Determining the required sample size

Sample size is an important concept in statistics. In order to make reliable inferences of a population based upon a sample of that population, it is necessary to have a statistically significant sample size. One method of calculating the required sample size is Cochran's formula, named after the statistician William Gemmell Cochran.

Cochran's formula is as follows:

Here, p is the expected percent of the population with the desired attribute. e is the desired precision or margin of error. Z is something called the z-score, which is found in a z-score table (it's a statistics thing). In order to find the z-score, we must choose a desired confidence value (how confident are we in the results).

This recipe implements Cochran's formula for determining sample size in DAX based upon the chosen confidence levels, the proportion of the population with the desired attribute, and the desired precision.