In: Statistics and Probability
A study was performed to examine whether the type of gasoline used is independent of a person's living status. The researcher surveyed 600 people and asked two questions. The first question asked was “Do you use regular, plus, or super gasoline when filling up?” The second question asked was “Do you rent or own your place of residence?” The results are shown below. Assume the significance level α = .05 is used for the test.
Type of Gas |
Living Status |
Regular |
Plus |
Super |
Rent |
173 |
68 |
49 |
Own |
154 |
87 |
69 |
Under the null hypothesis, the expected number of people to rent
and use regular gasoline is 158.05 and the expected number of
people to own and use plus gasoline is 80.08. The numerical value
of the chi-square test statistic is:
Observed Frequencies | ||||
Regular | Plus | Super | Total | |
Rent | 173 | 68 | 49 | 290 |
Own | 154 | 87 | 69 | 310 |
Total | 327 | 155 | 118 | 600 |
Expected Frequencies | ||||
Regular | Plus | Super | Total | |
Rent | 327 * 290 / 600 = 158.05 | 155 * 290 / 600 = 74.9167 | 118 * 290 / 600 = 57.0333 | 290 |
Own | 327 * 310 / 600 = 168.95 | 155 * 310 / 600 = 80.0833 | 118 * 310 / 600 = 60.9667 | 310 |
Total | 327 | 155 | 118 | 600 |
(fo-fe)²/fe | ||||
Rent | (173 - 158.05)²/158.05 = 1.4141 | (68 - 74.9167)²/74.9167 = 0.6386 | (49 - 57.0333)²/57.0333 = 1.1315 | |
Own | (154 - 168.95)²/168.95 = 1.3229 | (87 - 80.0833)²/80.0833 = 0.5974 | (69 - 60.9667)²/60.9667 = 1.0585 |
Test statistic:
χ² = ∑ ((fo-fe)²/fe) = 6.1630