In: Statistics and Probability
Students in a high school were asked whether good grades, athletic ability, or popularity was most important to them. A two-way table separating the students by gender and by choice of most important factor (factor) is shown below:
| 
 Boys  | 
 Girls  | 
||
| 
 Grades  | 
 12  | 
 16  | 
|
| 
 Popularity  | 
 16  | 
 32  | 
|
| 
 Sports  | 
 30  | 
 14  | 
|
| 
 100  | 
For the data above, do the following
a)
| Boys | Girls | Total | |
| Grades | 12 | 16 | 28 | 
| Popularity | 16 | 32 | 48 | 
| Sports | 30 | 14 | 44 | 
| Total | 58 | 62 | 120 | 
b)
| Gender by factor percentages | |||
| Boys | Girls | Total | |
| Grades | 42.85714% | 57.14286% | 100% | 
| Popularity | 33.33333% | 66.66667% | 100% | 
| Sports | 68.18182% | 31.81818% | 100% | 
c)
| Factor by gender percentages | ||
| Boys | Girls | |
| Grades | 20.68966% | 25.80645% | 
| Popularity | 27.58621% | 51.6129% | 
| Sports | 51.72414% | 22.58065% | 
| Total | 100% | 100% | 
d)
Chi square test of independence need to be done to test the association between gender and factor
e)

Where, r is the number of rows and c is the number of column in the table.
f)
The expected values are obtained using the formula,

| Boys | Girls | Total | |
| Grades | 
![]()  | 
![]()  | 
28 | 
| Popularity | 
![]()  | 
![]()  | 
48 | 
| Sports | 
![]()  | 
![]()  | 
44 | 
| Total | 58 | 62 | 120 | 
g)
The Chi-Square Value is obtained using the formula

Observed, ![]()  | 
Expected, ![]()  | 
![]()  | 
![]()  | 
![]()  | 
| 12 | 13.53333333 | -1.533 | 2.351 | 0.174 | 
| 16 | 23.2 | -7.200 | 51.840 | 2.234 | 
| 30 | 21.26666667 | 8.733 | 76.271 | 3.586 | 
| 16 | 14.46666667 | 1.533 | 2.351 | 0.163 | 
| 32 | 24.8 | 7.200 | 51.840 | 2.090 | 
| 14 | 22.73333333 | -8.733 | 76.271 | 3.355 | 
| 11.603 | 
h)

i)
From the chi square table,
For significance level = 0.05 and degree of freedom = 2, the critical value is,

j)
Since the chi square value is greater than the critical value, It can be concluded that the null hypothesis is rejected at 5% significance level.
Hence both the variables factor and gender are dependent.