DTM.1:5.2 - The same spread permits opposite evaluations

Suppose a conditional population model gives fractions 0.30, 0.25 and 0.45 in three modes of tool use. Scores of 1, 0.5 and 0.1 for a separately specified task give a mean of 0.47. Scores of 1, 1.4 and 0.9 for another justified assessment give 1.055.

The arithmetic is a constructed example, not measurements of users. Its purpose is to expose the missing premise: state fractions do not determine task performance. The practitioner must supply the task, capability criterion and evidence for the scores. A model of spread cannot supply them merely by naming one state “dependent.”