In: Statistics and Probability
A large gasoline distributor claims that 65% of all vehicle owners who use its service stations choose regular unleaded gas, 20% choose mid-grade unleaded gas, and 15% choose premium unleaded gas. To investigate this claim, researchers collected data from a random sample of drivers who filled up their vehicles at the distributor's service stations in a large city. The results were 305 regular, 121 mid-grade, and 74 premium gas purchases. Are the data from the sample consistent with the distributor's claim? Conduct an appropriate statistical test at the 5% significance level to support your conclusion. Make sure to include parameters, check conditions, and show calculations before formulating a conclusion.
null hypothesis:Ho: Sample proportion of different fuel is same as claimed by distributor
Alternate hypothesis:Ha: Sample proportion of different fuel is different from as claimed by distributor
degree of freedom =categories-1= | 2 |
for 2 df and 0.05 level of signifcance critical region χ2= | 5.991 |
Applying chi square goodness of fit test:
relative | observed | Expected | residual | Chi square | |
category | frequency | Oi | Ei=total*p | R2i=(Oi-Ei)/√Ei | R2i=(Oi-Ei)2/Ei |
Unleaded | 0.65 | 305 | 325.00 | -1.11 | 1.231 |
Mid-grade | 0.20 | 121 | 100.00 | 2.10 | 4.410 |
Premium | 0.15 | 74 | 75.00 | -0.12 | 0.013 |
total | 1.000 | 500 | 500 | 5.654 |
as test statistic 5.654 does not fall in rejection region; we can not reject null hypothesis
we do not have evidence to conclude that Sample proportion of different fuel is different from as claimed by distributor