Students’ evaluation about dictogloss procedures
74
The table below shows that the students agreed with most of the statements as the means scores are less than 3 (X8=2.56, X9=1.60, X10=2.16, X12=2.52, X14=2.12).
However, they disagreed with statement 11 and statement 13 as the means scores are above 3 (¥11=4.12, X13=4.08). Moreover, the standard deviation in the two items are low (SD11= .78, SD13= .70), which means there are not diverse distributions in the way they response to the statement.
Item xX SD
Item 8 2.56 a 1.19
Item 9 1.60 70
Item 10 2.16 80
Item 11 4.12 -78
Item 12 2.52 1.19
Item 13 4.08 70
Item 14 2.12 92
Table 4.8: Mean scores and Standard deviation of items 8-14
In short the student showed their positive attitudes towards the time length of the first three stages but the disagreement to the time length of the final stage in dictogloss task.
In the third stage they agreed with the recognition of the unknown or unclearly known grammatical points but disagree with the using of English for the whole discussion.
Moreover, in the last stage they indicated their favorable attitudes to false finding in using the grammatical structures.
4.2 Discussion of the findings
4.2.1 Findings on the students’ achievement
75
The t-test computed to test the mean differences between the pretest and posttest within each group indicated that the mean difference in the experimental group was statistically significant at .05 level. It means there was significant improvement in the subjects’ performance on the posttest in comparison to their performance on the pretest. Hence, dictogloss task clearly helped enhance the subjects’ grammar competence.
Obviously, the task gave the students the contexts in which a specific linguistic form had to be used. Moreover, in the task the students enhanced their involvement in interaction as the students had to show their agreement on a single solution. This fact required the students to engage in extensive discussion about the appropriate grammatical forms. Consequently, their awareness of the form was raised and their specific aspect of language was improved. Nabei (1996) determined the significance of interaction in dictogloss task as under the task the students truly involved in the language forms and importantly the interaction created not merely the meaning-based interaction but also the grammar-based interaction. Thanks to the awareness of the forms the accuracy of language use is significantly improved (Swain & Lapkin, 1998).
In addition, the task gave the students the opportunities to make comparisons between their current interlanguage (as realized in their output) and the target language. These opportunities were provided by means of the reconstruction activity which assists the students in recognizing their linguistic difficulties. The reconstruction stage also made allowance for the functions of output (Mayo, 2002) in that students tested their hypothesis about their target language, reflected on their production and processed the target language syntactically. With these functions, the students took more notice of their linguistic problems and the knowledge they needed to learn in L2 and the students could reflect on their own language use.
76
In brief, with the convincing results obtained from the students in the experimental group it is confirmed that dictogloss task produces the high achievement which is more fruitful than what the center-applied approach did. Its application truly enhances the students’ grammar competence. As a matter of fact, the results of the tests agree with the majority of the studies which claimed that the students under dictogloss task outperformed the other students as for grammar competence (Swain & Lapkin, 2000;
Kuiken & Vedder, 2002a; Akbari, 2009; Alsibai, 2008). Consequently, dictogloss has proved to be a positive task in grammar teaching and learning and lay significant impact on the acquisition of specific linguistic forms.
4.2.2 Findings on the students’ attitudes
As mentioned in chapter 4 the Likert scale questionnaire delivered to explore the students’ attitudes consists of 3 main sections:
(1) Item 1 to item 5 to measure the students’ overall satisfaction with dictogloss (2) Item 6 to item 7 to measure the students’ attitudes towards the texts
(3) Item 8 to item 14 to measure the students’ attitudes towards certain dictogloss procedures
The findings demonstrated that the students in the experimental group showed their positive reactions to the first section which aimed at finding out their level of satisfaction with dictogloss. Since the first section was positively responded to by the students in the experimental group it is concluded that doing dictogloss task was a positive experience for them. They found that dictogloss task was not only enjoyable but also lively as they could interact with their peers in the reconstruction stage and with the teacher in the other three stages. By doing so the upper group of students was
77
in the position to help the lower group of students and by doing so the students are able to strengthen their own understanding of the target language.. Lesser (2004 as cited in Stockwell, 2010) noted that students usually have better performance in classroom tasks when working together, rather than alone. Moreover, dictogloss task required the students to perform very specifically context-based task. That’s why, the students found that dictogloss task helped them possess the increasing motivation to use the grammatical structures thanks to the contexts created. Thanks to the liveliness and contextualization the students liked the teachers to spend much more time doing dictogloss task and teaching other grammatical points by using dictogloss.
The findings on the students’ satisfaction towards dictogloss in this thesis agree with the results in Alsibai (2008)’s study. In her study she explored the students’ attitudes toward dictogloss task and the findings were that they experienced the task positively thanks to its liveliness and interaction creation. In this sense, the results from this study corroborate Alsibai’ study on the fact that dictogloss task was truly a favorable experience for the students in the experimental group.
The findings on the second section which aimed at investigating the students’ opinions towards the texts in dictogloss task showed the contrast in the students’ opinions towards the two components of the text; namely, text length and text content. As for the text length most of the students confirmed its suitability. In this study, all the texts comprised 8-12 sentences fluctuating from 100-150 words, which met the standard mentioned by Nabei (1996). As for test content, in contrast, most of the students manifested their disagreement towards the interesting content. In this study all the texts were extracted from the textbook as the content couldn’t attract the students. In fact, the content of the texts can be selected from the authentic materials or the teacher can create his own (Jacob & Small, 2003) so that the text content can be interesting enough
78
to attract the students. However, in this study, owing to the text’ difficulty level, the texts were just selected in the textbook; therefore, the content was not attractive enough to satisfy the students. To the best of my knowledge since no other research has explored the subjects’ attitudes towards dictogloss text the findings cannot be evaluated and compared against other similar questionnaires.
The third section focused on exploring the students’ attitudes towards certain dictogloss procedures. The findings were that most of the students reacted negatively to the fact that they spent most of the time using English to discuss during the reconstruction stage. That is the main issue in doing dictogloss task since worries may arise when the students are on their own without the teacher’ intervention and remark at the reconstruction stage of dictogloss task. More concretely, they do not speak the target language and do not learn the mistakes mutually when the collaborative dialogs are taking place. The solution for these two problems mentioned in Alsibai (2008)’s study is that while students are involved in reconstruction it’s necessary for the teacher to observe the groups working, listen to their discourse without interference, and record the students’ common mistakes. In this way, the teacher is able to solve the two problems simultaneously; first, on closely monitoring the students the teacher can assure that everyone is speaking the target language most of the time and secondly, the teacher is able to hear the common mistakes, and thus in the error analysis and correction stage of dictogloss task he can spend more time explaining why the mistakes are made and how the students can avoid them in the future. In this section the students also expressed their negative attitudes towards the timing in the error analysis and correction stage as the time was inadequate for correcting all groups’ mistakes.
Although this stage was designed to help the students overcome the grammar misunderstandings and inaccuracies sometimes it came too late in the task.
Consequently, a possible solution to the problem mentioned in Alsibai (2008)’s study
79