Average a dataset's smallest and largest values to locate the center of its range. The calculation trail makes the reported midrange easier to reproduce.
Midrange is useful for a quick center estimate when endpoints are meaningful, such as a tolerance interval or a bounded measurement span. It is also a convenient check on the center of a symmetric range.
Midrange is the arithmetic mean of a dataset's minimum and maximum. It uses only the two endpoints, so it describes the center of the observed range rather than the balance point of every observation.
The operation is a compact expression of the midrange definition above.
For 12, 18, 21, 27, and 42, the minimum is 12 and maximum is 42. Their average is (12 + 42)/2 = 27, even though the arithmetic mean of all five values is 24. This Midrange example can be compared with spread around the mean.
Because only two observations determine midrange, one extreme outlier can move it sharply while every interior value remains unchanged. It should not replace mean or median when the full distribution matters. If the Midrange assumptions do not fit, consider mean and median.
Review the sign, scale, and unit of midrange after entering data values.
Scan the list for its smallest and largest values, add those endpoints, and divide by two. Sorting is optional, but it makes endpoint selection easier to verify in a short dataset.
Source values may arrive in a different order from the form. Map them explicitly to data values, normalize units, and retain enough precision for the next step after midrange.
Changing an interior observation leaves midrange unchanged, while moving either endpoint shifts the result by half that endpoint's change. Moving both endpoints equally shifts midrange by the same amount. Watching this response separates a data-entry mistake from an unexpected but valid value.
Midrange averages the endpoints, while range subtracts them and arithmetic mean uses every observation. These measures can coincide in symmetric data but describe different features. This page deliberately reports only the former interpretation.
If data values are exact counts, more result digits may be meaningful than when they are measured approximations. Let the least certain source guide the final presentation.
Label the output as midrange in notes or tables. Store data values beside it when the result will be reused in a later stage.
No. Range subtracts minimum from maximum; midrange averages them.
No. Only a change to the minimum or maximum can change it.
It always lies halfway between the observed endpoints, though it need not be an observed value.
It can be poor when endpoints are outliers or the distribution is strongly skewed.