# MMU The Value of The Estimated Effect of Smoking Question

Description

2 attachmentsSlide 1 of 2attachment_1attachment_1attachment_2attachment_2.slider-slide > img { width: 100%; display: block; }
.slider-slide > img:focus { margin: auto; }

Unformatted Attachment Preview

2. The following table summarizes some regressions using data from a random
sample of babies born in Pennsylvania in 1989. The data include the baby’s
birthweight (the dependent variable) together with various characteristics of
the mother, including whether she smoked during the pregnancy.
Birthweight=birth weight of infant (in grams)
Smoker=1 if the mother smoked during pregnancy, 0 otherwise.
Alcohol=1 if mother drank alcohol during pregnancy, 0 otherwise
Nprevist=total number of prenatal visits
Unmarried=1 if mother is unmarried, 0 otherwise
Age=age in years
Educ=years of educational attainment (more than 16 years coded as 17)
Each column contains results from regression estimates from a model explaining
Birthweight. An empty cell means that the variable was not included in the model for
that column. For example in column 1 Birthweight was the dependent variable and an
intercept and the smoking dummy variable were the only regressors. Robust Standard
errors are reported in brackets underneath the parameter estimates.
Regressor
Smoker
(1)
-253.2
(26.8)
Alcohol
(2)
-217.6
(26.1)
-30.5
(72.6)
34.1
(3.6)
(3)
– 175.4
(26.8)
-21.1
(73.0)
29.6
(3.6)
-187.1
(27.7)
(4)
– 177.0
(27.3)
– 14.8
(72.9)
29.8
(3.6)
-199.3
(31.0)
Nprevist
Unmarried
Age
Educ
-2.5
(2.4)
-0.238
(5.53)
3199.4
(90.6)
Intercept
3432.1
(11.9)
3051.2
(43.7)
3134.4
(44.1)
SER
583.7
0.028
3000
570.5
0.072
3000
565.7
0.087
3000
565.8
0.087
3000
n
a) What is the value of the estimated effect of smoking on birthweight in each of
the regressions? (include units)
b) Construct a 95% confidence interval for the effect of smoking on birthweight,
using each of the regressions.
c) Is it likely that the coefficient on Smoker in regression (1) suffers from omitted
variables bias? Explain.
d) Is it likely that the coefficient on smoker in regression (2) suffers from omitted
variable bias? Explain.
e) Consider the coefficient on Unmarried in regression (3).
(i) Construct a 95% confidence interval for the coefficient.
(ii) is the coefficient statistically significant? Explain.
(iii) is the magnitude of the coefficient large? Explain.
(iv) A family advocacy group notes that the large coefficient suggests that
public policies that encourage marriage will, on average, lead to healthier
babies. Do you agree?
(Hint: Review the discussion of control variables in section 7.5 of SW 3rd
updated edition or section 6.8 of the 4th edition. Think of the various factors
that Unmarried may be controlling for and how this affects the interpretation
of its coefficient. Is it likely that the coefficient on Unmarried represents a
causal effect of marriage or is it controlling for other factors?)