Very often, there are a number of different estimators that can be used to estimate unknown population parameters. When faced with such a choice, it is desirable to know that the estimator chosen is the "best" under the circumstances, that is, it has more desirable properties than any of the other options available to us. There are three desirable properties of estimators:

**unbiasedness****efficiency****consistency**

The single estimate of an unknown population parameter calculated as a sample mean is called a

A

For example, suppose that a 95% confidence interval for the population mean is 20 to 40. This means that:

- There is a 95% probability that the population mean lies in the range of 20 to 40.
- "95%" is the degree of confidence.
- "5%" is the level of significance.
- 20 and 40 are the lower and higher confidence limits, respectively.

danlan: level of significance = 1-degree of confidence |

achu: Note: strictly speaking we really can't say there's a "95% probablility" of the mean being between 20-40. See wikipedia.org/Confidence_intervals for a detail description. But for the exam, I guess it's probably not a big deal. |

vsimco: Tha above is correct, a confidence interval does not imply a probability statement of the estimated parameter being inside it (this is a given -- it is) nor does it give a probability of statement of the true mean. You cannot technically say the mean has a 95% probability of being inside a confidence interval. This is WRONG. The mean is either inside or outside the interval, there is no middle ground. THE TRUE MEAN IS NOT A RANDOM VARIABLE. What is being said is that 95% of all CONFIDENCE INTERVALS (note: the interval(S****)) contain the true mean. Its very subtle. |

sahilb7: UnEfCo: Unbiased, Efficiency, Consistency |

sahilb7: Unbiased: Mean = Intended ParameterEfficiency: Least variance among all parameters Consistency: Converges towards the actual value as the sample size increases |

yannick85: you are the best Sahilb7 |