Scribd
Upload
51 views1 page

Homework1 2023

1. The document presents 4 problems regarding regression analysis and estimators. Problem 1 involves comparing the mean, variance and other properties of 3 different estimators for a population mean μY. Problem 2 proves an identity used in deriving the OLS formula for β1. Problem 3 analyzes a regression of earnings on age and asks questions about interpreting coefficients, units, R-squared, predictions, and calculating average earnings. Problem 4 discusses potential bias in estimating the effect of training when ability is omitted and correlated with training.

Uploaded by

darinaveresk
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
51 views1 page

Homework1 2023

1. The document presents 4 problems regarding regression analysis and estimators. Problem 1 involves comparing the mean, variance and other properties of 3 different estimators for a population mean μY. Problem 2 proves an identity used in deriving the OLS formula for β1. Problem 3 analyzes a regression of earnings on age and asks questions about interpreting coefficients, units, R-squared, predictions, and calculating average earnings. Problem 4 discusses potential bias in estimating the effect of training when ability is omitted and correlated with training.

Uploaded by

darinaveresk
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 1

Problem 1 (30 points: 15+15)

Denote the mean value of random variable Y as μ Y. You want to estimate it using sample data. Consider a
sample of n independently and identically distributed observations Y 1, Y2, …, Yn (n is even: 2, 4, 6, etc.)
n
1 1
Consider three estimators for μY: Y̆ = (Y 1 +Y n ),
2
Y= ∑Y
n i=1 i
and

~ 1 1 3 1 3 1 3
Y = ( Y 1+ Y 2+ Y 3 + Y 4 +…+ Y n−1 + Y n ).
n 2 2 2 2 2 2

(a) Find the mean and variance of these estimators.


(b) What can you say about the unbiasedness, consistency and efficiency of these three estimators?

Problem 2 (15 points)


Show that
n n

∑ x i ( y i ¿− y )=∑ (x ¿¿ i−x)( y i ¿− y ) ¿ ¿ ¿
i=1 i=1

We use this result in the derivation of OLS formula for β1 in a simple model with one regressor.

Problem 3 (35 points: 5+5+5+5+5+5+5)


Consider the following regression of average weekly earnings (Earnings, measured in USD) on age
(measured in years) using a random sample of college-educated workers aged 25-65:

^
Earnings=696.7+9.6 Age

R2 = 0.023, SER = 624.1 (standard error of the regression).

(a) Explain what the coefficient values 696.7 and 9.6 mean?
(b) What is unit of measurement for the SER - USD, years, or is it unit-free?
(c) What is unit of measurement for the R2 - USD, years, or is it unit-free?
(d) What does the value of R2 mean?
(e) What is the regression's predicted earnings for a 25-year-old worker? A 45-year-old worker?
(f) Will the regression give reliable predictions for a 72-year-old worker? Why or why not?
(g) The average age in the sample is 41.6 years. What is the average value of Earnings in the sample?

Problem 4 (20 points)


Assume that average worker productivity in small firms of the service sector (avgprod) depends on two
factors, average hours of training (avgtrain) and average worker ability (avgabil):

avgprod = β0 + β1 avgtrain + β2 avgabil + u.

Suppose that the above regression model satisfies the Gauss-Markov assumptions. A local government starts
a training program aimed to increase the level of worker training in these firms. It turns out that firms whose
workers have above average ability benefit from this program disproportionately, resulting in a positive
correlation between avgtrain and avgabil. If we (as econometricians) do not observe average worker ability
(avgabil), what is the likely bias in the estimated value of β1 obtained from the simple regression of avgprod
on avgtrain? Explain.

You might also like