In: Math
Many years ago, Detroit had a six-month newspaper strike that closed down all its newspapers. During the strike, Detroit’s suicide rate fell very sharply (that is, there were far fewer suicides), but returned to its usual rate when the papers resumed publishing. Do the following: (a) Make up a theoretical model that would account for this observation and (b) generate a total of two interesting hypotheses from the model that could potentially be tested.
(a) We may assume the linear relationship between the variables and fit a linear regression line to capture the data.
The model will be : , where i=1(|) n , n is the total number of days
Here is the number of Detroit's suicides on ith day, is the number of newspapers sold on ith day and is a categorical variable which takes value 1 if there is a newspaper strike on ith day and 0 if there is no strike on ith day.
(b) (1) We can test :
This will test if there is exist a relationship between number of Detroit's suicides on ith day and number of newspapers sold on ith day.
(2) Again we can test:
This will test if there is exist a relationship between number of Detroit's suicides on ith day and
Another test we can conduct is
If the null hypothesis of this test is accepted then we need not to conduct the above two test that's why I want to conduct the above two tests separately.