CHAPTER V ANALYSIS AND DISSCUSSIONS 5.1 COMPARISON OF LISTENING COMPREHENSION TEST SCORES (T-TEST) 5.2 COMPARISON OF LISTENING COMPREHENSION TEST SCORE: ANALYSIS OF CO-VARIANCE 5.3 VIEWS OF HIGH SCHOOL ENGLISH TEACHERS REGARDING VARIOUS ASPECTS OF TEACHING AND LEARNING OF ENGLISH 5.4 RATINGS OF PUPILS ON THE SUITABILITY OF INSTRUCTIONAL MATERIAL IN DEVELOPING THE SUB-SKILLS OF LISTENING COMPREHENSION 5.5 EVALUATION OF THE INSTRUCTIONAL MATERIAL PREPARED FOR DEVELOPING LISTENING COMPREHENSION IN ENGLISH AT HIGH SCHOOL LEVEL BY EXPERTS
42
Embed
ANALYSIS AND DISSCUSSIONSshodhganga.inflibnet.ac.in/bitstream/10603/12703/13/13...Arithmetic Mean (AM) Standard Deviation (SD) Critical Ratio (CR) Level of Significance Experimental
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
CHAPTER V
ANALYSIS AND DISSCUSSIONS
5.1 COMPARISON OF LISTENING COMPREHENSION
TEST SCORES (T-TEST)
5.2 COMPARISON OF LISTENING COMPREHENSION
TEST SCORE: ANALYSIS OF CO-VARIANCE
5.3 VIEWS OF HIGH SCHOOL ENGLISH TEACHERS REGARDING
VARIOUS ASPECTS OF TEACHING AND LEARNING OF
ENGLISH
5.4 RATINGS OF PUPILS ON THE SUITABILITY OF
INSTRUCTIONAL MATERIAL IN DEVELOPING THE
SUB-SKILLS OF LISTENING COMPREHENSION
5.5 EVALUATION OF THE INSTRUCTIONAL MATERIAL
PREPARED FOR DEVELOPING
LISTENING COMPREHENSION IN ENGLISH AT
HIGH SCHOOL LEVEL BY EXPERTS
ANALYSIS AND DISCUSSIONS
A systematic organisation, classification and tabulation of data are the bench
marks of a good research study. Analysis of data means studying such organised material
in order to discover inherent facts. Goode and Hatt (1952) points out that “analysis is
among the first of the methods which a researcher will call upon his effort to reduce a
field of size, that is, to extract a manageable chunk from the infinite complexity of real
world”.
Explaining the importance of analysis of data, Aggarwal (1966) remarks,
“However valid, reliable, and adequate the data may be, it does not serve any worthwhile
purpose, unless it is carefully edited, systematically classified and tabulated, scientifically
analysed, intelligently interpreted and rationally concluded”. Analysis thus involves
breaking down existing complex factors into simpler parts and putting the parts together
in new arrangements for purposes of interpretation. It is the study of data from as many
angles as possible to explore new facts.
Discussion or interpretation is the critical examination of the results obtained in
the study. The data collected for the present study are analysed with the specific aims of
eliciting findings which are necessary to arrive at definite conclusions in relation to the
objectives of the study.
The major purpose of this study is to determine experimentally the effectiveness
of the prepared Instruction Material in developing listening comprehension in English at
Analysis and Discussions
212
the High School Level. As a preliminary step to this, an attempt was made to study the
views of high school teachers of English regarding various aspects of teaching - learning
of English, and teaching of listening comprehension at present and to collect suggestions
for improving the teaching of listening comprehension.
Experimental cum Survey method was adopted for the study. ‘Non-equivalent
pre-test – post-test control group’ design was used for the study in which the experimental
group (N=94) was exposed to the prepared Instructional Material and the control group
(N=96) was not given any treatment. The data relevant for the survey were collected from
high school English teachers (N=50).
A group of subject experts (N = 10) validated the prepared Instructional Material
and Listening Comprehension Test. The data thus collected were analysed using
appropriate statistical technique, so as to verify the hypothesis of the study. The details of
the analysis carried out are presented below:
Analysis and Discussions
213
SECTION I
5.1. COMPARISON OF LISTENING COMPREHENSION TEST SCORES:
EXPERIMENTAL GROUP AND CONTROL GROUP
(TOTAL SAMPLE AND SUB-SAMPLES)
This section of analysis is intended to test the effectiveness of the Instructional
Material (IM) in developing Listening Comprehension (LC) by comparing the pre-test
and post-test scores, of Listening Comprehension Test (LCT), of the experimental group
and control group for total sample, and for sub-samples (experimental group) based on
gender, locality of school and management of school
The mean and standard deviation of Pre-test and Post-test scores of LCT of the
experimental and control groups were computed and the difference in means was tested
for significance. The details are given under the following heads:
5.1.1 Comparison of Pre-LCT scores: Experimental and Control groups.
5.1.2 Comparison of Post-LCT scores: Experimental and Control groups.
5.1.3 Comparison of Pre-LCT and Post-LCT scores: Experimental group.
5.1.4 Comparison of Pre-LCT and Post-LCT scores: Control group.
5.1.5 Comparison of Post-LCT scores: Experimental group (Sub-samples).
Analysis and Discussions
214
5.1.1 COMPARISON OF PRE-LCT SCORES:
EXPERIMENTAL AND CONTROL GROUPS
The details regarding the data and results of comparison between Pre-test scores
of LCT of Experimental group and Control group are given in Table 5.1.
Table 5.1
t-test between the means of Experimental and Control groups:
Pre-LCT scores
Treatment groups Number
(N)
Arithmetic
Mean
(AM)
Standard
Deviation
(SD)
Critical
Ratio
(CR)
Level of
Significance
Experimental
group 94 12.64 4.03
0.36
(P>0.05)
Not
Significant Control
group 96 12.88 5.14
Statistical analysis of the data (Table 5.1) revealed that, there is no significant
difference (CR = 0.36; P>0.05) between Experimental and Control group with regard to
their Pre-LCT scores. It can be concluded from the analysis that the Experimental and
Control groups are more or less equal, with regard to Pre-LCT scores.
Tenability of Hypothesis
The t-test between the means of Experimental and Control groups revealed that,
the two groups do not differ significantly with regard to Pre-LCT scores. Hence
hypothesis formulated in this content viz., H1 (There will be significant difference
between the experimental group and control group with regard to the pre-LC test scores.)
is rejected.
Analysis and Discussions
215
5.1.2 COMPARISON OF POST-LCT SCORES:
EXPERIMENTAL AND CONTROL GROUPS
The details regarding the data and result of comparison between Post-test scores
of LCT of Experimental and Control Group are given in Table 5.2.
Table 5.2
t-test between the means of Experimental and Control groups:
Post-LCT scores
Treatment groups Numbers
(N)
Arithmetic
Mean
(AM)
Standard
Deviation
(SD)
Critical
Ratio
(CR)
Level of
Significance
Experimental
group 94 21.94 6.31
9.83
(P<0.01) 0.01
Control
group 96 13.27 5.83
Statistical analysis of the data (Table 5.2) revealed that, there is significant
difference (CR = 9.83; P < 0.01) between Experimental and Control Group with regard to
Post-LCT scores. The difference is in favour of the Experimental group.
Tenability of Hypothesis
The t-test between the means of Experimental and Control group revealed that, the
two groups differ significantly with regard to Post-LCT scores. Hence hypothesis
formulated in this content viz., H2 (There will be significant difference between the
experimental group and control group with regard to the post-LC test scores.) is accepted.
Analysis and Discussions
216
5.1.3. COMPARISON OF PRE-LCT AND POST-LCT SCORES:
EXPERIMENTAL GROUP
The details regarding the data and results of comparison between Pre-test and
Post-test scores of LCT of Experimental Group are given in Table 5.3.
Table 5.3
t-test between the means of Pre-LCT and Post-LCT scores:
Experimental Group
Test Numbers
(N)
Arithmetic
Mean
(AM)
Standard
Deviation
(SD)
Critical
Ratio
(CR)
Level of
Significance
Pre-test 94 12.64 4.03 12.04
(P<0.01) 0.01
Post-test 94 21.94 6.31
Statistical analysis of the data (Table 5.3) revealed that, there is significant
difference (CR = 12.04; P < 0.01) between the means of Pre-LCT and Post-LCT scores of
Experimental group. The difference is in favour of Post-LCT scores.
Tenability of Hypothesis
The t-test between the means of Pre-LCT and Post-LCT scores revealed that, there
is significant difference between the Pre- and Post-LCT scores of Experimental group.
Hence the hypothesis formulated in this context viz., H3 (There will be significant
difference between the means of pre-LC test and post-LC test scores of the experimental
group) is accepted.
Analysis and Discussions
217
5.1.4 COMPARISON OF PRE-LCT AND POST-LCT SCORES: CONTROL GROUP
The details regarding the data and results of comparison between Pre-test and
Post-test scores of LCT of Control group are given in Table 5.4.
Table 5.4
t-test between the means of Pre-LCT and Post-LCT scores: Control Group
Test Numbers
(N)
Arithmetic
Mean
(AM)
Standard
Deviation
(SD)
Critical
Ratio
(CR)
Level of
Significance
Pre-test 96 12.88 5.14 0.49
(P>0.05)
Not
Significant Post-test 96 13.27 5.83
Statistical analysis of the data (Table 5.4) revealed that there is no significant
difference (CR = 0.49; P > 0.05) between the means of Pre-LCT and Post-LCT scores of
Control Group.
Tenability of Hypothesis
The t-test between the means of Pre-LCT and Post-LCT scores revealed that, there
is no significant difference between the Pre and Post-LCT scores of Control Group.
Hence the hypothesis formulated in this content viz., H4 (There will be significant
difference between the means of pre-LC test and post-LC test scores of the control group)
is rejected.
Analysis and Discussions
218
5.1.5 COMPARISON OF POST-LCT SCORES:
EXPERIMENTAL GROUP (SUB-SAMPLES)
To determine whether variables like gender, locality of school and management of
school affect the Post-LCT scores of pupils, the differential effect of the above mentioned
variables was studied. The details regarding the data and results of the test of significance
for difference between the means of Post-LCT scores for the sub-samples of
Experimental group are given in Table.5.5.
Table 5.5
t-test between the means of Post-LCT scores:
Experimental Group (Sub-samples)
Sl.
No. Variable Category
Number
(N)
Arithmetic
Mean
(AM)
Standard
Deviation
(SD)
Critical
Ratio
(CR)
Level of
Significance
1 Gender Male 46 21.43 5.47
0.76 Not
Significant Female 48 22.42 7.04
2
Locality
of
School
Rural 68 21.88 6.38
0.14 Not
Significant Urban 26 22.08 6.23
3
Manage-
ment of
School
Government 34 24 6.3
2.42 0.05 Aided 60 20.77 6.05
Statistical analysis of data (Table 5.5) revealed that there is no significant
difference (CR = 0.76; P > 0.05) between the means of Post-LCT scores of male and
female pupils.
Analysis and Discussions
219
It also revealed that there is no significant difference (CR= 0.14; P > 0.05)
between the means of Post-LCT scores of pupils belonging to rural and urban schools.
It was found that there is significant difference (CR= 2.42, P < 0.05) between the
means of Post-LCT scores of pupils studying in government and aided schools. The
difference is in favour of pupils studying in government school.
It can be concluded that Post-LCT scores were influenced by the variable viz.,
management of school. As regards gender and locality of school, no significant difference
was noted. Hence it can be concluded that Post-LCT scores were influenced only by the
management of school.
Tenability of Hypothesis
The t-test between the means of Post-LCT scores of experimental group (sub-
sample) revealed that, there is significant difference between pupils studying in
government and aided schools. Hence hypothesis formulated in this context viz.,
H5(iii) (There will be significant difference between pupils studying in government schools
and in aided schools with regard to post-LC test scores) is accepted.
The t-test between the means of Post-LCT scores of experimental groups (sub-
sample) also revealed that, there is no significant difference between male and female
pupils, and pupils belonging to urban and rural schools. Hence hypotheses formulated in
this contest viz., H5(i) (There will be significant difference between male and female
pupils with regard to post-LC test scores) and H5(ii) (There will be significant difference
between pupils studying in schools in urban locality and rural locality with regard to post-
LC test scores) are rejected.
Analysis and Discussions
220
SECTION II
5.2 COMPARISON OF LISTENING COMPREHENSION TEST SCORE:
EXPERIMENTAL AND CONTROL GROUP
USING ANALYSIS OF CO-VARIANCE
The experimental group exhibited better performance after being exposed to the
Instructional Material. Though it could be tentatively decided that the prepared
Instructional Material is effective in developing listening comprehension, it was difficult
to decide whether the difference between the pre-LCT and post-LCT scores resulted from
the experimental factor or from other variables, as it was highly inconvenient to equate
the group before the experiment. Therefore, it become necessary that the scores be
analyzed using the technique of Analysis of Co-variance (ANCOVA), which is a
statistical contrivance used to compare groups that are initially unlike, either in the
variable under study or some presumably related variable. Difference in the initial status
of the groups can be removed statistically using ANCOVA, so that they can be compared
as though their initial status has been equated. The use of ANCOVA is thus justified for
the analysis of the scores in the present study. The statistical procedure of ANCOVA, as
suggested by Garrett (1981), is strictly followed.
Analysis and Discussions
221
5.2.1 COMPARISON OF PRE-LCT AND POST-LCT SCORES:
EXPERIMENTAL AND CONTROL GROUP
USING ANALYSIS OF VARIANCE (ANOVA)
In this part of the analysis the pre-LCT and post-LCT scores, sum of squares,
mean square variance and F-ratios for the Pre- and Post-LCT scores of the experimental
and control group were computed as follows:
Step 1: Correction Term
There are three correction terms to be applied to the sum of squares – correction
for X (pre-test scores), correction for Y (post-test scores) and correction for XY. Cx, Cy,
and Cxy are calculated using the formulae:
= = 30950.66
= = 58573.14
= = 42577.89
Step II: Sum of Squares for Total (SS)
There are three SS’s for Total: SSX, SSY, and SSXY.
Total SS for X; SSX = ΣX2 − = 34971 – 30950.66 = 4020.34
Total SS for Y; SSY = ΣY2
− = 69068 – 58573.14 = 10494.86
Total SS for XY; SSXY = ΣXY − = 47139 – 42577.89 = 4561.11
Analysis and Discussions
222
Step III: Sums of Squares among Means
Sum of squares among the group means are calculated using the formulae:
SS among Group Means for X =
= = 2.43
SS among Group Means for Y =
= = 3566.28
SS among Group Means for XY =
= = (– 93.03)
Step IV: Sum of Squares within Group
Sum of squares within groups are calculated as follows:
SS within Groups for X = Total SS for X − SS among Group Means for X
= 4020.34 – 2.43 = 4017.91
SS within Groups for Y = Total SS for Y − SS among Group Means for Y
= 10494.86 – 3566.28 = 6928.58
Analysis and Discussions
223
SS within Groups for XY = Total SS for XY − SS among Group Means for XY
= 4561.11 – (– 93.03) = 4654.14
Step V: Analysis of Variance of X and Y scores, taken separately
The ‘F’ test is applied to the initial and final scores, to decide whether the scores
approach closer to significance.
Degrees of freedom (df)
Among Group means df = Total number of groups – 1 = 2 – 1 = 1
Within group df = Total sample – Number of groups = 190 – 2 = 188
Mean Square Variance
MSx
MSx Among group means = = = 2.43
MSx Within group = = = 21.37
MSy
MSy Among group means = = = 3566.28
MSy Within group = = = 36.85
Analysis and Discussions
224
Fx = = = 0.114
Fy = = = 96.78
The Analysis of Variance (ANOVA) for the Pre- and Post-LCT scores of the
experimental and control groups are given below in table 5.6.
Table 5.6
Summary of Analysis of Variance of Pre-LCT and Post-LCT scores:
Experimental and Control group
Source of
Variation df SSx SSy MSx MSy
Among
means 1 2.43 3566.28 2.43 3566.28
Within
Groups 188 4017.91 6928.58 21.37 36.85
Total 189 4020.34 10494.86 23.8 3603.13
Fx = 0.114 from table F, for df 1/188
Fy = 96.78 F at 0.05 = 3.90
F at 0.01 = 6.81
The F-ratios for the Pre-LCT and Post-LCT scores were tested for significance.
The Fx value obtained is 0.114 (Fx = 0.114; P>0.05) and is very low. Fx value obtained
revealed that, there is no significant difference between the Pre-test scores of the
experimental and control groups. The two groups are more or less equal with regard to the
Pre-LCT scores.
Analysis and Discussions
225
The Fy value obtained is 96.78 (Fy = 96.78; P<0.01) is very high. This indicates
that, there is significant difference between the Post-LCT scores of experimental and
control group.
Tenability of Hypothesis
Analysis of Variance between the experimental and control groups revealed that,
the two groups differ significantly with respect to Post-LCT scores. Hence the hypothesis
formulated in this contest viz., H6(i) (There will be significant difference in the Analysis
of variance (ANOVA) between the experimental and control group with regard to post-
LC test scores) is accepted.
5.2.2 COMPARISON OF PRE-LCT AND POST-LCT SCORES:
EXPERIMENTAL AND CONTROL GROUP
USING ANALYSIS OF CO-VARIANCE (ANCOVA)
Analysis of Co-variance (ANCOVA) represents an extension of ANOVA to allow
for the correlation between initial and final scores. ANCOVA was applied for analysing
the data in order to effect adjustments in final or terminal scores to determine the extent
of difference between the experimental and control group and hence determine the
effectiveness of the prepared Instructional Material.
For correcting the post-LCT scores (y-scores) for the difference in the pre-LCT
scores (x-scores), the adjusted sum of squares, and adjusted mean square variance were
computed. The F-ratio was also calculated as follows and the results are presented below
in Table 5.7.
Analysis and Discussions
226
Step VI: Computations of Adjusted SS for Y (SSY.X)
The computations of adjusted SS carried out in this step are for the purpose of
correcting the final (Y – post-test) scores for differences in initial (X – pre-test) scores.
The symbol SSy.x, means that the SSy have been adjusted for any variability in Y scores
contributed by X scores, or that the variability in X is held constant. The general formula
for adjusted SS is,
SSyx = SSy −
Total SS = Total SS for Y − = 10494.86 − = 5320.24
Within SS =Within SS for Y − = 6928.58 − = 1537.46
Among Mean SS = Total SS − Within SS = 5320.24 − 1537.46 = 3782.78
Adjusted Mean Squares
From the various adjusted sum of squares of the variance (MSY.X) can be computed
by dividing each SS by its appropriate degrees of freedom. The ‘F’ test is applied to the
adjusted, among and within Y means.
MSY.X (Among means) = = = 3782.78
MSY.X (Within group) = = = 8.22
Analysis and Discussions
227
FY.X = = = 460.19
Step VII: Significance of differences among adjusted Y Means SDyx
SDyx = = = 2.87
Table 5.7
Summary of Analysis of Co-variance of Pre-LCT and Post-LCT scores:
Experimental and Control Groups
Source of
variation df SSx SSy SSx.y SSy.x MSy.x SDy.x
Among
means 1 2.43 3566.28 -93.03 3782.78 3782.78
2.87
Within
groups 187 4017.91 6928.58 4654.14 1537.46 8.22
Total 188 4020.34 10494.86 4561.11 5320.24 3791
Fy.x = 460.19 from Table F for df = 1/187
F at 0.05 level = 3.90
F at 0.01 level = 6.81
The obtained value of F-ratio is 460.19 and is greater than the table value at
0.01 level and hence is significant (Fyx = 460.19; P < 0.01). This significant F-ratio for the
adjusted post-LCT scores shows that the post-LCT scores of pupils in the Experimental
Analysis and Discussions
228
and Control Group differ significantly after they have been adjusted for the differences in
the pre-LCT scores.
It further necessitates, proceeding to test for the difference between the adjusted
post-LCT means of the experimental and control groups. The adjusted means for the post-
LCT scores of pupils in the two groups were computed using correlation and regression
as follows and the details are given in Table 5.8.
Step VIII: Correlation and Regression Means
The regression coefficients for total, among means and within groups have been
calculated by use of the formula
b ( Within) = = = 1.16
Step IX: Calculation of adjusted Y means
Y means can be adjusted directly for difference in the X means by use of the formula,