In: Economics
Suppose there is a task at hand to check whether one can really save a bunch of money on the car insurance by switching from Allstate to GEICO. Their monthly insurance premiums for various car brands are shown in the table below. Conduct a two-sample test of hypothesis at α = 0.05 significance level to determine whether it is worth switching to GEICO.
Car Brands | GEICO | Allstate |
Honda | 120 | 136 |
Toyota | 125 | 135 |
Nissan | 122 | 135 |
Ford | 107 | 102 |
Chevrolet | 126 | 128 |
Chrysler | 110 | 104 |
BMW | 154 | 176 |
Mercedes | 160 | 182 |
Volkswagen | 148 | 164 |
We have the following information
Average Premium of GEICO = (1,172 ÷ 9) = 130.2
Average Premium of Allstate = (1,262 ÷ 9) = 140.2
Car Brands |
GEICO |
Allstate |
Deviation from Mean (GEICO) |
Square of Deviation from Mean (GEICO) |
Deviation from Mean (Allstate) |
Square of Deviation from Mean (Allstate) |
Honda |
120 |
136 |
-10.2 |
104.5 |
-4.2 |
17.8 |
Toyota |
125 |
135 |
-5.2 |
27.3 |
-5.2 |
27.3 |
Nissan |
122 |
135 |
-8.2 |
67.6 |
-5.2 |
27.3 |
Ford |
107 |
102 |
-23.2 |
539.3 |
-38.2 |
1460.9 |
Chevrolet |
126 |
128 |
-4.2 |
17.8 |
-12.2 |
149.4 |
Chrysler |
110 |
104 |
-20.2 |
408.9 |
-36.2 |
1312.0 |
BMW |
154 |
176 |
23.8 |
565.4 |
35.8 |
1280.0 |
Mercedes |
160 |
182 |
29.8 |
886.7 |
41.8 |
1745.4 |
Volkswagen |
148 |
164 |
17.8 |
316.0 |
23.8 |
565.4 |
Total |
1172 |
1262 |
2933.6 |
6585.6 |
Standard Deviation of GEICO = √(Square of Deviation from Mean)÷(N – 1); where N is the number of observations
Standard Deviation of GEICO = √(2933.6 ÷ 8) = 19.1
Standard Deviation of Allstat = √(Square of Deviation from Mean)÷(N – 1); where N is the number of observations
Standard Deviation of Allstate = √(6585.6 ÷ 8) = 28.7
Null Hypothesis: There is no significant difference between the average premium of GEICO and Allstate
T-test for mean comparison with unequal standard deviations
T-test statistic = (Mean of GEICO – Mean of Allstate) ÷ √((Standard Deviation of GEICO)2/N1) + ((Standard Deviation of Allstat)2/N2)
N1 and N2 are number of observations. We will be using 95% confidence level
T Stat |
-0.71466 |
P(T<=t) one-tail |
0.244245 |
t Critical one-tail |
1.782288 |
P(T<=t) two-tail |
0.488489 |
t Critical two-tail |
2.178813 |
Since, the calculated T-state value (-0.71466) is less than the critical value (2.178813), so we accept the null hypothesis that there is no significant difference between the average premium of GEICO and Allstate. So, one can say that it is not worth switching to GEICO.