The null hypothesis for ipsative measurement would generate an expected
average point allocation of 4.5 points (100/22=4.5) per issue or need. (The
total of twenty-two comes from the twenty-one given plus approximately one
additional issue – actually, 0.6— suggested.) The first eight issues listed in
Table 1 exceed 4.5, and are thus considered as relevant priorities for initial
implementation efforts. Strict statistical analyses would focus more attention
on the actual distribution of point allocations; traditional Pareto or scree test
analyses would examine the breakpoint of mean scores for the point allocations
to use as an appropriate number of items for selection (Box et al., 1978).