60 JOURNAL OF COSMETIC SCIENCE or longer data from these theme park studies to the same criterion from the US Women's Hair Length Survey taken in 1972 (see Table I). This older telephone survey of women selected randomly across the US, from ages 14 to 60, indicated that 24% of the women or 12% of the population (assuming 50% men in the US population) rated their hair as shoulder-length or longer, which is similar but slight lower than the 13.2% at shoulder- length or longer found in this theme park study. The slightly higher value of the theme park study may be a result of the difference in time (1972 vs 2001) or a slightly younger population in the. theme park study. Nevertheless, these results suggest that the theme park data represent a good approximation for the percentages of persons with different hair lengths in the US population. Possible complications from hairstyles and headcovers. Since our hair length estimations are by observation at a distance and do not involve direct contact with the subjects, hairstyles and headcovers or hats could conceivably interfere with free-falling hair and thus provide the potential to interfere with our hair length estimations. For example, in some in- stances hair worn in a bun or a related hairstyle interferes to the extent that we do not count a person with this type of style. Therefore, we decided to try to obtain a quan- titative estimate of the percentage of people at theme parks that wear (a) hairstyles or (b) headcovers that might affect our data. (a) Hairstyles Table IV shows that about 3 percent of the people observed had a hairstyle that interfered with our estimation of hair length by anatomical site, and we therefore did not count these persons. Even though the exact length could not be estimated for this 3% of the people, the vast majority of these were clearly in the chin-to-shoulder or shoulder- length groups. In fact, we believe it is safe to say from the size of the "bun or coil of hair" that none of these subjects had waist-length or longer hair. Therefore, if a correction needs to be applied to our data for hairstyle, it most likely would increase the percent- ages for the shoulder- or chin-to-shoulder-length groups rather than affect the longer hair estimates, i.e., waist-length or longer. (b) Headcovers and hats On May 9, we went to Epcot to try to estimate the percentages of persons who wear headcovers that might possibly interfere with our estimates of free-hanging hair lengths. On this day we counted 2,470 persons and found only five (0.2%) that were wearing headcovers that interfered with our hair length estimation (see Table V). From the type of scarf worn, the hair of these five individuals could be seen to be short, that is clearly not waist-length or longer. Interestingly, on this day we did not observe Table IV Types of Hair Styles Type of style Number of persons Percentage Free-hanging hair styles 1322 84.8% Ponytail or braid(s) 190 12.2% Bun or interfering style 47 3.0% Totals 1559 100.0%

HAIR LENGTH IN THE USA 61 Table V Headcovers at Theme Parks Type of headcover Number of persons Percentage No headcover 1790 72.47 % Caps a 604 24.45 % Hats 70 2.83% Scarves/hoods 6 b 0.24% Totals 2470 99.99% Includes baseball-type caps and sun-visor-type caps. One person wore a scarf, but her hair could be seen hanging to her shoulders. any persons wearing a hood or a headcover of any type that could cover very long hair, whereas during a few of the other studies, we did note one person in each of two studies with a hood that may or may not have been covering extra long hair. One of these persons appeared to be a Muslim and the other was of Indian (Asian) origin. From the 2.83% of persons wearing hats, we could clearly see free-hanging hair in all cases. For two persons, the hat was worn at a steep angle, covering part of the neck nevertheless, a sufficient amount of the hair/hairline could still be seen and an estimate of hair length could still be made. From these data and our notes in the previous studies, we conclude that headcovers did not interfere with our hair length estimates. Shorter hair. To obtain some limited data on shorter hair we decided to count approxi- mately 1000 persons, using the three groupings described in the hair survey of Table I, and we did this at Disney's MGM Studios. Chin-length or shorter -- short hair chin- to-shoulder length -- medium-length hair shoulder-length or longer = long hair. Table ¾I summarizes the data that we obtained for these three groups in this abb'reviated study. The shoulder-length or longer hair at 15.9% is in reasonable agreement with the value of 14.53% for the long hair studied at MGM studios (see Table II). Obviously, a larger amount of data is necessary to provide greater reliability for these numbers for shorter hair. However, these data show that somewhere in the vicinity of 60 percent of the people in the US have hair chin-length or shorter and that about 20-some percent have hair of medium length. The larger group of 62% with chin- or-shorter- length hair is probably largely comprised of males who wear predominately shorter hairstyles than females. SUMMARY Variables affecting hair length are summarized along with data on abnormally long hair from several different sources. By observing the hair of 24,300 adults in central Florida Table VI Short, Medium-Length, and Long Hair Groups Theme park Shoulder-length or longer Chin-to-shoulder length Chin-length or shorter MGM Studios 185 (15.9%) 257 (22.1%) 721 (62.0%) N= 1163.