Explain what is meant by autocorrelated errors. What effects does autocorrelation have on the results of a regression?
What will be an ideal response?
Autocorrelated errors occur when the errors in a regression model are no longer independent. The hypothesis tests and confidence intervals in multiple regression assume that the errors are independent. If the errors are not independent, then the estimated standard errors for the coefficients are biased. For example, it can be shown that, if there is a positive correlation between the error terms, then the least-squares estimate of the coefficient standard coefficient will be too large. This could lead us to conclude that certain coefficients are significantly different from 0 - by rejecting H0 : ?j = 0 - when, in fact, H0 should not be rejected. In addition, estimated confidence intervals would be too narrow.
You might also like to view...
Jay Co., CPAs, audited the financial statements of Maco Corp. Jay intentionally expressed an unmodified opinion on the financial statements even though material misstatements were discovered. The financial statements and Jay's unmodified opinion were included in a registration statement and prospectus for an original public offering of Maco stock. Which of the following statements is true regarding Jay's liability to a purchaser of the offering under Section 10(b) and Rule 10b-5 of the Securities Exchange Act of 1934?
A. Jay will be liable if the purchaser relied on Jay's unmodified opinion on the financial statements. B. Jay will be liable if Jay was negligent in conducting the audit. C. Jay will not be liable if the purchaser's loss was under $500. D. Jay will not be liable if the misstatement resulted from an omission of a material fact by Jay.
When Mercedes Benz promises the finest luxury ride in an automobile, the positioning approach being used is:
A) competitors B) price-quality relationship C) use or application D) cultural symbol