You are on page 1of 10

Analysis

Hypotheses:

H0: There is no statistically significant relationship between student satisfaction based on


traditional learning and that of understanding of subject, faculty interaction, team work,
convenience, academic stress and presentation.

H1: There is statistically significant relationship between student satisfaction based on


traditional learning and that of understanding of subject, faculty interaction, team work,
convenience, academic stress and presentation.

 Dependent variable: student satisfaction


 Independent variables: understanding of subject, faculty interaction, team work,
convenience, academic stress and presentation

Descriptive Statistics
Std.
Mean Deviation N
Overall academic 3.67 1.219 145
satisfaction
with offline class
Understanding of the 3.5052 .82626 145
subject
Faculty 3.8897 .84105 145
Interaction
Convenience 4.0690 .91500 145
Team 3.9052 .84578 145
Work
Academic 2.5931 .88729 145
Stress
Presentation 3.8241 .94102 145
Correlations
Overall
academi
c
satisfacti
on Tea
with Understand Faculty m Acade
offline ing of the Interacti Convenie Wor mic Presentat
class subject on nce k Stress ion
Pearson Overall 1.000 .574 .470 .472 .484 -.483 .683
Correlati academic
on satisfaction
with
offline
class
Understand .574 1.000 .654 .284 .519 -.512 .606
ing of the
subject
Faculty .470 .654 1.000 .435 .557 -.487 .601
Interaction
Convenien .472 .284 .435 1.000 .606 -.462 .563
ce
Team .484 .519 .557 .606 1.00 -.619 .646
Work 0
Academic -.483 -.512 -.487 -.462 -.61 1.000 -.548
Stress 9
Presentatio .683 .606 .601 .563 .646 -.548 1.000
n
Sig. (1- Overall . .000 .000 .000 .000 .000 .000
tailed) academic
satisfaction
with
offline
class
Understand .000 . .000 .000 .000 .000 .000
ing of the
subject
Faculty .000 .000 . .000 .000 .000 .000
Interaction
Convenien .000 .000 .000 . .000 .000 .000
ce
Team .000 .000 .000 .000 . .000 .000
Work
Academic .000 .000 .000 .000 .000 . .000
Stress
Presentatio .000 .000 .000 .000 .000 .000 .
n
N Overall 145 145 145 145 145 145 145
academic
satisfaction
with
offline
class
Understand 145 145 145 145 145 145 145
ing of the
subject
Faculty 145 145 145 145 145 145 145
Interaction
Convenien 145 145 145 145 145 145 145
ce
Team 145 145 145 145 145 145 145
Work
Academic 145 145 145 145 145 145 145
Stress
Presentatio 145 145 145 145 145 145 145
n
Variables Entered/Removeda
Variables Variables
Model Entered Removed Method
1 Presentation, . Enter
Academic
Stress,
Convenience,
Faculty
Interaction,
Understandin
g of the
subject, Team
Workb
a. Dependent Variable: Overall academic
satisfaction
with offline class
b. All requested variables entered.

1. EXPLANATORY POWER:

Model Summaryb
Adjusted R Std. Error of Durbin-
Model R R Square Square the Estimate Watson
a
1 .728 .530 .509 .854 2.036
a. Predictors: (Constant), Presentation, Academic
Stress, Convenience, Faculty
Interaction, Understanding of the subject, Team
Work
b. Dependent Variable: Overall academic satisfaction
with offline class

Adjusted R Square value is 50.9%.


It means 50.9% of variance in student satisfaction can be explained by understanding of
subject, faculty interaction, team work, convenience, academic stress and presentation.
It is a good score as far as social science is concerned.
2. GOODNESS OF FIT:

ANOVAa
Sum of
Model Squares df Mean Square F Sig.
1 Regression 113.391 6 18.899 25.894 .000b
Residual 100.719 138 .730
Total 214.110 144
a. Dependent Variable: Overall academic satisfaction
with offline class
b. Predictors: (Constant), Presentation, Academic
Stress, Convenience, Faculty
Interaction, Understanding of the subject, Team
Work

This table indicates that the regression model predicts the dependent variable significantly
well. The Sig value is 0.000.Here, p < 0.0005, which is less than 0.05, and indicates that,
overall, the regression model statistically significantly predicts the outcome variable (i.e., it is
a good fit for the data).

3. REGRESSION COEFFICIENTS:

Coefficientsa
Unstandardized Standardized Collinearity
Coefficients Coefficients Statistics
Std.
Model B Error Beta t Sig. Tolerance VIF
1 (Constant) .143 .753 .189 .850
Understanding of .425 .128 .288 3.331 .001 .455 2.198
the subject
Faculty -.092 .123 -.063 -.745 .458 .474 2.110
Interaction
Convenience .227 .105 .171 2.157 .033 .545 1.836
Team -.119 .131 -.082 -.903 .368 .411 2.434
Work
Academic -.123 .109 -.090 - .258 .546 1.832
Stress 1.136
Presentation .588 .118 .454 5.004 .000 .414 2.413
a. Dependent Variable: Overall academic satisfaction
with offline class

 The Coefficients table provides us with the necessary information to predict student
satisfaction from understanding of subject, faculty interaction, team work, convenience,
academic stress and presentation as well as determine whether understanding of subject,
faculty interaction, team work, convenience, academic stress and presentation contributes
statistically significantly. (By referring to the ‘Sig.’ column)
 The table predicts that for every unit increase in Presentation .588 unit increase in student
satisfaction is predicted. For every unit increase in Faculty Interaction, there is a 0.92 unit
decrease in the predicted student satisfaction. Similar prediction can be done for other
variables as well.
 The coefficient for understanding of subject (0.425) is statistically significant using
alpha of 0.05 because its p-value is 0.001, which is smaller than 0.05.

The coefficient for faculty interaction (-0.092) is not statistically significant at the 0.05 level
since the p-value is 0.458, which is greater than .05.

The coefficient for team work (-0.119) is not statistically significant using alpha of 0.05
because its p-value is 0.368, which is greater than 0.05.

The coefficient for convenience (0.227) is statistically significant because its p-value of
0.033 which is less than 0.05.
The coefficient for academic stress (-0.123) is not statistically significant because its p-value
of 0.258 which is greater than 0.05.

The coefficient for presentation (0.588) is statistically significant because its p-value of
0.000 which is lesser than 0.05

Data should meet certain assumptions, in order for the analysis to be


reliable and valid. It includes:
 Assumption #1: The relationship between the IVs and the DV is linear.
 Assumption #2: There is no multicollinearity in the data.
 Assumption #3: The values of the residuals are independent.
 Assumption #4: The variance of the residuals is constant (homoscedasticity).
 Assumption #5: The values of the residuals are normally distributed.
Collinearity Diagnosticsa
Variance Proportions
Understan Tea
Condit ding of Faculty m Acade
Mo Dimen Eigenv ion (Const the Interac Conveni Wo mic Presenta
del sion alue Index ant) subject tion ence rk Stress tion
1 1 6.725 1.000 .00 .00 .00 .00 .00 .00 .00
2 .179 6.122 .00 .01 .00 .00 .00 .24 .01
3 .039 13.185 .00 .24 .05 .31 .02 .00 .00
4 .019 18.685 .01 .00 .21 .03 .04 .03 .88
5 .017 20.063 .02 .07 .45 .12 .50 .00 .04
6 .014 21.898 .00 .57 .28 .46 .24 .01 .07
7 .007 30.755 .98 .11 .01 .07 .19 .71 .00
a. Dependent Variable: Overall academic satisfaction
with offline class

Residuals Statisticsa
Minimu Maximu Std.
m m Mean Deviation N
Predicted Value .62 5.17 3.67 .887 145
Residual -3.389 2.289 .000 .836 145
Std. Predicted -3.432 1.695 .000 1.000 145
Value
Std. Residual -3.966 2.680 .000 .979 145
a. Dependent Variable: Overall academic satisfaction
with offline class

Assumption #1: The relationship between the IVs and the DV is linear.
It’s linear. Because most of data points laying linearly.
Assumption #2: There is no multicollinearity in the data.

We can test this assumption by looking at the Coefficients table. This allows us to formally
check that our predictors (or IVs) are not too highly correlated. We can use Variance Inflation
Factor (VIF) and Tolerance statistics to assess this assumption.
Coefficientsa
Unstandardized Standardized Collinearity
Coefficients Coefficients Statistics
Std.
Model B Error Beta t Sig. Tolerance VIF
1 (Constant) .143 .753 .189 .850
Understanding of .425 .128 .288 3.331 .001 .455 2.198
the subject
Faculty -.092 .123 -.063 -.745 .458 .474 2.110
Interaction
Convenience .227 .105 .171 2.157 .033 .545 1.836
Team -.119 .131 -.082 -.903 .368 .411 2.434
Work
Academic -.123 .109 -.090 - .258 .546 1.832
Stress 1.136
Presentation .588 .118 .454 5.004 .000 .414 2.413
a. Dependent Variable: Overall academic satisfaction
with offline class
Conclusion: For the assumption to be met we want VIF scores to be well below 10, and
tolerance scores to be above 0.2; which is the case in this example. Thus, there is no
multicollinearity in our data.

Assumption #3: The values of the residuals (error in our model) are independent

To check the next assumption we need to look at is the Model Summary box. Here, we can
use the Durbin-Watson statistic to test the assumption that the residuals are independent.
This statistic can vary from 0 to 4. For assumption #3 to be met, we want this value to be in
between 1.5 to 2.5. Values below 1 and above 3 are cause for concern and may render our
analysis invalid.

Model Summaryb
Adjusted R Std. Error of Durbin-
Model R R Square Square the Estimate Watson
1 .728a .530 .509 .854 2.036
a. Predictors: (Constant), Presentation, Academic
Stress, Convenience, Faculty
Interaction, Understanding of the subject, Team
Work
b. Dependent Variable: Overall academic satisfaction
with offline class
Conclusion: In this case, the value is 2.036; the Durbin-Watson statistic showed that this
assumption had been met, as the obtained value was close to 2.
Assumption #4: The variance of the residuals is constant (homoscedasticity).
Homoscedasticity means “having the same scatter.” the points must be about the same
distance from the line. The opposite is heteroscedasticity (“different scatter”), where points
are at widely varying distances from the regression line.
From above scatter plot, where some data points are at widely varying distances from the
regression line. Therefore this model having less Homoscedasticity.

Assumption #5: The values of the residuals are normally distributed

This assumption can be tested by looking at the P-P plot for the model. The closer the dots lie
to the diagonal line, the closer to normal the residuals are distributed. In this case, our data
points touch the line.

REGRESSION EQUATION OF THIS MODEL


For standardized coefficients of beta: 0 + (0.288) x1 - (0.063) x2 + (0.171) x3 -(0.082) x4 –
(0.090)x5+ 0.454 x6
For unstandardized coefficients of beta: 0.143 + (0.425) x1 - (0.092) x2 + (0.227) x3 -
(0.119) x4 –(0.123)x5- 0.588 x6

You might also like