Sandy was interested in where her friends' responses to an event
invitation(definite yes, definite no, maybe, no response) would
differ depending on their primary method of travel (car or bus).
The results are displayed in the following table. Using the .05
significance level, is the type of response independent of type of
travel?
Definite Yes Definite No Maybe No response
car 10 3 9 8
bus 20 10 8 32
\Use the fives steps of hypothesis testing
Sketch the chi-square distribution. Be sure your sketch gives a rough indication of its shape and shows the cutoff score, sample score, and region of rejection
Answer
Step 1 -> Null hypothesis H0 : no association between mode of
transport and response.
Alternative hypothesis H1 : association between the two
variables.
step 2 -> Make the table.
Definite yes |
definite no |
maybe |
no reponse |
Total |
|
Car |
10 |
3 |
9 |
8 |
30 |
train |
20 |
10 |
8 |
32 |
70 |
Total |
30 |
13 |
17 |
40 |
100 |
Step 3 : Calculate the expected frequencies ( =Row total * Column total/total)
Definite yes |
definite no |
maybe |
no reponse |
Total |
|
Car |
9 |
3.9 |
5.1 |
12 |
30 |
train |
21 |
9.1 |
11.9 |
28 |
70 |
Total |
30 |
13 |
17 |
40 |
100 |
Step 4 : Calculate chi-square statistic.
Chi-square = (observed -
expected)2/expected = 6.620699
Step 5 : Critical value = Upper 95% point of chi-square
distribution with df = (#rows-1)*(#columns-1) = 3
=7.814728
Since calculated chi-square < critical value, we accept H0 and
conclude that there is no significant association between the two
variables.
p-value = | 0.08505 |
Sandy was interested in where her friends' responses to an event invitation(definite yes, definite no, maybe,...