In: Statistics and Probability
Question 3 (50 pts)
The following data show the information of “sex” and “being afraid of walking at night for 12 respondents.” Based on the data, answer the following questions,
Case # Sex Afraid of walking at nights
1 Male No
2 Male No
3 Male No
4 Female Yes
5 Female Yes
6 Male No
7 Female Yes
8 Female Yes
9 Female No
10 Male Yes
11 Female No
12 Female No
1.)Produce the crosstabulation, be sure to include the cell percentages, the row margins, the column margins, and the total number of cases in the table (30 pts)
(30 pts) 2) Compute its lambda (10 pts)
3) Interpret the lambda (10 pts)
(1)
Afraid to walk night | Total | ||
Sex | Yes | No | |
Male | 1 | 4 | 5 |
Female | 4 | 3 | 7 |
Total | 5 | 7 | 12 |
We simply take count the males and females and assign them the row totals and count yeses ad nos and assign them column totals. We can count 1 of the individual cells them rest of the values can be found using subtraction.
(2) Lambda shows the association between the dependent and the independent variable. It ranges from 0 to 1.
Where
Where E1 is the non modal frequency (lower total of the dependent variable ) or the error as we say. E2 is the error within the individual cell. (The non modal frequencies of the independent variable)
Here we can have two lambda depending on which variable is assigned what. We use the non-model frequencies.
(a) If sex is the independent variable then
(yes total - 5) (1 is lower in male and 3 is lower in female)
= 0.2
(b) If afraid of walking at night is independent then
(male total - 5 ) (1 is lower in yes and 3 is lower in no)
= 0.2
(3) So we can see that at both places there is an association of 0.2 between the variables. It means that the modal category of independent (in a: male and in b:no) and the overall modal frequency have an association of 0.2 (20%).