Z: Understanding Unbiased Estimators in Regression Analysis

  • Thread starter Thread starter wow007051
  • Start date Start date
Click For Summary

Homework Help Overview

The discussion revolves around the concept of unbiased estimators in the context of regression analysis. Participants are exploring the definition of unbiased estimators and the conditions under which regression models may yield biased estimates.

Discussion Character

  • Conceptual clarification, Assumption checking

Approaches and Questions Raised

  • The original poster attempts to understand what constitutes an unbiased estimator and how to verify if a regression provides one. Some participants discuss factors that can lead to biased estimators, such as omitted variable bias and model mis-specification. Others question the relationship between the coefficient of determination (r²) and the bias of an estimator.

Discussion Status

The discussion is active, with participants providing insights into the nature of unbiased estimators and raising questions about the implications of low r² values. There is a clear distinction made between bias and r², indicating a productive exploration of the topic.

Contextual Notes

Participants are navigating the complexities of statistical estimation and regression analysis, with some uncertainty regarding the implications of various statistical measures on bias.

wow007051
Messages
5
Reaction score
0
first of all...what is an unbiased estimator??
how to check whether a reggression provide an unbiased estimator?

thanks!
 
Physics news on Phys.org
www.talkstats.com
stats.stackexchange.com.

The following three events can cause biased estimators:

1) Omitted variable bias.
2) cov(error,regressors) \not= 0
3) cov(regressor1, regressor2) \not= 0
4) Model mis-specification (eg not including a squared term when you should - do a RAMSEY RESET test).
 
An unbiased estimator is a sample function:
<br /> Z_n = f(X_1, \ldots, X_n)<br />
such that, for an i.i.d. sample with a parameter of the distribution θ that we are trying to estimate, has the property:
<br /> \mathrm{E}\left[Z_n \right] = \theta<br />
If this does not hold for a finite n, but is true as n \rightarrow \infty, then we say that the estimator is asymptotically unbiased.

In general, if the function f is some non-polynomial function, it is very hard to check the bias of the estimator. If, on the other hand, the estimator is a (symmetric) polynomial of degree p (pth moment), we may use some rules for the expectation values. For example, the mean:
<br /> \bar{X}_n \equiv \frac{1}{n} \, \sum_{k = 1}^{n}{X_k}<br />
has the property:
<br /> \mathrm{E} \left[\bar{X}_n \right] = \frac{1}{n} \, \sum_{k = 1}^{n}{\mathrm{E} \left[ X_k \right]} = E \left[ X \right]<br />
is the unbiased estimator of the mathematical expectation of the random variable X.
 
can i say if a regression with very low r^2 it doesn't provide unbiased estimator?
 
can i say if a regression with very low r^2 it doesn't provide unbiased estimator?
 
wow007051 said:
can i say if a regression with very low r^2 it doesn't provide unbiased estimator?

No. Bias and r^2 are not really related.

RGV
 

Similar threads

  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 7 ·
Replies
7
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 23 ·
Replies
23
Views
4K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
19
Views
5K
Replies
3
Views
1K
  • · Replies 8 ·
Replies
8
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
Replies
7
Views
11K