##### Page tree
Go to start of banner

# Estimating the mean of a Normal distribution when the distribution's standard deviation is known

Occasionally, it is possible that the mean is unknown but the standard deviation is known. For example, the manual of a laser ruler that architects use might specify the standard deviation of the error at a certain measured distance. For a given set of n data values randomly sampled from a Normal distribution, with unknown mean m and known standard deviation s, the distribution of uncertainty of the true mean is calculated from a Normal distribution:

$\mu=Normal\Big(\bar{x},\sigma \Big/ \sqrt{n}\big)$                                    (1)

This can be rewritten as:

$\mu=Normal(0,1)*\sigma \Big/_{\sqrt{n}}+\bar{x}$                          (2)

The model Estimate Mean for Normal Distribution When StDev is Known lets you generate values for the above uncertainty distribution for m for a data set.

The links to the Estimate Mean for Normal Distribution When StDev is Known software specific models are provided here:

Crystal Ball Estimate_mean_for_Normal_distribution_when_std_is_known

@Risk Estimate_mean_for_Normal_distribution_when_std_is_known

Since the unit Normal distribution has a standard deviation smaller than the Student distribution, Equation 2 provides a narrower range of uncertainty, especially at low n, compared with the situation where the standard deviation is unknown, which makes sense because we know more under these circumstances than when we did not know the standard deviation.

Comparison with the Bayesian approach

The Bayesian derivation of the same formula as Equation 1 is given here.

• No labels