Why did you choose this topic? What are you trying to find out?
It is important to know the factors that affect the mortality rate in a region. This analysis has economic significance for appropriate policy decisions. If the factors that have a positive effect on mortality rate or increases the mortality rate can be known, those factors can be controlled by the government. Again if the factors that can reduce the mortality rate can be identified we can increase the availability of these factors in the economy so that mortality rate is reduced. Thus the study of the factors affecting mortality rate is quite important from the point of few of framing a good health policy.
The objective of my study is to find out how the death rate of a region is related to the availability of doctors, availability of hospitals, per capita income and population density of the region. Thus the variables are given as:
What you think the model will predict. (What do you expect the “b” values for each variable to be? (positive? Negative? Why?)
According to my idea about the death rate I am of the opinion that doctor and hospital availability will have negative coefficients as more hospitals and doctors provide better healthcare to reduce the mortality rate. Per capita income should also have negative coefficient. But population density may have a positive coefficient as a higher density will mean more crowding at the hospitals and the doctors gets less time to attend the medical cases as there will be too many patients.
What are the coefficient parameters (b’s) and what do they mean?
The coefficient is given below:
X1 = 0.0074X2 + 0.0006X3 -0.33X4 - 0.0095X5
We can see that X2 has a coefficient of 0.0074. If X2 or doctor availability increases by 1per cent mortality rate will increase by 0.0074 per cent. Similarly X3 has a coefficient of 0.0006 and X4 has a negative coefficient which implies that increase in per capita income will reduce mortality rate. Population density also has a negative impact on death rate. That is as density increases death rate falls.
Which of these coefficient parameters are individually significant? Are they economically significant?
X4 and X5 has statistically significant coefficients. This is economically significant also. This result suggests that affluence tends to reduce mortality. Thus growth of the economy should lead to lower mortality rate. Increase in population density also reduced mortality. The reason is a densely populated area has more amenities and transport facility. The illness can be reported on time and addressed in a timely manner.
How well do the predictor variables explain the dependent variable?
The value of multiple R square is quite low at 0.38. That means the model has 38% predictability. It can explain only 38% of the variations in the dependent variable. Thus the predictor variable cannot explain the dependent variable to a significant extent.
Is the model significant as a whole? How do you know?
The model is significant as a whole as the F value is higher that the critical F value. Thus the independent variables do not have zero coefficients. They do influence the dependent variable.
Did the model and the signs and values of the parameter agree with your prediction? If no, why do you think it didn’t? (was there perhaps a violation of the 5 assumptions?)
The signs of the coefficient were not according to expectation. Availability of doctor and hospital should have a negative coefficient but the coefficients are positive. But the coefficients are insignificant and too small. Thus the availability of doctors and hospitals cannot affect mortality individually.
How do you think you might improve your model? Be specific. If you were to add another variable in there, which one would it be and why?
The independent variables should be more specific like including the quality of medical facility available. The transport and communication system is also important determinant in this case. The level of education should be one more variable to be added.
How do you think you might improve your research next time?
I would include diagnostic tests after conducting the regression to verify that all the assumptions like autocorrelation and homoscedasticity hold good for the model. I would check the incidence of genetically transmitted diseases in the area as another independent variable.
Submit the actual data set as well (and tell me the link where you got the data)