grouping error

n. An experimental error caused by the manner in which the data were combined or grouped. In most statistical procedures there are assumptions that data in a group or range have an average near the median of the range and a normal distribution. If these assumptions are not met, distortion is introduced into the data. Thus, if we have two groups of children, one of which includes children from 1 to 5 years of age and one which includes children from 6 to 19 years of age, we would expect different results if one experiment had groups with nothing but 5- and 6-year-olds in it and in another experiment the groups had nothing but 1- and 10-year-olds in it.