In: Statistics and Probability
a.Sketch a graph of x=age and y=annual medical expenses.
Hint: Suppose expenses tend to be relatively high for the newborn and for the elderly.
b.Can we use simple linear regression? why or why not?
NO DATA SET WAS PROVIDED. IT IS IMPROVISED WITH NO INFO (WE CAN MAKE IT UP)
The above graph clearly shows that medical expenses are high for a newborn and gradually decrease as one approaches the youth and thrafter again start incresing as a person gets old. Thus the graph is first decreasing with increase in age and then increasing with increase in age.
b) Since no data is given to check if a linear regression line can be fit to age and medical expense we can consider two cases-
Firstly we consider a case when age and medical expense have a direct relationship and are not affected by other variables, that is one is dependant on other then we can consider a linear regression for the two variables.
Secondly, In another case if as in reality we consider that medical expenses depend not only on age but other variables also then we might have to consider a multiple linear regression model. Other factors we can consider which affest medical expense are food,environment , sanitation and hygiene etc. Thus the multiple linear model can take these two in consideration also.