In: Statistics and Probability
5. h
The next problem comes directly from a student project from three years ago. It is on a very serious topic so I am putting it last on the final.
A former student in Math& 146 did a project that asked people whether they had ever committed a crime, whether they were convicted, and whether they suffered from mental health problems. She thought there would be an association between being convicted and having mental health problems. She collected data through a confidential survey in her office, and then she also collected data from homeless people through in-person interviews.
The table below shows all people in her sample who said they
committed a crime.
|
No - Mental Health Problems |
Yes - Mental Health Problems |
Total |
Convicted |
20 |
22 |
|
Not convicted |
27 |
16 |
|
Total |
Does the data suggest there is an association between being convicted of a crime and having mental health problems? Provide a comprehensive answer using concepts from two-way tables and probabilities. For full credit, explain at least one possible source of bias in this research.
Expected frequency of a cell = sum of row*sum of column / total sum | |||||||
Expected Frequencies | |||||||
no | yes | F | P | Total | |||
convicted | 47*42/85=23.224 | 38*42/85=18.776 | 42 | ||||
not convicted | 47*43/85=23.776 | 38*43/85=19.224 | 43 |
(fo-fe)^2/fe | ||||||
convicted | 0.447 | 0.553 | ||||
not convicted | 0.437 | 0.541 |
Ho: given two variable are independent
H1: Given two variables are not independent
Chi-Square Test Statistic,χ² = Σ(fo-fe)^2/fe =
1.978
Level of Significance = 0.05
Number of Rows = 2
Number of Columns = 2
Degrees of Freedom=(#row - 1)(#column -1) = (2- 1 ) * ( 2- 1 )
= 1
p-Value = 0.1596 [Excel function:
=CHISQ.DIST.RT(χ²,df) ]
Decision: p value > α , do not reject
Ho
There is not sufficient evidence that there is an association between being convicted of a crime and having mental health problems
Response bias could a major concern
THANKS
revert back for doubt
please upvote