In fact, some elements of the scales are currently based only on clinical experience rather than on any empirical data. Thus, for example, the presence of unicorns in the human figure drawing is taken as a potential indicator of molestation. So are Xs used in the place of eyes. Even when elements have some empirical data to support their use as trauma indicators, these data may be scant. The authors report, for instance, that butterflies were found in 4.4 percent of drawings made by molested children versus less than .8 percent of drawings by all other children, and so they suggest that butterflies might be taken as a possible indicator.