This preview shows page 1. Sign up to view the full content.
Unformatted text preview: researcher is interested in testing whether the level of
beer consumption of individuals is independent of their level of
cigarette consumption. A random sample of 1,000 individuals are
interviewed about their weekly consumption of beer and cigarettes,
yielding the following bivariate frequency distribution: nil
nil
Cigarette 120
cons.
2160
>60 256
105
71
22 Beer consumption
13
410
glasses
glasses
68
54
110
87
56
25
26
17 Over 10
glasses
25
42
26
10 Given the data, perform a test of independence of beer and cigarette
consumption using a 5% significance level. The expected cell frequencies are given in brackets in the following table.
nil
nil Cig.
cons. 256
(182.962)
120
105
(156.176)
2160
71
(80.812)
>60
22
(34.05)
Col.
454
Total Beer consumption
13
410
Over 10
Row
glasses
glasses
glasses
total
68
54
25
403
(104.78) (73.749) (41.509)
110
87
42
344
(89.44) (62.952) (35.432)
56
25
26
178
(46.28) (32.574) (18.334)
26
17
10
75
(19.5) (13.725)
(7.725)
260
183
103
1000 4 e.g. ( )( ) Now the test statistic is given by
∑∑ ( ) ( And since
we reject
consumption are not independent. ) ( ) and conclude that beer and cigarette 5...
View Full
Document
 '11
 DenzilGFiebig

Click to edit the document details