Eschholz trained three individuals to code the characters in each movie. Lead characters were specifically looked at; leads were determined by billing on the movie. The coders looked at several factors when analyzing each film. Prestige was determined by using the 198o U.S. Census Bureau Occupational Categories and the corresponding NORC/GSS 1989 Occupational Prestige Score (Eschholz et al., 2002, p.310). Occupation was gender determined by the top ten positions men held in 1990 or if literature had determined a job to be male. The same procedure was used to determine female jobs. If a job seemed neither male nor female they were put into a gender-neutral category. Age of characters was looked into as well. Gender traits were also examined. Here the Bem Sex Role Inventory was used as a basis for gender traits. The list of traits was downsized to those traits that could be observed rather than self reported because the subjects were film characters (Eschholz et al., 2002, pp.311-312). .
Many interesting results came out of this study. Not surprisingly females are underrepresented in movies. In 1990 women made up 51 percent of the population according to the U.S. Census Bureau, but only appear in leading roles 35 percent of the time (as cited in Eschholz et al., 2002, p.314). Age played an interesting role in characters as well. "Thirty three percent of female characters were 30 years of age or younger, compared to 13 percent of the male characters- (Eschholz et al., 2002, p.315). When the age is changed to 50, ten percent of male characters were over fifty, compared to only four percent of female characters. The U.S. Census Bureau (1990) estimated that men over fifty made up 23.7% of the population and women were 28.2% (Eschholz et al., 2002, p. 315).
Employment rates for male characters in movies were higher than those of female characters (Eschholz et al., 2002, p.317). An interesting point made is that 29 percent of white women were portrayed as unemployed but having the option to work.