Professional Documents
Culture Documents
Business Analytics (MGT 555) Individual Assignment Assignment 2
Business Analytics (MGT 555) Individual Assignment Assignment 2
(MGT 555)
INDIVIDUAL ASSIGNMENT
ASSIGNMENT 2
Prepared by:
NURUL SYAKIRAH BINTI MOHD HUSNI
2018263578
BA232 2A
Prepared for:
DR. NORSHIMA BINTI HUMAIDI
QUESTION 1
Based on the data student grades:
20
0
-20 50 60 70 80 90 100 110
-40
Final Exam
Based on the result findings from the data student grade, the data in the figure 1 above
shows the key regression results that include regression statistics, ANOVA and the
coefficients of the student grades. Based on the hypothesis, the H0: 𝛽1 is equal to 0 student
grades has no effect on final exam and the H1: 𝛽1 is not equal to 0 student grades has
significant effect on final exam. According to the result findings in the regression statistics, it
shows the R-Square is 0.300622 which is equals to 30% meanwhile, the standard error is
9.431160618. This indicates that 30% of variation in final exam is explained by midterm
while the remaining 70% is explained by other variables. This indicates that the model is not
in good fit.
Next, according to the ANOVA, the p-value indicate the value of (0.000) for the F-
test (23.2114906) which is less than the level of significance (0.05). Therefore, the null
hypothesis is rejected. Meaning which, final exam is statistically significant variable in
explaining the variation in student. Moreover, the p-value (0.000) for final exam regression
coefficient is less than 0.005. While, the β 1 is equals to 0.486, which is between the lower
value (0.283) and the upper value (0.688). Therefore, we can conclude that the coefficient is
statistically not equal to zero and the null hypothesis is rejected. Based on the figure 3 below,
it shows the residual analysis, the result indicates that there is no outliers appear to exist
because all the standard residuals value is less than negative or positive 3.
Based on the result findings from the data home price, the data in the figure 4 above
shows the multiple regression model that include regression statistics, ANOVA and the
coefficients of the home price. Based on the hypothesis, the hypothesis 1 is beds is positively
influence home price meanwhile, hypothesis 2 is sqft and baths is positively influence home
price. Based on to the result findings, this model also shown that there is no multicollinearity
issue because the association between Sqft and Beds is below 0.7 which are (0.530) and
(0.284). In the regression statistics, it shows the R-Square is 0.72367178 which is equals to
72% of variation in home price explained by sqft, baths and beds. Meanwhile, the standard
error is 74984.98417. This indicates that 72% of variation in home price indicates that the
model is on good fit.
On the other hand, the p-value (0.000) for the F-test is less than the level of
significance (0.05). For the first model of both sqft and baths shows that the p-value is
0.010779404 and 0.000854652. We can conclude that that sqft and baths is significantly
influence the home price where the p-value is less than 0.05. Meanwhile, the beds p-value is
0.9782636 which is more than 0.05 was not shown significant compare to others variable. So,
the beds need to eliminate in this case.
Figure 5: Best Model
Based on the figure 5 above, it shows the best model after eliminate the beds variable
since the variable was not significant. In conclusion, the best model for this case is eliminate
the one factors mentioned above which is beds. Thus, the model should has only sqft and
baths variable which is shown significantly influence home price.