We are interested in studying whether a mother's smoking affects the birthweight of her baby.

The dataset for this question is birthweight2019.csv available on Moodle. The variables you

need are as follows:

bweight: infant birthweight (grams)

smoke2: 1 if mother smokes 1-5 cigarettes per day, 0 otherwise

smoke3: 1 if mother smokes 6-10 cigarettes per day, 0 otherwise

smoke4: 1 if mother smokes 11 or more cigarettes per day, 0 otherwise

mmarried: 1 if mother married, 0 otherwise

mage: mother's age

Questions:

1. (a). Estimate the following regression model, and report the estimated coefficients and the

associated standard errors. (1 points)

bweighti = 1 + 2smoke2i + 3smoke3i + 4smoke4i + 5magei + 6mmariedi + ei

(1)

(b). Interpret the estimated coefficients of smoke2, smoke3 and smoke4. (3 points)

1

2. What is the estimated difference in the expected birthweight of a baby whose mother is

married and smokes 6-10 cigarettes a day relative to that of a baby whose mother is not

married and smokes 11 or more cigarettes a day? Keep 4 decimals for the calculation

involved. (2 points)

3. (a).

Now consider a similar model but with log of birthweight as dependent variable.

Estimate this model, and report the estimated coefficients and the associated standard

errors. (1 points)

ln(bweight)i = 1 + 2smoke2i + 3smoke3i + 4smoke4i + 5magei + 6mmariedi + ei

(2)

(b). Calculate and interpret the percentage differences in birthweight by comparing groups

smoke2, smoke3 and smoke4 with the reference group. (3 points)

4. Based on the model (2), test at the 5% significance level the null hypothesis that smoking 11

or more cigarettes per day reduces birthweight by no more than smoking 6-10 cigarettes per

day, against the hypothesis that smoking 11 or more cigarettes per day reduces birthweight

by more than smoking 6-10 cigarettes per day. Write down all the step