Question

“Larger hospitals in South Florida display higher readmission rates and thus lower quality of care compared

to smaller hospitals.”

A). Identify the implied dependent and independent variables.

B). Does the headline demonstrate a positive or negative correlation between the two?

C). Is there a plausible mechanism (confounding factor) through which we could explain why the independent variable has a

causal impact on the outcome?

D). Now try to think of reasons why this could be a spurious relationship instead and construct alternative explanations.

Answer 1

A) In the above example, the possible dependent variable could be the 'Readmission Rate per 1000 Patients'. Here we have selected number of readmissions because larger hospital have greater possibilities to have higher number of the patient and again high chance of getting re-admission

The independent variable could be 'Number of Patients coming every day'

B) By looking at the above headline, we came to know that it is a positive correlation because Larger hospitals in South Florida display higher readmission rates.

C) Since larger hospitals have a highly sophisticated facility and good care for chronic and complex disease & casualties. So the cases coming to larger hospitals have more chances to be serious and critical which requires good number of visits to hospitals. Whereas, the cases coming to the smaller hospitals are not that critical and serious and can be cured in less number of visits.

So, the 'Type of disease being cured' can be a potential confounder to this relationship and that is why it is showing this type of causal relationship.

D) Since the relationship between 'Larger hospitals and readmission rate' is not directly related rather it is being explained by another variable "Type of disease" so we can that the above relationship is spurious.

Because it is not a larger hospital that causes a higher readmission rate rather it is a type of disease which controls the readmission rate.

An alternate hypothesis might be that the 'type of disease' has a significant relationship with readmission rate.