2008 TRI/PRINCETON CONFERENCE 119 LIGHTNESS DEPENDENCE Technical measures of shine typically produce larger values for dark substrates as absorp- tion tilts the balance between specular and diffuse refl ected light towards the specular. This trend is seen in Figures 3 and 4. The report that judgment of shine depends on skew and not on color (10), however, led us to make this comparison as well. We fi nd, in Figure 3, Table II Luster Results (polarized imaging) on Damaged Hair Hair type Metric Untreated 20000X Comb- damaged Combed vs untreated UV + conditioner Conditioned vs untreated Combed vs conditioned Indian TRI 30.4 30.5 No 33.4 Med. Brown 23.6 24.4 No 25.9 Piedmont 2.6 3.2 No 2.5 No No Red 27.2 24.4 27.2 MB bleached 10.4 11.3 No 10.4 No Dyed black 21.8 21.5 21.1 No No Dyed brown 12.4 12.8 No 7.8 No No Dyed red 14.6 15.0 No 7.9 No No Indian Stamm 93.2 92.3 92.4 No Med. brown 85.9 85.7 83.4 No No Piedmont 22.1 24.8 No 21.6 No No Red 86.5 84.9 84.1 No No MB bleached 64.3 65.8 No 66.6 Dyed black 84.6 83.7 76.7 No No Dyed brown 64.5 62.9 51.9 No No Dyed red 71.2 70.0 53.8 No No Indian Reich- Robbins 275.0 245.4 272.1 No Med. brown 103.4 105.4 No 96.6 No No Piedmont 2.1 2.7 No 2.0 No No Red 124.9 99.8 105.6 No MB bleached 17.9 20.4 No 19.2 No Dyed black 87.4 81.6 56.5 No No Dyed brown 21.6 21.4 10.0 No No Dyed red 31.6 31.0 10.6 No No Indian Guiolet 13.7 12.0 12.2 No Med. brown 6.1 6.0 5.0 No No Piedmont 0.3 0.3 No 0.3 No No Red 6.4 5.6 5.3 No No MB bleached 1.8 1.9 No 2.0 Dyed black 5.5 5.1 3.3 No No Dyed brown 1.8 1.7 1.1 No No Dyed red 2.5 2.3 1.2 No No Indian Boosa-Nova 46.9 38.8 53.3 Med. Brown 26.2 26.9 No 27.1 Piedmont 1.1 1.4 No 1.1 No No Red 32.5 24.4 30.5 No MB bleached 7.9 8.2 No 8.6 Dyed black 27.0 22.0 20.2 No No Dyed brown 9.6 8.3 6.1 No No Dyed red 14.2 11.1 6.2 No No The text in the comparison column indicates if the directional change matches the perceptual results in Figure 6.

JOURNAL OF COSMETIC SCIENCE 120 that the histogram skew produced by hair varies inversely with the lightness of the hair. The fact that this behavior is single valued, unlike the other measures, leads us to specu- late that there is an intrinsic relationship between skew and hair color. We have at- tempted to model this dependence and use variations from the model as predictors of the small changes in shine with damage and with control treatments to no avail. Skew, how- ever, remains an interesting quantity for future investigations. PERCEPTUAL CONSISTENCY The tests of perceptual consistency in this paper are stringent in that all treatments and damage produced only very small changes in tress appearance. It is well established that Figure 6. Perception of shine of damaged and treated hair. (a) Choice of shinier tress (combed vs untreated) for various hair colors in 2AFC test. (b) Choice of shinier tress (UV exposed + conditioner vs untreated) for various hair colors in 2AFC test.