In: Statistics and Probability
A sample of 340 randomly selected people in their 20s were asked about their favorite place to go on a Friday night. The results are given in
the table, together with a person’s gender:
Movies Sports game Restaurant Shopping Concert
Total
Men 43 50 46 19 42
200
Women 29 21 31 31 28
140
Total 72 71 77
50 70 340
(a) [1] Give the marginal distribution of gender.
(b) [2] Find the conditional distribution of preferred activity for Women.
Test at a 5% level of significance whether the preferred activity depends on the gender.
(c) [1] State the null and alternative hypotheses.
(d) [1] Calculate the test statistic.
(e) [1] Find the corresponding p-value.
(f) [1] Make the decision according to (e).
(g) [1] Give the conclusion by the decision made from (f).
a)
marginal distribution of gender:
Men : 200 / 340 = 0.5882
Women : 140 / 340 = 0.4118
b) Conditional distribution of preferred activity for Women.
Sports | Game | Restaurant | Shopping | Concert | |
Women | 29 / 140 = 0.2071 | 21 / 140 = 0.15 | 31 / 140 = 0.2214 | 31 / 140 = 0.2214 | 28 / 140 = 0.20 |
c)
Observed Frequencies | ||||||
Sports | Game | Restaurant | Shopping | Concert | Total | |
Men | 43 | 50 | 46 | 19 | 42 | 200 |
Women | 29 | 21 | 31 | 31 | 28 | 140 |
Total | 72 | 71 | 77 | 50 | 70 | 340 |
Expected Frequencies | ||||||
Sports | Game | Restaurant | Shopping | Concert | Total | |
Men | 72 * 200 / 340 = 42.3529 | 71 * 200 / 340 = 41.7647 | 77 * 200 / 340 = 45.2941 | 50 * 200 / 340 = 29.4118 | 70 * 200 / 340 = 41.1765 | 200 |
Women | 72 * 140 / 340 = 29.6471 | 71 * 140 / 340 = 29.2353 | 77 * 140 / 340 = 31.7059 | 50 * 140 / 340 = 20.5882 | 70 * 140 / 340 = 28.8235 | 140 |
Total | 72 | 71 | 77 | 50 | 70 | 340 |
(fo-fe)²/fe | ||||||
Men | (43 - 42.3529)²/42.3529 = 0.0099 | (50 - 41.7647)²/41.7647 = 1.6239 | (46 - 45.2941)²/45.2941 = 0.011 | (19 - 29.4118)²/29.4118 = 3.6858 | (42 - 41.1765)²/41.1765 = 0.0165 | |
Women | (29 - 29.6471)²/29.6471 = 0.0141 | (21 - 29.2353)²/29.2353 = 2.3198 | (31 - 31.7059)²/31.7059 = 0.0157 | (31 - 20.5882)²/20.5882 = 5.2654 | (28 - 28.8235)²/28.8235 = 0.0235 |
Null and Alternative hypothesis:
Ho: Preferred activity is independent of the gender.
H1: Preferred activity dependent on the gender.
Test statistic:
χ² = ∑ ((fo-fe)²/fe) = 12.9855
df = (r-1)(c-1) = 4
p-value = CHISQ.DIST.RT(12.9855, 4) = 0.0113
Decision:
p-value < α, Reject the null hypothesis.
There is enough evidence to conclude that preferred activity depends on the gender.