Sample Size Neglect

In some countries, you can be arrested by the Stats Police for sample size neglect. Always make sure your samples have ample food, shelter, and unconditional love.

Or more accurately, sample size neglect happens when we try to draw conclusions about a population from a sample that is just too small. A too-small sample might have just one data point that is extremely large (or small). Consequently, when we calculate the sample means and/or standard deviations of those samples, they may be not at all close to the mean and/or standard deviation of the population as a whole.

When we use these erroneous results to make our decisions, we’ll probably make bad decisions. Larger samples can also have those extremely high or low values, but they’re often balanced out by extremely low or high values also included in the sample because there’s just more data there all together.

Of course, it’s possible that even a large sample will randomly give us a mean and/or standard deviation that is far from the population’s mean and/or standard deviation, but it’s just much less likely to happen with large samples than smaller ones. How small is too small? Well, there are loads of websites and books out there that will tell you how large your sample should be for various populations sizes and statistical processes.