In: Statistics and Probability
Examine your classmate’s problem to assess the appropriateness
and accuracy of using a linear regression model.
Discuss the meaning of the standard error of the estimate and how
it affects the predicted values of Y for that analysis.
The problem I am interested in dealing with is the rate of illnesses as it compares to people that do or do not wash there hands frequently. This data would be collected throughout the year to see if there is a correlation of illness with those that do or do not wash their hands frequently.
A regression analysis would be appropriate due to the fact that illness are normally higher when personnel do not wash their hands on a consistent basis.
The data collection process could occur through a number of
ways, I think the most effective way would be to ask health care
professionals to poll their ill patients and ask if they are
frequent or infrequent hand washers. Data could also be collected
on how many times the patient came in for the same illness or
similar illness throughout the year.
Standard Error measures the accurateness of the estimates of the parameter of any linear regression model. Alternatively, the lower the value of standard error the more accuarte is the estimates of the parameter. So, if the SE is lower then we can say that the predictors predict the response variable Y more accurately and efficiently and also the fiiting is good.
In the given problem it is said that the data by asking health care professionals to poll their ill patients and ask if they are frequent or infrequent hand washers. So actually the response has two outcomes such as YES or NO i.e. binary. So, we use Logistic Regression model rather than Linear Regression model.
On the other hand if the data could also be collected on how many times the patient came in for the same illness or similar illness throughout the year. Here the response is a count or frequency data i.e. the response variable counts the occurences. So here we use Poisson Regression model.