“…Kohlberg (1966) suggested that there is a 2-year lag between the time a child learns to self-label correctly and the time he or she learns to label others correctly according to social-conventional cues. In fact, although gender constancy for self and others are significantly correlated (e.g., Zucker, Bradley, Kuksis, et al, 1999), children evidence gender constancy for self earlier than they evidence gender constancy for others (e.g., Bem, 1989;Gouze & Nadelman, 1980;Marcus & Overton, 1978;Szkrybalo & Ruble, 1999;Wehrens & De Lisi, 1983) and especially for opposite-sex others (Leonard & Archer, 1989). For instance, Shields and Duveen (1986) found that although the majority of children aged 3½-5½ indicated that their own sex could not change, 80% of them thought that the gender of another child, shown to them in a figure drawing, could change with variations in external appearance (i.e.…”