Educational computer software

The increasing use of computer technology in schools and households is exposing children to a wide variety of new media resources. Public debate over the effects of computers has most often concerned children's exposure to violent and sexually explicit material in computer games and on the Internet. Less national attention has been paid to other potentially detrimental effects, such as the perpetuation of social biases and stereotyping, and to other forms of technology, such as educational software. Educational software tends to be viewed as more wholesome, "family-oriented," and scholastic than other forms of computer software and is therefore generally ignored as a potential source of negative influence on children. Little systematic research has been conducted on biases and stereotypes found in educational software, despite its widespread use. Therefore, the current study was designed to investigate one form of stereotyping that may occur in this educational medium--gender stereotyping.

THE ROLE OF GENDER STEREOTYPES

Feminist scholars (e.g., Kilbourne, 1999; Ruscher, 2001; Tavris, 1992) have asserted that feminine and masculine gender role stereotypes are constraining and therefore limit individuals' potential. Pervasive negative stereotyping in a culture serves to normalize biased portrayals of individuals, thereby preserving the status quo (French, 1992; Ruscher, 2001). In addition, researchers in the field of cognitive psychology have shown that to simplify and conserve mental resources, individuals tend to develop stereotypes, which are then used for filtering, organizing, and remembering information (Macrae, Milne, & Bodenhausen, 1994; Sherman & Frost, 2000). Such stereotypes, as simplified and generalized constructs of complex phenomena, can then affect, often negatively, individuals' attitudes and performance (Aronson, Quinn, & Spencer, 1998; Stangor, Carr, & Kiang, 1998).

STEREOTYPING IN THE MEDIA

One powerful source of stereotyping in societies is the media to which individuals are exposed. Messages concerning stereotyped gender roles are relayed to individuals of all ages and from a multitude of media sources. Research has shown the existence of highly stereotyped gender roles in television aimed at both children and adults (Elasmar, Hasegawa, & Brain, 1999; Huntemann & Morgan, 2001; Signorielli, 2001; Thompson & Zerbinos, 1997), magazines (Vigorito & Curry, 1998; Willemsen, 1998), children's books (Gooden & Gooden, 2001; Oskamp, Kaufman, & Wolterbeek, 1996), comics (Brabant & Mooney, 1997), advertisements (Bartsch, Burnett, Diller, & Rankin-Williams, 2000; Furnham & Mak, 1999), and video games (Dietz, 1998; Subrahmanyam, Kraut, Greenfield, & Gross, 2001).

In the field of education, researchers have provided evidence that gender stereotyping is present in various forms of educational media, such as textbooks (Hogben & Waterman, 1997; Peterson & Kroner, 1992), preschool picture books (Oskamp et al., 1996; Tepper & Cassidy, 1999), award-winning children's literature (Dougherty & Engel, 1987), and educational television (Barner, 1999). Research has also informed teachers and administrators about the ways in which gender stereotyping and biases can negatively impact students and about the importance of inclusion and diversity in education curricula and classroom climate (Mueleners, 2001; Sapon-Shevin, 1996).

The educational media that have taken on an increasingly important role in the past years have been technology-based media, especially in the form of the Internet and educational software. In 1993, with funding from the United States Department of Education, the International Society for Technology in Education (ISTE) drafted its first set of technology standards for teachers (ISTE, 2002). These standards, currently in the third edition, have become known as the National Educational Technology Standards for Teachers, which are used by accredited education programs for preschool through 12th grade teachers. Specifically listed in these national standards are statements that the teacher should (a) "apply technology resources to enable and empower learners with diverse backgrounds, characteristics, and abilities" and (b) "identify and use technology resources that affirm diversity" (ISTE, 2002, p. 9). These standards may in part be an attempt to address the large "technology gap" that exists between the genders (American Association of University Women Educational Foundation, 1998, 2000). Thus, educators have been alerted to the fact that gender stereotyping and biases may have a causal effect on the creation of the technological gender gap.

Research concerning gender stereotyping in computer-based educational resources has been surprisingly lacking, despite the increased use of technology in schools, the ISTE standards that address the importance of affirming diversity with technology, and the large body of literature that shows the negative effects of stereotyping (e.g., Blair & Sanford, 1999; Brown, Steele, & Walsh-Childers, 2002; Signorielli, 2001). Chappell's study (Chappell, 1996) of 17 mathematics educational software packages, Milburn, Carney, and Ramirez's research (Milburn, Carney, & Ramirez, 2001) concerning computer clipart, and Drees and Phye's study of 34 language arts software packages (Drees & Phye, 2001) are the only known content analyses of gender biases in software used in educational settings. All three studies showed that images of girls and women and female characters were portrayed substantially less often than images of boys and men and male characters; however, the studies differed in their findings regarding gender stereotyping.

Chappell's study (Chappell, 1996) of educational software for preschoolers through 12th-graders did not include ratings of characters' genderstereotyped behaviors and traits; however, the researcher assessed the degree to which activities in the software were based on two masculine stereotyped actions--violence and competition. Chapell found that the degree of competition and violence increased with software grade level. Preschool software (n = 3) was virtually free of violence and contained no competition; thus, it was the least masculine gender role stereotyped.

The other studies concerned the extent to which the characters in software exhibit gender stereotyped behaviors and traits. Milburn et al.'s research (Milburn et al., 2001) revealed that clipart presented human characters in highly gender-stereotyped ways. Female characters were more often passive, nurturant, and engaged in feminine stereotyped activities (e.g., setting the table), whereas male characters were more often active, non-nurturant, and engaged in masculine stereotyped activities (e.g., sawing wood). Drees and Phye (2001), however, found no statistically significant (p < .05) differences between male and female characters in terms of gender role stereotyping. They coded children's language arts software using the list of 21 gender-stereotyped traits (e.g., directive, passive, nurturant, competitive) that Oskamp et al. (1996) utilized in their study of preschool picture books. Drees and Phye speculated that their results differed from Oskamp et al.'s partly because of the use of different statistical analyses in the two studies and partly because of an improvement over the past several years in media portrayals of male and female characters.

Based on Gooden and Gooden's recommendation that more content analyses should be performed on preschool educational materials (Gooden & Gooden, 2001), the topic under investigation in the current study is the degree to which gender role stereotyping exists in educational software for preschoolers. Software specifically targeted at young children has been found to contain very little gender role stereotyping (Chappell, 1996; Drees & Phye, 2001) even though other media aimed at this age group have been found to be highly gender stereotyped (Barner, 1999; Dougherty & Engel, 1987; Oskamp et al., 1996; Tepper & Cassidy, 1999). In the current study, analyses were performed on a greater number of software packages and a broader range of topic areas than those used by Chappell (1996) and by Drees and Phye (2001) in order to increase the generalizability of the results. In addition, a comprehensive coding scheme was developed in accordance with those used by others, so that the current findings could be compared to those of previous studies.

METHOD

Sample