94 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS EVALUATORS Twelve women, 20 to 65 years old, were randomly chosen from regular users of moisturizers. These consumer evaluators did not receive formal training, other than the necessary instruction about procedures. With this panel, the conditions for evalua- tion could be better standardized than if individual subjects reported on themselves, and individual user bias in such reporting was eliminated. These 12 judges were to evaluate for overall effect. They were to do so by determining if they could see a difference in appearance between the two sides of the face and, if so, to choose the side with fewer lines and smoother appearance. If both sides were equal in appearance, the evaluators were to so indicate. The subjects were evaluated indi- vidually in a random sequence, with separate observation forms, at a uniform distance of 2 ft, under uniform lighting conditions of light from above, with the same interval between test series (one week), and at the same time of day so as to minimize the effect of any daily variation on subjects or judges. RESULTS By the end of the 6 weeks, 864 evaluations had been completed, at the rate of 144 ap- praisals (12 evaluators times 12 subjects) for each weekly test series. The number of judgments favoring a specific treatment, as manifested 3 hr after the last application and as revealed after breaking of the code, is shown in Fig. 1. The major features of M01STURIZER 0 WATER BOTH SIDES EOUAL MOISTURE LOTION ML WATER BOTH SLOES EOUAL EXPER. FORM. EF WATER BOTH SIDES EQUAL MOISTURE FILM MF WATER BOTH SLOES EQUAL LOTION L WATER BOTH SlDES EOUAL NO TREATMENT WATER BOTH SIDES EQUAL 0 20 40 60 80 100 120 144 NUMBER OF FAVORABLE EVALUATIONS Figure 1. Overall results by treatment (total of 864 single-blind evaluations--144 in each of 6 randomized test series)

MOISTURIZERS AND FACIAL LINES 95 these results were: a) Water was judged to have a perceptible effect and was significantly better (P 0.01) than no treatment. In turn, moisturizers O, ML, and EF were significantly su- perior to water (P 0.001, P 0.001, and P 0.01 respectively). Moisturizers MF and L were not significantly different from water. b) A fairly distinct ranking developed in the extent to which any one of.these prepara- tions achieved a possible maximum of 144 favorable evaluations when compared against water. In percentage form and in absolute numbers this ranking was as follows: Identity Code Oil of Olay © O Commercial Moisture Lotion ML Experimental Formulation EF Commercial Moisture Film MF Commercial Lotion L % Favorable Evaluations 83% (119 out of 144) 76% (109/144) 64% (92/144) 55% (79/144) 53% (77/144) There was a numerical trend toward moisturizer O being assessed more favorably more often than moisturizer ML, although the difference was .not statistically significant at the 95 per cent confidence level. Moisturizer O was superior to and differed statistically from moisturizers EF, MF, and L (P 0.001), and similarly, ML differed from MF and L (P 0.01). c) Although water was assessed less favorably than the five cosmetic formulations, it was assessed more favorably than no treatment (89 vs. 51). Table I shows the individual components of the panel ratings. The effect of the two moisturizers with the greatest lead over their control was manifested in that for every subject a clear majority of the evaluators chose the side of the face treated with one of these agents. Other preparations were proportionately less effective as their lead gradually eroded away throughout the group of subjects, rather than this lead making an abrupt, pronounced transition in particular subjects. In other words, the extent of effectiveness existing for a particular cosmetic tended to prevail throughout the study group, and was not concentrated in particular subjects. Table I also provides the opportunity to examine the results in terms of subjects, and thereby to test the validity of the method. Overall, for the six-test series, there was no erratic pattern indicative of a failure of randomization. Figure 2 shows the distribution of favorable evaluations by the 12-member evaluator panel for each test agent. The values clustered within one-third to less than one-half of the possible range. DISCUSSION Application of many of those same features, which differentiate the controlled clinical study from an open investigation, to the method of evaluation by a nonprofessional and untrained consumer panel has similarly increased the value of the results from this lat- ter mode of inquiry. These features, in conjunction with a qualitative appraisal of com- parative overall performance, appropriate for use by nonprofessional consumer evalua-