Page 33 - ASBIRES-2017_Preceedings

Rahubaddha & Panahatipola

Table 1: Advanced analysis table As in table 1, value of the model
2
p-values is 0.64; about 64% of total variation in the
Coefficient P-Value VIF GPA value of the students can be explained
by the model.
(Constant) 2.747 0
Involve in We can build final regression
Sport 0.023 0.004 1.039 model for the GPA value of the third-year
Participatio students in faculty of applied sciences.
n of
informal Fitted Model Was,
revision -0.014 0.039 1.132
class GPA Value = 2.747 + 0.023(Good
conducted Foundation in first Year) -
by Students 0.014(Participation to Informal Revision
Have Good Classes Conducted by Students) +
Foundation 0.07 0.002 1.103
in 1st Year 0.023(Involve in Sport) - 0.098(Total
Total -0.098 0 1.128 Number of repeats)
repeats
6 CONCLUSION
Model Summary
R
Model R When considering 3rd year
Square undergraduates there were four factors
0.8 0.64 which related to the GPA value.

ANOVA Table Foundation got in the first year,
participation for the informal revision
Sum of classes conducted by students; total
df Sig.
Squares repeats and involvement in sport were
identified factors which were related
directly to the GPA value of third year
b
Regression 18.799 4 .000
undergraduates in faculty of applied
Residual 10.562 134
sciences. Therefore, we can come to the
Total 29.361 138 conclusion that these factors are the most

important factors that should be
As in Table 1, Enter regression is considered out of all other factors which
used to identify a suitable model for affect the GPA value of the 3 year
rd
predicting GPA value using the factors students. From those four factors number
which we identified as having impact on of repeaters has high impact for the GPA
GPA. According to the P-values, Good value.
Foundation in first Year, Participation in
the informal revision classes conducted Factors like using the library
by students, Involve in Sports and Total facilities, number of times sat for the A/L
Number of repeats have become as examination, no sisters and brothers and
statistically significant at 5% significance place of stay have no significant impact
level. on the GPA according to the preliminary
analysis of third year undergraduates in
Since the P-value of the final the Faculty of Applied Sciences.
model is less than 0.05, that model is
useful in predicting the response variable. The preliminary analysis implies us
Considering table 1, VIF values all values the conclusion that the Gender of the
are less than 5 suggesting that student has a relation to the GPA directly
multicollinearity does not exist in this however in the regression model Gender
model. is not included it as a significant factor to
predict GPA. Similar thing happens in the

23