This preview shows pages 1–2. Sign up to view the full content.
Homework #5 —
Model MisSpecification/Equality Constraints/Group Comparisons — Page 1
Soc 63993, Advanced Social Statistics II
Homework No. 5
Model MisSpecification/Equality Constraints/Group Comparisons
1.
Model misspecification.
A campaign manager has found that the amount of time spent
watching TV political ads is negatively correlated with favorable opinion of her candidate.
Two
models have been proposed to explain this relationship:
(i)
Amount watch TV
Political Ads
Education
Opinion of
Candidate

+
+
(ii)
Amount watch TV
Political Ads
Opinion of
Candidate

A.
Suppose that model (i) is correct.
What harm will result from estimating model (ii) and
relying on the results?
If appropriate, discuss such things as biased coefficients, inflated standard
errors, and misguided policy decisions (particularly with regards to the use of TV advertising).
Similarly, discuss the harm that will result if Model (ii) is correct and model (i) is mistakenly
estimated and relied upon.
B.
Model (i) is estimated, yielding the following results.
Based on this information,
determine what the regression coefficient would be for model (ii).
Compute the regression
coefficient using both the formula for omitted variable bias and the formula for the slope
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
This is the end of the preview. Sign up
to
access the rest of the document.
 Spring '11
 RichardWilliams

Click to edit the document details