Case Study 2
Khaled Al Akrah
1- The dealership’s management believes that customers’ place of residence
may influence their concern about fuel consumption. Determine if customers
from different cities have differing concerns about fuel consumption.
Dependent and independent variables: dependent variable is fuel and the
independent variable is the place of residence.
H0: 𝜇fuel_beirut = 𝜇fuel-aley
Hypothesis:
H1: 𝜇fuel_beirut ≠ 𝜇fuel_aley
Test Statistics:
First, we use the Q-Q plot (Analyze -> Descriptive statistics) to check for
normality. Utilizing the fuel in the dependant list, create Q-Q charts. The
figure shows that the GPA has a distribution that is more similar to the
normal distribution. We may think about parametric tests since we can
assume normal distribution. We must now determine whether it is an
independent t-test, an ANOVA, or a paired t-test. We may apply an
independent t-test since we are comparing the means of two independent
samples, the independent variable is binary (a categorical variable with two
values), and the dependent variable is scale.
Finally, Analyze, Compare Means, and Independent-Samples T Test. Select
the city as the grouping variable and fuel consumption as the test variable.
Select the first and second groups you want to use from the define groups
buttons. The confidence coefficient, which is set by default to 95%, can be
modified from the settings menu. It will remain unchanged.
Interpret the results:
Levene's Test
for Equality of t-test for Equality of Means
Variances
95%
Confidenc
Sig. (2- Mean Std. Error
F Sig. t df e Interval lower upper
tailed) Difference Difference
of the
Difference
Equal
variances 4.728 0.043 -0.303 18 0.766 -0.119 0.393 -0.945 0.707
assumed
Equal
variances
-0.374 15.946 0.713 -0.119 0.318 -0.794 0.556
not
assumed
Use the Levene's test to determine whether you can use the second line in
the t-test if the significance level is less than 0.05. If not, we apply the first
line. In this instance, it is less than 0.05. (0.043). We can therefore utilize the
second line.
We can assume equal means if the t-test significance is larger than 0.05.
However, since our case's significance is substantially higher than 0.05
(0.713), we have reason to believe that H1 cannot be true. Furthermore, the
confidence interval shows that, if the same experiment were repeated with
different samples chosen based on the same criteria, we would have a 95%
confidence that the average fuel consumption of Beirut (1) is higher than the
average fuel consumption of Aley (2) by a value between 0.556 and 0.794.
There is sufficient data to conclude that the average fuel consumption
difference between Beirut and Aley is approximately 0.119 points ("Mean
Difference"). The difference in consumption between Beirut and Aleem,
with a 95% confidence interval, is between 0.556 and 0.794 points ("Upper"
& "Lower").
2. Determine if customers are more concerned about safety or price,
when deciding which car to buy.
Dependent and independent variables:
Brand is the dependent variable, and price and safety concerns are the
independent variables. Since there are more than two samples, we must
perform a "Two-Way ANOVA" test.
Test Statistics:
First, Place price on the Horizontal axis of Plots and safety on the Separate
lines, then click Add and Continue. Second, Move safety to Horizontal
axis and price to Separate lines since more under Plots, then click Add and
Continue. Third, Move safety to Post Hoc Tests for under Post Hoc, then
click Continue after choosing Tukey under Equal Variances Assumed.
Finally, Select Descriptive statistics under Options, then click Continue and
OK.
Interpret the results:
Sig. of price*safety: there is no statistically significant interaction between
price and safety because the value of price*safety is 0.623>0.05.
Sig of price: is 0.0000.05, indicating a statistically significant difference
between the averages of yearly car purchases among the prices.
Sig of safety: is 0.2811>0.05, indicating that there is no statistically
significant difference between the methods for purchasing a car among those
who are concerned about safety.
Tests of Between-Subjects Effects
Type III Sum
Source of Squares df Mean Square F Sig.
Corrected Model 17.300a 9 1.922 7.689 .002
Intercept 86.282 1 86.282 345.128 .000
Concern_for_safety .722 2 .361 1.445 .281
Concern_for_price 15.618 3 5.206 20.824 .000
Concern_for_safety *
.678 4 .170 .678 .623
Concern_for_price
Error 2.500 10 .250
Total 108.000 20
Corrected Total 19.800 19
3. The dealership’s management believes that customers’ concern for
safety may differ according to the customers’ ages.
Dependent and independent variables:
We must perform a "one-way ANOVA" test because there are more than
two samples, the age is the independent variable, and the safety concern
(scale) is the dependent variable.
Test statistics:
Move safety concern to the "Dependent List" and age to the "Factor" after
performing Analyze, Compare Means, and One-Way ANOVA. Choose
"Descriptive" from the "Options" menu, then click "Continue" and "OK."
Interpret the results:
There is no statistically significant difference in the means of safety concern
when compared with age because the significance level "Sig." is 0.281>0.05.
ANOVA
Sum of
Squares df Mean Square F Sig.
Between Groups 2.443 3 .814 1.392 .281
ANOVA
Sum of Mean
Squares df Square F Sig.
Within Groups 9.357 16 .585
Total Between 11.800 17.53019 3 5.843 20.914 .000
Groups
Within Groups 4.470 16 .279
Total 22.000 19
4. Management believes that older customers are wealthier and less
concerned about price than younger customers.
Dependent and independent variables:
We must conduct a "one-way ANOVA" test because there are more than
two samples, age is the independent variable, and price concern is the
dependent variable.
Test statistics:
Move age to the Factor list and price concern to the Dependent List after
selecting Analyze, Compare Means, and One-Way ANOVA. Select
Descriptive under Options, then click Continue and OK.
Interpret the results:
The statistical difference between the means of safety concern when
compared with age is statistically significant (Sig. = 0.000>0.05).
5. Management believes that the ages of customers are not associated
with the brand of cars they prefer to buy.
Dependent and independent variables:
We must perform a "one-way ANOVA" test because the brand is the
dependent variable, the age is the independent variable, and we have more
than two samples.
Test statistics:
Move the brand to the Dependent List and the age to the Factor after
selecting Analyze, Compare Means, One-Way ANOVA. Select Descriptive
under Options, then click Continue and ok.
Interpret the results:
There is no statistically significant difference between the means of brand
when compared with age because the significance level is 0.0240.05.
Anova
Sum of Mean
Squares df Square F Sig.
Between
5.000 1 5.000 6.081 .024
Groups
Within
14.800 18 .822
Groups
Total 19.800 19