Scribd
Upload
3 views3 pages

Tutorial 2 Sol

The document discusses key concepts in econometrics, focusing on the importance of the normality assumption in OLS regression for constructing interval estimators and hypothesis testing. It addresses various questions related to regression analysis, including the interpretation of intercepts, predicted values, model explanatory power, and issues like heteroskedasticity and omitted variable bias. Additionally, it explores correlations between house prices and environmental factors, highlighting potential biases in estimations.

Uploaded by

202330416
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views3 pages

Tutorial 2 Sol

The document discusses key concepts in econometrics, focusing on the importance of the normality assumption in OLS regression for constructing interval estimators and hypothesis testing. It addresses various questions related to regression analysis, including the interpretation of intercepts, predicted values, model explanatory power, and issues like heteroskedasticity and omitted variable bias. Additionally, it explores correlations between house prices and environmental factors, highlighting potential biases in estimations.

Uploaded by

202330416
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Econometrics 1, 2026

Tutorial 2 Sol

Question 1. The classical normal linear regression model assumes that each 𝑢𝑖 is distributed
normally: 𝑢𝑖 ~𝑁(0, 𝜎 2 ). Explain why this is important in the OLS regression.
Asymptotic Normality: Why we need this? Consistency is a good property of point estimators
but it says nothing about the shape of the estimator’s distribution. We need to normality
assumption to approximate distribution for constructing interval estimators and hypotheses.
Central Limit Theorem (CLT): One of the most powerful results in probability and statistics. It
states that the average from a random sample for any population (with finite variance), when
standardized, has an asymptotic standard normal distribution.
Law of Large Numbers (LLN) and CLT: Most estimators in statistics and econometrics can be
written as functions of sample averages. Without normality assumption; these inferences are only
approximately valid in large samples.
Question 2. (Woolridge, P. 59)

a) Interpret the intercept in this equation and comment on its sign and magnitude.
b) What is the predicted consumption when family income is $30,000?
c) What does the r-squared say about this model?
d) With inc on the x-axis, draw a graph of the estimated MPC and APC.
Answers: a. The intercept represents the predicted annual consumption when the annual income
is zero. A family with no income, on average would have to borrow (or dissave) $124 to maintain
some level of consumption.
b. cons= -124.84 + 0.853(30000) = 25465.16
c. The explanatory variables in the model can explain only 69.2% of the variations in the
explained variable or income can only explain 69.2% of the variations in family consumption.

Question 3.
i. Heteroskedasticity
ii. Omitting an important variable
iii. A sample correlation coefficient of .95 between two independent variables both included
in the model
a. Which of the above can cause OLS estimators to be biased?
Only (ii). (i) and (iii) only contributes to inefficiency, not biasedness.
b. Which of the above can cause the usual OLS t- statistics to be invalid (that is, not to have
t distribution under 𝐻0 )?
(i) and (ii). When (i) is present, the standard errors calculated by the OLS formulas are incorrect,
making the OLS t-statistics invalid and omitting an important variable causes biasedness
therefore the t-statistics based on this biased estimator is invalid.
Multicollinearity only affects the precision of the estimates, not the t-statistics.
Question 4. (Woolridge P. 104)

The negative bias as average training has a positive impact on average product and it should not
be negatively correlated.
Question 5. (Woolridge P. 104)

(i) House prices and (nox) would be negatively correlated because areas where there is high
concentration of pollution may be less desirable while house prices will be positively correlated
with rooms as house with higher number of bedrooms tend to be expensive.
(ii) negatively correlated because (nox) are less desirable and may have smaller rooms,
especially in big cities and industrial areas. Downward biased estimation as it is under-estimating
the true impact of pollution.

You might also like