Examining the results in Table 9-7, a general rule seems to emerge. A change in at least 15 percent of the ration items appears to be sufficient to produce an increase in calorie intake and improved consumer perception of the overall ration. The one exception concerns MRE VII (Popper et al., 1987) where the changes to the ration consisted of adding fruit-flavored beverages to all menus and hot sauce to three. In this instance, where both items repeated, the definition of ration change probably overstates the actual change. In addition, the failure of caloric beverages to increase energy intake in this study appears to be the exception (see Engell, Chapter 12 in this volume, for a complete discussion of the issue). It is clear that the definition of the magnitude of ration change needs refinement. However, it is equally clear that the vague term substantial change used earlier can be taken to mean approximately 15 percent.
An obvious corollary to this rule, revealed in Table 9-7, is that the new items must be acceptable to the consumer. In every case where a real improvement in the ration was demonstrated, the new items were highly acceptable. The mean acceptance ratings of the new items in the MRE studies ranged from 7.33 to 8.43. Specification of the exact parameters for ration improvement await further data, but on the basis of the current analysis, a 15 percent change in the ration, where new items are acceptable to the consumer, appears to be a good guideline for the ration developer.
It is not clear from Table 9-7 whether ration changes that led to increased food consumption and consumer acceptance resulted from replacing poorly rated food items by more acceptable ones, or whether these changes were due to increasing the amount of variety within the ration. In two of the three MRE studies described in Table 9-7 (Lester et al., 1993; Popper et al., 1987) these factors are confounded and it is impossible to separate out item replacement effects from variety effects. In the third MRE study (Edwards et al., 1989), only new items were added to the ration in the form of a supplement, and it appears that increased variety was responsible for the improvement in ration performance. The two T Ration studies listed in Table 9-7 can be regarded as item replacement studies, but they do not provide any real evidence on the issue of item replacement versus variety as the magnitude of ration change was probably too small to realistically expect changes in food consumption and overall ration acceptance.
One interpretation of the improved consumption and consumer acceptance in the ration modification studies is that these changes resulted from increased variety in the ration system. In every instance where there was clear evidence for improvement, the changes to the ration involved greater variety. Rolls and her colleagues have convincingly demonstrated that within the context of a