Reliability of Estimates

An estimate is reliable when it is made under conditions that are well supported by the data. In regression, we should first check whether the value being used is within the range of the original data; this is called interpolation and is usually more reliable.

If the value is outside the data range, it is extrapolation, which is less reliable because the same pattern may not continue beyond the observed data. We should also check whether there is a strong linear correlation between the two variables, because regression estimates are only dependable when the data follows a clear linear trend.

So, a reliable estimate usually requires both conditions: the value is within the data range and the variables have a strong linear correlation.

Interpolation - The process of carrying out estimation within the given data range.
Extrapolation – The process of carrying out estimation beyond the given data range.

Flowchart showing reliability of estimates based on interpolation and linear correlation