Malaria Is A Disease That Is Highly Prevalent Throughout Much Of The Developing World: Statistics Assignment, UQ, Australia

Q1: 

Malaria is a disease that is highly prevalent throughout much of the developing world. It takes the form of a parasite that is carried by mosquitoes, and hence it is particularly severe in warm and wet climates where the insects thrive. Given that malaria has a high mortality rate and can incapacitate a large proportion of the working-age population, the disease is thought to contribute greatly to poverty throughout Africa and Asia.

There is a large research effort dedicated to assessing the effectiveness of preventative measures in order to control the disease, including the use of mosquito nets. Your task is to examine the effectiveness of a mosquito net program using data collected from villages in Senegal. In some villages, the researchers distributed mosquito nets (the treatment group) while in others they did not (the control group). The data was obtained by recording details of each village twice, before and after the nets were handed out.

The data can be found in the file malaria.wf1 where the variables are defined as follows: malaria gives the percent infection rate in the village, income gives the average income level (in USD), rain gives the annual rainfall, water is a dummy variable which is equal to one if there is a stagnant water source, and temperature is the average temperature in degrees Celsius. In addition, there are two dummy variables. Period denotes the time period where 0 indicates the period before the nets were distributed, and 1 denotes the period after.  MN denotes the villages that received the nets with a value of 1.

1. To assess the effectiveness of the program a difference-in-differences estimator is used. Briefly explain the concept of a “parallel trend” for this estimator and what this assumption implies for the mosquito net program.

2. Why can’t we estimate the effectiveness of the program by simply averaging the malarial rates (in the latter period) for villages that did and did not receive the nets? Briefly explain

3. Calculate the difference-in-differences estimate of the effectiveness of the program (in terms of reduction in percent infections) based only upon average rates before and after the nets were distributed.

4. Give the regression equation that would be used to calculate the full difference-in-difference estimate employing the full set of control variables. Also, estimate this equation and interpret the results. Why is this estimate preferable to the estimate given in part 3?

5. Briefly perform some diagnostics on your model. Do the coefficients make sense? Are there any signs of misspecification? Are the residuals in line with your assumptions?

Q2:

A financial firm has hired you to study the empirical links between an aggregate of monthly returns on a US tech portfolio, and returns on similar portfolios based on European firms and firms from Hong Kong.  The file tech.wf1 has 240 observations on these variables, denoted US, Eur and HK respectively. You are to estimate an) model to capture the dynamic interplay between these variables. Note that since you have two “x” covariates, you may like to think of the model as having the form ), where the Q terms refer to lags of separate lengths for the two independent variables.

1. Select the appropriate lag structures to ensure that your model (i) has desirable goodness of fit properties, and (ii) is free of autocorrelation. Give your estimated equation and provide appropriate supporting outputs to demonstrate the quality of your model.

2. Write a paragraph explaining how you settled on your specification in the previous question. Why is it better than other possible specifications?

3. Write a paragraph or two explaining the dynamics of returns as estimated by your model. Which index seems more important for returns in the US? How long does it take for movements in these markets to impact the US index? Is there evidence of feedback effects in these variables? How could your model be used to inform investment or trading decisions for US tech companies?

Q3:

Smoking is known to have a variety of negative health effects including an increased risk of lung cancer, stroke, and heart disease. One less-well-studied effect is the impact of maternal cigarette consumption while pregnant on fetal health. In this question, you are to estimate this impact, while allowing for the fact that smoking is potentially endogenous to infant health. The data set smoking.wf1 has data on birth weights, maternal age (and age squared), income, education, BMI, and whether the birth is a girl or boy.  Also included are two instruments – a tax exemption on cigarettes and a dummy indicating anti-smoking advertising intensity.

1. Regress birth weights on age, age squared, income, education, BMI, girl, and smoking. How much lower are the birth weights of mothers who smoke?

2. Perform the same estimation where you instrument smoking using the price of cigarettes and advertising intensity. Do your estimates change very much?

3. Discuss why smoking might be endogenous, and explain why your instrumental variable estimation will help get around this problem.

Q4: 

The file infant.wf1 contains World Bank data on country-level infant mortality rates across the developed and developing world. The data set contains 394 observations over a 25-year period (from 1990-2014) and includes a number of potential covariates that might plausibly affect health outcomes. The variables are defined as follows:

Infant Mortality – The number of deaths per 1000 live births.

GDP – The real GDP per capita (in US dollars) for each country.

Year – The year the observation was recorded.

Contraceptive Prevalence– The proportion of individuals with access to contraceptives.

Physicians – The number of doctors per 1000 people.

Sanitation – The proportion of individuals with access to modern sanitation facilities.

Education – A measure of educational attainment in terms of years of post-primary attendance.

Your task is to produce an econometric model that will yield useful information for understanding the factors that drive infant mortality. The idea is ultimately to inform policies that could alleviate problems associated with infant mortality in developing countries.

Be sure to highlight the strengths and weaknesses of your model. Also include an Eviews output of your final model and of other related outputs that helped you make your decisions.

You may like to (i) use alternative functional forms and (ii) include different transformations of the same variable in your chosen regression.  You should also consider at least some of the following:

(1) The quality of the fit

(2) The significance of the coefficients

(3) Whether the specification of your model is appropriate for the data

(4) The parsimony of your model (i.e. there is a slight preference for simple models over complex ones)

(5) Whether the standard errors are correct

(6) The extent to which multicollinearity affects your model

(7) What the model ultimately implies about infant mortality rates.