A quasi-experimental procedure with a pretest-posttest design was implemented to identify the relationships of the intervention, parents’ training for participation, by comparing treated units (kids) to control units. The purpose was to recognize the relationships among language teaching, parents’ training for participation, kids’ activities, parents’ involvement, and learning level in the three groups. Thus, a comparison was made between two groups of kids and a control group of kids that their parents received no training. The aim was to evaluate performances between the groups.
Participants
The Statistical Population comprises 60 Iranian children between 9 and 10 and their parents who neither are familiar with the English language at all (they know just several English words). The 60 children were divided into two groups of experimental 1 and experimental 2 and the control group.
The Experimental Group
The experimental group 1 consisted of 20 children learning English through face-to-face classroom. At the same time, their parents were taught through Telegram (the most favorite social network in Iran) with the use of educational videos and short texts. The parents of this group participated in a 6_hour workshop on how to involve in the program. The experimental group 2 consisted of 20 children learning English through face-to-face classrooms. Their parents only participated in the six-hour workshop mentioned above and were not taught the English language at all.
The Control Group
In the control group, the children were just taught the English language and their parents neither learned English nor participated in the workshop on how to involve.
The intervention
The intervention in the experimental and the control group as mentioned earlier, both experimental groups participated separately in a workshop before the program. They were taught the methods and ways of involvement and were encouraged to participate more actively in their children's learning process. They were told that they can help their children in story reading, doing the exercises, memorizing words, reading English poems and watching animated cartoons.
The instruments
A pretest and a posttest were conducted. An English Language Course was used to assess kids’ learning performance. On a scale of 0.00 to 0.99, reliability coefficients generally fall between 0.75 and 0.86 (Educational Testing Service, 2005).
Questionnaire
A set of questionnaires, called “PI-SHBScienceLAS” adopted from Karaçöp et al. (2016) were answered by participants before and after the course.
Questionnaire items
Active Involvement of Parents: This questionnaire consisted of seven items and measured parents’ involvement level in the learning process of kids. The related items are: “I allocate time to make studies with my child improving his/her English Language skills”, “I am telling my child that I like to learn new things about English language”, “I have enough knowledge to help my child with his/her assignment of English language”, “I establish clear rules for my child to do his/her assignment at home”, and “I enjoy helping my child with his/her assignment of English language lesson”, “While helping my child with his/her assignment of English language lesson, I can make explanations by giving examples apart from the ones given in the book”, and “By helping my child with his/her assignment of English language lesson I think that I make a difference in his/her school performance” ”. A reliability test on the 5-item (1 = totally disagree, 5 = totally agree) scale indicated an acceptable internal consistency (α =.75).
Engagement: Respondents were divided in two groups. A group has received training and another one, has not received training.
Parents and student’s joint activities: This scale contained five items. Respondents were asked to determine the activities including Homework, Word, Reading the story, Poem, and Cartoon in which have done along with their kids.
Level of students’ learning: The scale includes an item on a four-degree spectrum, from 1 (weak) to 4 (excellent).
Language teaching: The experimental groups were received language teaching to participate the in the learning process of their kids.
Analytical procedure
The statistical procedures employed to answer the research questions in chronological order, are:
- Multivariate Analysis of Variance (MANOVA)
- Pearson correlation analysis
- Partial Least Square (PLS)
- Multiple regression analysis
The data was also analyzed by the Statistical Package for the Social Sciences, version 22.0 and XLSTAT Premium version 19.
First question: How will parents' training affect their children's participation, actions, and learning levels?
Table number 2 shows the mean and standard deviation for some variables, including students' learning, parents’ participation, and joint activities of parents and students. Descriptive statistics show that the mean scores related to the students' level of skills and learning (3.70) and the parents' participation level (3.49) were higher in the first experimental group than the second group and the control group. In addition, in the first experimental group, the mean of activities including assignments (14.55), writings (52.75), reading stories (19.99), poetry (4.25), and cartoon comprehension (11.55) was more than the second experimental group and also the control group. Furthermore, the second experiment group had higher means in the three variables above than the control group.
Table 2. Descriptive Statistics
Variable
|
Dimension
|
Group
|
Mean
|
Std. Deviation
|
N
|
Learning level of kids
|
-
|
Experimental group 1
|
3.70
|
0.66
|
20
|
Experimental group 2
|
2.70
|
0.47
|
20
|
Control group
|
2.10
|
0.45
|
20
|
Parents participation
|
-
|
Experimental group 1
|
3.49
|
0.25
|
20
|
Experimental group 2
|
2.50
|
0.20
|
20
|
Control group
|
1.24
|
0.23
|
20
|
Joint activities of parents and students
|
Homework
|
Experimental group 1
|
14.55
|
0.60
|
20
|
Experimental group 2
|
13.60
|
1.19
|
20
|
Control group
|
11.20
|
1.47
|
20
|
Word
|
Experimental group 1
|
52.75
|
2.15
|
20
|
Experimental group 2
|
43.85
|
3.34
|
20
|
Control group
|
21.00
|
5.74
|
20
|
Reading the story
|
Experimental group 1
|
19.90
|
1.29
|
20
|
Experimental group 2
|
18.60
|
1.64
|
20
|
Control group
|
8.90
|
1.77
|
20
|
Poem
|
Experimental group 1
|
4.25
|
0.55
|
20
|
Experimental group 2
|
2.75
|
0.44
|
20
|
Control group
|
0.90
|
0.64
|
20
|
Cartoon
|
Experimental group 1
|
11.55
|
1.39
|
20
|
Experimental group 2
|
8.60
|
2.35
|
20
|
Control group
|
2.65
|
0.49
|
20
|
Multivariate Analysis of Variance (MANOVA) was used in order to review training of parents concerning their participation, the amount it had affected their participation, actions, and learning level of students, and also by regarding specific training for parents (participatory and verbal) that is in complete contrast to the first and second experimental group and control group.
In this test, related variables should have a meaningful relationship. For example, the dependent variables of parents' participation, mutual actions (parents and students), and learning level of students have theoretically close relationships with each other by means that it is expected to increase parents' participation, actions, and learning level students increase. Before applying the MANOVA test, the test assumptions were examined by Box1, Levene 2, Pillai's Trace 3, Wilks' Lambda 4, Hotelling's Trace 5, and Roy's Largest Root. The box test was used for three variables to examine the equality of the variance-covariance matrix. The results of the box test are shown in the table below. Table 2 shows the level of significance (p = 0.05), which indicates that the equality condition of the variance-covariance matrix is well observed (P <0.05, F = 2.069).
Table 3. Box's Test of Equality of Covariance Matrices
Box's M
|
141.067
|
F
|
2.069
|
df1
|
56
|
df2
|
9280.317
|
Sig.
|
.000
|
Levene test was used to review equality of variances of learning level, skill learning of students, parents' participation, and joint actions of parents and students. The Levene test results have been represented below table. This table shows that the variances of these three variables are not equal in the three groups (1 and 2) and the control group. However, they are significantly distinctive (P≤0.05), which indicates the reliability of the following represented results.
Table 4. Levene's Test of Equality of Error Variances
|
F
|
df1
|
df2
|
Sig.
|
Learning level
|
2.019
|
2
|
57
|
.042
|
Parents participation
|
3.831
|
2
|
57
|
.031
|
Homework
|
14.084
|
2
|
57
|
.000
|
Word
|
2.731
|
2
|
57
|
.036
|
Reading the story
|
5.721
|
2
|
57
|
.018
|
Poem
|
2.335
|
2
|
57
|
.027
|
Cartoon
|
7.591
|
2
|
57
|
.001
|
In the below table, the results of the Wilks' Lambda test are represented. The Wilks Lambda test statistics fluctuate between zero and one. Moreover, when approaching zero, it shows the more significant difference in the average of quantitative variables that, as approaching one, means no difference in means between groups.
Results showed that among the three experimented groups, there was a meaningful distinction at least among one of the below variables: learning level, skill learning of students, parents' participation, and joint actions of parents and students (130/87 F= 05/0P≤ ).
Table 5. Multivariate Tests
Effect
|
Value
|
F
|
Hypothesis df
|
Error df
|
Sig.
|
Intercept
|
Wilks' Lambda
|
0.001
|
6470.744b
|
7
|
51
|
0.000
|
Group
|
0.006
|
87.130b
|
14
|
102
|
0.000
|
Effects of each one of the quantitative variables (levels of learning and skills learning of students, parental participation, and joint activities of parents and children) are studied in the form of three independent test groups (1 and 2) and control groups as well as the interactive effects of these three variables on each other according to the results of Box, Levene, Wilks' Lambda tests. The results are shown in the following table. Based on this table, there are meaningful distinction between the three groups (1 and 2) and control groups in the level of learning and skills learning level of students (FF45 = 975 and P0.05), parental participation rate (F = 483.5115, 0P≤), classroom activities and homework (F = 339.45 and P0.05), word maintenance activity (330 / 114F = and 05 / 0P≤), reading story activity (F = 289243/0 and P0.05 / 24) , Reading poetry (FF = 185.555, P≤0.05) and cartoon perception activity (160 F = 149 and P≤0.05).
Therefore, there is a significant difference between the responses of the test and control groups for the three variables presented in the following table. In other words, the level of learning and student skills, parental participation, and joint activities of parents and children in the experimental groups (1 and 2) and the control group are not similar. Furthermore, the variance of training for participation is %61/7. The variance of students' level of learning and skills is 61/7 percent, parental participation 94.4 percent, homework, and activities 61.4 percent. The vocabulary conservation and learning activity variance is 92.1 percent, poetry 91 percent, and cartoon perception activities 84/9.
Table 6. Tests of Between-Subjects Effects
Dependent Variable
|
R Squared
|
Adjusted R Squared
|
Sum of Squares
|
df
|
Mean Square
|
F
|
Sig.
|
Learning level
|
.617
|
.604
|
26.133
|
2
|
13.067
|
45.975
|
.000
|
Parent participation
|
.944
|
.942
|
50.550
|
2
|
25.275
|
483.511
|
.000
|
Homework
|
.614
|
.600
|
119.233
|
2
|
59.617
|
45.339
|
.000
|
Word
|
.921
|
.918
|
10729.300
|
2
|
5364.650
|
330.114
|
.000
|
Reading the story
|
.910
|
.907
|
1445.200
|
2
|
722.600
|
289.243
|
.000
|
Poem
|
.867
|
.862
|
112.633
|
2
|
56.317
|
185.552
|
.000
|
Cartoon
|
.849
|
.844
|
822.100
|
2
|
411.050
|
160.149
|
.000
|
Now post hoc test is used for understanding if there are differences among these groups. In this study, an LSD follow-up test was used to measure the issue mentioned above (see below table), which is neutral compared to the sample size of the various groups and has the prerequisite for equality of variances. If the difference between the two groups is more than the constant value of LSD, then there is a significant difference between the two groups.
It is important to note two points in performing this test: first, this test should be used when the value of the F is significant in the analysis table of the variance; the second number of groups should not be so many. Therefore, both of these preconditions are true in this research. Second, according to the results of the post hoc test between the experimental group 1 and 2, the experimental group No. 1 and the control group, and the experimental group 2 and the control group, variables' mean of the level of learning and student skills, parental participation, and joint activities of parents and children have significant differences (P0.05). Also, for the lower and upper bound mark, the mean of these three variables was higher for experimental group 1 than control group 2 and control group. In addition, both groups 1 and 2 had a greater mean than the control group.
Table 7. Multiple Comparisons (LSD posthoc test)
Dependent Variable
|
(I) group
|
(J) group
|
Mean Difference (I-J)
|
Std. Error
|
Sig.
|
95% Confidence Interval
|
Lower Bound
|
Upper Bound
|
Learning level
|
Experimental group 1
|
Experimental group 2
|
1.00*
|
0.17
|
.000
|
0.66
|
1.34
|
Control group
|
1.60*
|
0.17
|
.000
|
1.26
|
1.94
|
Experimental group 2
|
Experimental group 1
|
-1.00*
|
0.17
|
.000
|
-1.33
|
-0.66
|
Control group
|
.60*
|
0.17
|
.001
|
0.26
|
0.94
|
Control group
|
Experimental group 1
|
-1.60*
|
0.17
|
.000
|
-1.94
|
-1.26
|
experimental group 2
|
-.60*
|
0.17
|
.001
|
-0.94
|
-0.26
|
Parent participation
|
Experimental group 1
|
experimental group 2
|
.98*
|
0.07
|
.000
|
0.84
|
1.13
|
Control group
|
2.24*
|
0.07
|
.000
|
2.10
|
2.39
|
Experimental group 2
|
Experimental group 1
|
-.98*
|
0.07
|
.000
|
-1.13
|
-0.84
|
Control group
|
1.25*
|
0.07
|
.000
|
1.11
|
1.40
|
Control group
|
Experimental group 1
|
-2.24*
|
0.07
|
.000
|
-2.39
|
-2.10
|
experimental group 2
|
-1.25*
|
0.07
|
.000
|
-1.40
|
-1.11
|
Homework
|
Experimental group 1
|
experimental group 2
|
.95*
|
0.36
|
.011
|
0.22
|
1.68
|
Control group
|
3.35*
|
0.36
|
.000
|
2.62
|
4.08
|
Experimental group 2
|
Experimental group 1
|
-.9500*
|
0.36
|
.011
|
-1.68
|
-0.22
|
control group
|
2.40*
|
0.36
|
.000
|
1.67
|
3.13
|
Control group
|
Experimental group 1
|
-3.35*
|
0.36
|
.000
|
-4.08
|
-2.62
|
Experimental group 2
|
-2.40*
|
0.36
|
.000
|
-3.13
|
-1.67
|
Word
|
Experimental group 1
|
Experimental group 2
|
8.90*
|
1.27
|
.000
|
6.35
|
11.45
|
control group
|
31.75*
|
1.27
|
.000
|
29.20
|
34.30
|
Experimental group 2
|
Experimental group 1
|
-8.90*
|
1.27
|
.000
|
-11.45
|
-6.35
|
control group
|
22.85*
|
1.27
|
.000
|
20.30
|
25.40
|
Control group
|
Experimental group 1
|
-31.75*
|
1.27
|
.000
|
-34.30
|
-29.20
|
Experimental group 2
|
-22.85*
|
1.27
|
.000
|
-25.40
|
-20.30
|
Reading the story
|
Experimental group 1
|
Experimental group 2
|
1.30*
|
0.50
|
.012
|
0.30
|
2.30
|
Control group
|
11.00*
|
0.50
|
.000
|
10.00
|
12.00
|
Experimental group 2
|
Experimental group 1
|
-1.30*
|
0.50
|
.012
|
-2.30
|
-0.30
|
Control group
|
9.70*
|
0.50
|
.000
|
8.70
|
10.70
|
Control group
|
Experimental group 1
|
-11.00*
|
0.50
|
.000
|
-12.00
|
-10.00
|
Experimental group 2
|
-9.70*
|
0.50
|
.000
|
-10.70
|
-8.70
|
Poem
|
Experimental group 1
|
Experimental group 2
|
1.50*
|
0.17
|
.000
|
1.15
|
1.85
|
Control group
|
3.35*
|
0.17
|
.000
|
3.00
|
3.70
|
Experimental group 2
|
Experimental group 1
|
-1.50*
|
0.17
|
.000
|
-1.85
|
-1.15
|
Control group
|
1.85*
|
0.17
|
.000
|
1.50
|
2.20
|
Control group
|
Experimental group 1
|
-3.35*
|
0.17
|
.000
|
-3.70
|
-3.00
|
Experimental group 2
|
-1.85*
|
0.17
|
.000
|
-2.20
|
-1.50
|
Cartoon
|
Experimental group 1
|
Experimental group 2
|
2.95*
|
0.51
|
.000
|
1.94
|
3.96
|
Control group
|
8.90*
|
0.51
|
.000
|
7.89
|
9.91
|
Experimental group 2
|
Experimental group 1
|
-2.95*
|
0.51
|
.000
|
-3.96
|
-1.94
|
Control group
|
5.95*
|
0.51
|
.000
|
4.94
|
6.96
|
Control group
|
Experimental group 1
|
-8.90*
|
0.51
|
.000
|
-9.91
|
-7.89
|
Experimental group 2
|
-5.95*
|
0.51
|
.000
|
-6.96
|
-4.94
|
*the mean difference is significant at the 0.05 level
|
Second question. What kind of relationship is among parents' participation and level of their children's successes while learning a foreign language?
Pearson correlation analysis was used to examine the relationship between the two variables of parental participation and the success rate of children in learning foreign languages by considering these two variables are quantitative. According to this table, the Correlation between the two variables of parents' participation and the success rate of children's learning for the first group was equal to 0/519, for the second group was 0/391. For the third group, it was 0/488, which is This rate is significant concerning the error level of 5 to 10% (P <0.05, P <0.1).
Therefore, it can be concluded that there is a moderate correlation between the two variables of parental participation and the success level of their children's learning in all three groups. Furthermore, it shows a direct and positive relationship meaning that if the parental participation level increases, the success level of students in learning foreign language skills will also improve.
Table 8. Pearsons’ Correlations for the Groups
|
Learning level for Group 1
|
Learning level for Group 2
|
Learning level for Group 3
|
Parent participation
|
Pearson Correlation
|
.519*
|
.391**
|
.488*
|
Sig. (2-tailed)
|
.019
|
.089
|
.029
|
N
|
20
|
20
|
20
|
*Correlation is significant at the 0.05 level, ** Correlation is significant at the 0.1 level
|
Question 3: What is the relationship between language teaching and teaching engagement to parents and child activities in the three groups? (what is the relationship between research variables, and which variable has the most influence on learning degree in children?)
The Partial Least Square (PLS) approach as the second generation of structural equation modeling methods has created new insights for behavioral sciences researchers. Through presenting the latent variable modeling approach, it was concluded that by calculating the measurement error in scales that reduce the estimated relationship, more exact estimates could be obtained on interaction effects.
Two models are examined in PLS:
- outer models
- inner models
The outer model is similar to measuring the confirmatory factor analysis (CFA), and the inner model is similar to path analysis in structural equation modeling. After the outer model test, the inner model indicates the relationship between latent variables in the research. The model research hypotheses can be investigated through the inner model. The PLS analysis is performed in the XLSTAT environment.
PLS model estimation
The present research model is as follows:
yt=fX1, X2,...Xn
Y, the dependent variables, and X, the explanatory variables, are the factors that influence this variable.
The dependent and explanatory variables in the present research are as follows:
explanatory variables:
X1: language teaching
X2: teaching engagement to parents
Dependent variables:
X3: degree of parent involvement
Y1: foreign language learning level
Y2: child activities
Using the PLS, the number of variables in the two components can be categorized.
The following table and graph show the quality of the estimated PLS model based on the number of components selected. When the R2 values for the dependent and independent variables of X and Y and the Q2 values approach 1, the PLS model is estimated more accurately.
Table 9. model efficacy indices with the two components
Index
|
Comp1
|
Comp2
|
Q² cum
|
0.727
|
0.736
|
R²Y cum
|
0.739
|
0.754
|
R²X cum
|
0.749
|
1.000
|
A summary of the PLS model results in the first stage is presented in the following. This graph shows the Correlation between dependent variables and estimated independent variables. These variables are correlated less whenever the axes approach the center of the circle.
As observed, there is a strong positive correlation between the dependent variables (blue points) and the independent variables.
The following table shows the correlation matrix between explanatory variables of the model and each of the estimated components. As observed, there is a relatively strong correlation between the dependent variables of X3 (degree of parent involvement), Y1 (foreign language learning level), and Y2 (child activities) and the first component, namely 0.972, 0.593, and 0.960, respectively.
Table 10. correlation matrix between model variables and each of the estimated components
Variable
|
t1
|
t2
|
X1-0
|
-0.840
|
0.542
|
X1-1
|
0.840
|
-0.542
|
X2-0
|
-0.890
|
-0.457
|
X2-1
|
0.890
|
0.457
|
X3
|
0.972
|
0.020
|
Y1
|
0.593
|
-0.002
|
Y2
|
0.960
|
0.208
|
Based on the results in the following table, variables of X2 (teaching engagement to parents) and X1 (language teaching) have the highest impact on the estimated components, respectively.
In this table, the Upper and Lower bounds represent the acceptable domain of the calculated indices (confidence interval).
Table 11. variables most important in the first two components, respectively
Variable
|
VIP
|
Standard deviation
|
The lower bound(95%)
|
Upper bound(95%)
|
VIP
|
Standard deviation
|
The lower bound(95%)
|
Upper bound(95%)
|
X2-0
|
1.082
|
0.027
|
1.029
|
1.134
|
1.079
|
0.045
|
0.991
|
1.166
|
X2-1
|
1.082
|
0.027
|
1.029
|
1.134
|
1.079
|
0.045
|
0.991
|
1.166
|
X1-0
|
0.911
|
0.032
|
0.849
|
0.973
|
0.915
|
0.053
|
0.810
|
1.019
|
X1-1
|
0.911
|
0.032
|
0.849
|
0.973
|
0.915
|
0.053
|
0.810
|
1.019
|
Given the results delineated in the following table, it can be specified that all independent variables under study (language teaching and teaching engagement to parents) have a significant influence on the dependent variables.
Table 12. estimated coefficients for independent variables of the model and their standard deviations
Variable
|
Coefficient
|
Std. deviation
|
The lower bound (95%)
|
Upper bound (95%)
|
X1-0
|
-0.146
|
0.015
|
-0.177
|
-0.116
|
X1-1
|
0.146
|
0.015
|
0.116
|
0.177
|
X2-0
|
-0.401
|
0.014
|
-0.428
|
-0.375
|
X2-1
|
0.401
|
0.014
|
0.375
|
0.428
|
Regression model estimation
After recognizing essential variables influencing the degree of parent involvement, foreign language learning level, and child activities, regression models are specified to determine the influence of independent variables on the dependent variable. To do the analyses related to the linear regressions, SAS was utilized. Since three different groups (experimental1, experimental2, and control) were investigated in the present study, the influence of the independent variables on the dependent variables was calculated within several separate regression models while considering variables available in each experiment for groups 1 and 2. A summary of the results is presented in the following table. As the regression analysis shows, there is a significant relationship between all variables in the conceptual model of the research (P<0.001).
Table 13. Results of regression analysis with different independent and dependent variables
Model
|
|
RMSE
|
R2
|
X3 = 1.872 + 1.614 (0.81**)X1
|
|
0.57
|
0.65
|
X3 = 1.243 + 1.750 (0.87**)X2
|
|
0.76
|
0.47
|
X3 = 1.243 + 0.985 (0.49**)X1 + 1.258 (0.63**)X2
|
|
0.23
|
0.94
|
Y2 = 66.025 + 36.975 (0.69**)X1
|
|
18.38
|
0.48
|
Y2 = 44.650 + 50.550 (0.95**)X2
|
|
8.04
|
0.90
|
Y2 = 19.060 + 24.608 (0.93**)X3
|
|
9.67
|
0.86
|
Y2 = 44.650 + 15.600 (0.29**)X1 + 42.750 (0.80**)X2
|
|
4.80
|
0.97
|
Y2 = 48.903 - 1.631 (0.06ns)X1 + 9.362 (0.69**)X3
|
|
9.95
|
0.42
|
Y2 = 31.213 + 31.622 (0.59**)X2 + 10.814 (0.41**)X3
|
|
6.31
|
0.94
|
Y2 = 51.678 + 21.168 (0.40**)X1 + 49.865 (0.94**)X2 – 5.656 (0.21*)X3
|
|
4.67
|
0.97
|
Y1 = 66.425 + 13.475 (0.50**)X1
|
|
11.21
|
0.25
|
Y1 = 61.450 + 14.200 (0.53**)X2
|
|
11.00
|
0.28
|
Y1 = 49.940 + 8.706 (0.65**)X3
|
|
9.87
|
0.42
|
Y1 = 47.756 + 0.296 (0.58**)Y2
|
|
10.51
|
0.34
|
Y1 = 61.450 + 8.500 (0.31*)X1 + 9.950 (0.37**)X2
|
|
10.51
|
0.35
|
Y1 = 48.903 - 1.631 (0.06ns)X1 + 9.362 (0.69**)X3
|
|
9.95
|
0.42
|
Y1 = 51.136 + 4.912 (0.18ns)X1 + 0.232 (0.46**)Y2
|
|
10.46
|
0.36
|
Y1 = 48.253 - 4.390 (0.16ns)X2 + 10.622 (0.79**)X3
|
|
9.91
|
0.42
|
Y1 = 42.287 - 7.494 (0.28ns)X2 + 0.429 (0.85 *)Y2
|
|
10.54
|
0.35
|
Y1 = 50.93 – 9.987 (0.74**)X3 - 0.052 (0.10ns)Y2
|
|
9.95
|
0.42
|
Y1 = 40.168 - 8.363 (0.31ns)X1 – 11.598 (0.43ns)X2 + 17.128 (1.27**)X3
|
|
9.85
|
0.44
|
Y1 = 51.845 + 5.144 (0.19ns)X1 + 0.754 (0.03ns)X2 + 0.215 (0.42ns)Y2
|
|
10.55
|
0.36
|
Y1 = 46.113 – 6.558 (0.24ns)X2 + 9.880 (0.73**)X3 + 0.069 (0.14ns)Y2
|
|
9.99
|
0.43
|
Y1 = 49.842 – 2.148 (0.08ns)X1 + 11.206 (0.83*)X3 - 0.066 (0.13ns)Y2
|
|
10.02
|
0.42
|
Y1 = 16.045 - 18.244 (0.68*)X1 – 34.875 (1.29*)X2 + 19.769 (1.47**)X3 + 0.467 (0.92ns)Y2
|
|
9.69
|
0.47
|
Notes: The numbers in parentheses are standardized regression coefficients; ns P > 0.05; * P ≤ 0.05; ** P ≤ 0.01
In order to recognize the most important variables influencing the learning level in students, standard regression coefficients related to the relationship between each pair of the research variables are presented in the following graph. Results showed that from among variables explaining the degree of parent involvement, the variable of teaching engagement to parents was the most effective (P<0.001; β=0.87), followed by the variable of language teaching (P<0.001; β=0.81). Moreover, the results showed that from among variables predicting child activities, teaching engagement to parents (engagement <0.001; β=0.95), parent involvement (parent involvement<0.001; β=0.93), language teaching (Language teaching <0.001; β=0.69) had the most influence respectively.
It was interesting that the direct influence of language teaching (β=0.69) on child activities was less than its indirect influence of parent involvement degree (parent involvement) (0.81×0.93=0.75). Compared to the direct influence, language teaching through the variable of parent involvement degree had more influence on child activities. Finally, parent involvement (parent involvement <0.001; β=0.65), child activities (activities<0.001; β=0.58), and language teaching (language teaching<0.001; β=0.50) had the most influence on the dependent variable of foreign language learning level respectively.
Question number 4. What kind of relationship is there among students' and parents' actions with the level of foreign language learning in three experimental groups?
Pearson correlation test was used to answer this question. According to the table below, the correlation coefficient ® between the two variables of parental participation and the success rate of students learning for the first group was 0.384, which is significant at a 10% error level (P-value <0.1). Therefore, it can be concluded that there is a weak correlation between students' and their parents' activities and learning in the first group. Furthermore, this is a direct and positive relationship, meaning that the higher the parents' and students' activities, the greater the students' success in learning foreign language skills. However, this relationship was not significant for the second experimental and control groups (P-value> 0.5).
Table 14. Pearsons’ Correlations for the Groups
|
Learning level for Group 1
|
Learning level for Group 2
|
Learning level for Group 3
|
Activities
|
Pearson Correlation
|
0.384
|
0.209
|
-0.250
|
Sig. (2-tailed)
|
0.095
|
0.377
|
0.288
|
N
|
20
|
20
|
20
|
* Correlation is significant at the 0.1 level
|
For better representation of this figure, the MONOVA was recalculated by computing the mean of activities shown in this figure.
Table 15. Tests of Between-Subjects Effects
Tests of Between-Subjects Effects
|
Source
|
Dependent Variable
|
Type III Sum of Squares
|
Df
|
Mean Square
|
F
|
Sig.
|
Corrected Model
|
Learning level
|
26.133a
|
2
|
13.067
|
45.975
|
.000
|
Parent participation
|
50.550b
|
2
|
25.275
|
483.511
|
.000
|
activities
|
36504.300c
|
2
|
18252.150
|
790.947
|
.000
|
Intercept
|
Learning level
|
481.667
|
1
|
481.667
|
1694.753
|
.000
|
Parent participation
|
348.348
|
1
|
348.348
|
6663.951
|
.000
|
activities
|
368323.350
|
1
|
368323.350
|
15961.099
|
.000
|
group
|
Learning level
|
26.133
|
2
|
13.067
|
45.975
|
.000
|
Parent participation
|
50.550
|
2
|
25.275
|
483.511
|
.000
|
activities
|
36504.300
|
2
|
18252.150
|
790.947
|
.000
|
Error
|
Learning level
|
16.200
|
57
|
.284
|
|
|
Parent participation
|
2.980
|
57
|
.052
|
|
|
activities
|
1315.350
|
57
|
23.076
|
|
|
Total
|
Learning level
|
524.000
|
60
|
|
|
|
Parent participation
|
401.878
|
60
|
|
|
|
activities
|
406143.000
|
60
|
|
|
|
Corrected Total
|
Learning level
|
42.333
|
59
|
|
|
|
Parent participation
|
53.529
|
59
|
|
|
|
activities
|
37819.650
|
59
|
|
|
|
a. R Squared = .617 (Adjusted R Squared = .604)
|
b. R Squared = .944 (Adjusted R Squared = .942)
|
c. R Squared = .965 (Adjusted R Squared = .964)
|
Figurative display of results
In the following figures, the results of MANOVA and correlation analysis are summarized. The numbers on the line of chains are results of the MANOVA test, and the numbers on the full lines represent the results of the correlation test