heteroskedasticity &amp; serial correlation-sport

# heteroskedasticity & serial correlation-sport -...

This preview shows pages 1–15. Sign up to view the full content.

Slide # 1 * * * * * * * * * * * * * * * * * * * *

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 2 NOTE Read all slides in the full set of slides for this topic. There is other information in those slides that you must know.
3 Exam 3 Cumulative exam!! Includes 10 multiple choice questions from Exam 1

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 4 Research Project Report Due Class #14
Slide # 5 Research Project: Order of Testing 1 st . Omitted variables and incorrect functional form (F, adj. R 2 , plots) 2 nd : Note: do either B. or C. but not both B. Serial correlation ( time series) C. Heteroskedasticity (cross section) 3 rd . Multicollinearity (corr. matrix, VIF) 4 th . Irrelevant variables (t-statistics)

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 6 Heteroskedasticity (see notes) * * * * * * * * * * * * * * * * * * * *
Slide # 7 Heteroskedasticity & Serial Correlation are Sneaky You probably wouldn’t even think that they exist Yet, they can mess up t-statistics so badly that you might reach wrong conclusions about which IVs matter (i.e., affect dependent variable) and which do not

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 8 II. Introduction A. The word itself 1. Learn to spell "Heteroskedasticity" 2. Will have to spell it on next exam B. Who should learn this? 1. Those using cross-sectional data in their projects (“cross-sectional”??) 2. Those who won't o a) Still need to learn this o b) Can you guess why?
Slide # 9 Y X * * * * * * * * * * * * * * * * * * * Want dispersion of DV around regression line Want dispersion of DV around regression line ( ( σ σ 2 ) roughly constant across sample ) roughly constant across sample NO HETEROSKEDASTICITY Regression line

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 10 Y X * * * * * * * * * * * * * * * * * * * same throughout sample: same throughout sample: β β constant constant NO HETEROSKEDASTICITY Regression line
Slide # 11 Y X * * * * * * * * * * * * * * * * * * * Common case: dispersion of DV around Common case: dispersion of DV around regression line ( regression line ( σ σ 2 ) not constant ) not constant * * * * * * * HETEROSKEDASTICITY Regression line

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 12 average relationship between DV & IV average relationship between DV & IV NOT NOT same throughout sample: same throughout sample: β β NOT constant NOT constant Y X * * * * * * * * * * * * * * * * * * * * * * * * * * HETEROSKEDASTICITY Regression line
Slide # 13 When can it occur? 1. Example 2. MLB payrolls and market size o smaller market teams Less variability in payroll sizes Lower revenues available for payrolls o Larger market teams More variability in payroll sizes Greater revenues available for payrolls Some teams spend more on players Other teams spend relatively less on players

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Slide # 14 When can it occur? (cont.) MLB payrolls and market size LARGE markets (populations > 4,000,000)** o Standard deviation = \$27,940,000 o Range = \$97,470,000 Min = \$40,960,000 Max = \$138,420,000
This is the end of the preview. Sign up to access the rest of the document.

## This note was uploaded on 09/09/2011 for the course ECON 6416 taught by Professor Richardhofler during the Fall '11 term at University of Central Florida.

### Page1 / 77

heteroskedasticity & serial correlation-sport -...

This preview shows document pages 1 - 15. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online