Unbiased Estimator

What is an Unbiased Estimator?

An unbiased estimator is a statistical term used to describe an estimator that is expected to hit the true parameter value of the population from which the data is sampled. In other words, an unbiased estimator provides correct parameter estimates on average across many samples. The concept of unbiasedness is central to the field of statistical inference, where the goal is to make conclusions about a population based on a sample.

Mathematically, an estimator is unbiased if its expected value equals the parameter it is estimating. If θ represents the true parameter value and ˆθ (theta hat) represents the estimator, then ˆθ is unbiased if:

E[ˆθ] = θ

where E[ˆθ] is the expected value of the estimator. If an estimator is not unbiased, it is said to be biased.

Importance of Unbiased Estimators

Unbiased estimators are important because they do not systematically overestimate or underestimate the parameter being estimated. This is crucial for the reliability and credibility of statistical analysis. When an estimator is biased, it may consistently lead to incorrect conclusions, which can affect decision-making processes in various fields such as economics, medicine, and social sciences.

Examples of Unbiased Estimators

A classic example of an unbiased estimator is the sample mean. If we have a random sample X1, X2, ..., Xn from a population with a true mean μ, then the sample mean:

X̄ = (1/n) ∑ Xi

is an unbiased estimator of μ, because its expected value is equal to the true population mean.

Another example is the sample variance. However, it is important to note that to make the sample variance an unbiased estimator of the population variance σ², a correction factor is used. The unbiased estimator of the variance is calculated as:

S² = (1/(n-1)) ∑ (Xi - X̄)²

where n is the sample size. The use of (n-1) instead of n corrects for the bias that results from estimating the population mean with the sample mean.

Biased vs. Unbiased Estimators

While unbiasedness is a desirable property, it is not the only criterion for choosing an estimator. In some cases, a biased estimator may be preferred if it has lower variance, leading to more precise estimates. This trade-off between bias and variance is a key consideration in statistical estimation and is known as the bias-variance trade-off.

For example, in the context of linear regression, the ordinary least squares (OLS) estimator is unbiased under certain conditions. However, if there is multicollinearity in the data, ridge regression, which introduces a small amount of bias, can result in better predictions due to a significant reduction in variance.

Consistency of Estimators

Another important property related to unbiasedness is consistency. An estimator is consistent if it converges to the true parameter value as the sample size increases. While unbiasedness is concerned with the expectation of the estimator for a fixed sample size, consistency is about the behavior of the estimator as the sample size grows. It is possible for an estimator to be biased but consistent if the bias diminishes as the sample size increases.


In summary, an unbiased estimator is a statistical tool that does not systematically deviate from the true parameter it aims to estimate. While unbiasedness is a valuable property, it is not the sole factor to consider when selecting an estimator. The overall performance of an estimator, including its variance and consistency, should be taken into account to ensure the reliability of statistical inference.

Please sign up or login with your details

Forgot password? Click here to reset