Collin Fellows


Gender Roles and Representation in Children’s Picture Books:

A survey of Caldecott award winners and honorees from 1992 to 2004


Main research questions


Our primary research question is: What do children learn about gender from preschool picture books? To answer this question, we looked at three areas of research that have been done in the past; gender visibility, stereotypical role characteristics and racial/cultural diversity. In addition, we explored the additional variables of non-human characters, and the gender of the author/illustrator.

Why is this topic of concern?


The header announcement for the 2004 Amelia Bloomer Project Awards sums up the importance of this research topic.

As the political and social climate in the United States becomes more conservative, the need for books depicting strong, self-actualizing girls and women grows. While reproductive rights are challenged and the basic right to free expression is diminished, it is more important than ever for children and teens to read about girls and women who solve problems, gain personal power, and empower others to reach their goals. The Amelia Bloomer Project, sponsored by the Feminist Task Force of the Social Responsibilities Round Table of the American Library Association, is proud to announce the third annual Amelia Bloomer Project list of books that celebrate girls and women as a vibrant, vital force in the world. (http://www.libr.org/FTF/bloomer.html) Research by Brown (1956) and Hartley (1960) shows that children have a primary understanding of gender appropriate behavior by the time they are kindergarten age. At only four years old "children realize that the primary feminine role is housekeeping, and the primary masculine role is wage earning" (cited in Weitzman, Eifler, Hokada, and Ross, 1972, p. 1125). Picture books are often a child’s first exposure to a world outside their parents influence. This is a time of transition for children. These books are aimed at children during the developmental period (ages 3-9) where, in the 1930s George Herbert Mead explained that they are learning to see "their self-activity in relationship to a small group … (and) they are able to take part in more complex social activity" (Ashley and Orenstein, 1998, p.479). This is a time when messages of the normative gender roles, though not overt, do come across. Children have learned how one family unit is structured from their parents, but have not yet entered the world of structured learning in school. It is important for both parents and writers to understand the messages of sex-role expectations, attitudes toward cultural diversity, and relative worth of men and women that children are receiving at this formative age. In a time where there is so much emphasis on gender equality, and recognition of the imbalance of access to resources based on gender, it is crucial to become aware of the socialization that children receive from these seemingly innocent books.

Literature Review

The easiest measurement of gender inequality is a simple count of the number of boys versus girls within the illustrations. The research in this area is rich. Beginning in the 1970s, researchers began to document the unequal frequency of female characters in the illustrations within picture books. Weitzman et al. (1972) found that during a five year period in the late 1960s, Caldecott winning books showed 261 pictures of males, but only 23 pictures of females. As Clark, Lennen, and Morris (2005) note in their research, visibility, measured by a count of single character, female illustrations in Caldecott winning books grew from 11.7% in the 1960s to 37.1% in the early 1980s. In order to update prior studies Kortenhaus (1993) reanalyzed all Caldecott books from the 1940s through the 1980s and found that in spite of criticisms of methodology, earlier research findings held generally true. They found a linear decline in sexism for all decades until the 1970s, when they saw the ratio of male to female character level out below 2:1. This ratio was down sharply from 5:1 in the early decades.

These numbers improved as we moved into the 1990s. Oskamp, Kaufman and Wolterbeek (1996) found that the single female character percentage (43%) had increased, and the balance between total male and female characters depicted had come much closer (72 females, 79 males). Gooden and Gooden (2001) found much the same thing for the years 1995-1999, that far more males were portrayed alone than females. Females were portrayed alone in 19% (280 of 1464) of the illustrations while males were portrayed in 23% (337) of the illustrations. When they included animals that difference increased, with female illustrations totaling 24% (358) and male illustrations totaling 31% (453)

Based on the trend found in prior research, we expected to see in an analysis of Caldecott Award winners and honorees from 1992-2004, even more parity between the number of female and male characters depicted.

Even though research shows that the number of female characters is coming into balance with males, it also finds that both genders are limited in the available role scripts that they see. Three areas have been studied to show stereotypes are reproduced in children’s book illustrations; artifacts used, location depicted, and behavioral characteristics.

Boys use shovels, girls use vacuums. This may seem trite and cliché, but when Crabb and Bielawski (1994) looked to representations of material culture, they found that objects tended to be very gender-marked. They defined material culture as "that part of the environment that has been intentionally constructed for practical purposes according to dictated plans" (p. 1), and broke these artifacts down to household, production, and personal, and looked to see if the use of these tools was restricted by gender. It is not surprising that they found the proportion of female characters shown using household artifacts was larger than the corresponding proportion of male characters, and that the proportion of male characters shown using production artifacts was larger than the corresponding proportion of female characters.

The history of behavioral attribution in all of these past studies has not made replication of this variable an easy task. In the earliest study of Caldecott winning books, Weitzman analyzed behavioral traits. Davis (1984) criticized Weitzman’s study for failure to provide operational definitions and for questionable data gathering methodology. Using the refined set of values for behavioral trait that Davis developed, Williams (1987) tried to address the methodological issues of Weitzman’s study by only coding a trait that was present in an illustration, and addressing issues of intercoder reliability. Oskamp et al. (1996) further reduced the number of values for this variable from 20 to 13, (Dependent, Independent, Cooperative, Competitive, Directive, Submissive, Persistent, Explorative, Creative, Imitative, Nurturant, Aggressive, Emotional, Active and Passive). Their study found four attributes that showed significant male-female differences: dependent, submissive, independent and creative. This was considerably less than the number of significant gender differences in the Williams, Williams, and Malecha (1987) study of 1980-1985 Caldecott winners. In contrast to Williams, they found female characters being portrayed as fuller individuals with a greater number and variety of attributes. Results have differed greatly in all studies using behavioral traits and we chose to reduce, yet again, the number of observed traits to reduce the potential for error. One of the observed characteristics that has remained relatively consistent over past research has been level of activity.

Another area that researchers found gender restrictive portrayal is the physical location of the characters in an illustration. Weitzman et al. (1972) commented on their observation that girls were more often found indoors. "While boys play in the real world, girls sit and watch them—cut off from that world by the window, porch, or fence around their home" (p. 1133). Oskamp (1996) revisited this and found that by the 1980s, the trend had reversed, with only 50% of girls shown indoors compared to 69% of boys.

In areas of representation of gender we expected that, while balance may be in sight for location and activity level, production and domestic artifacts would still have gendered assignments.

Although it was not one of the primary research questions of their study, Oskamp et al. (1996) noted that only eight of the twenty-two books that they analyzed contained any "cross-cultural themes, touched upon life in different nations, racial groups, or religions"(p. 9).

We believed that due to the rapid expansion of global economy and the increased awareness of different cultures, our sample would show a marked increase in the number of books showing cultural diversity.

Diane Turner-Bowker (1996) demonstrated that there is no statistical difference in sexism in children’s books based solely on the author’s/illustrator’s gender. While Turner-Bowker used written text in her study, we believed that the same would be true when the books were analyzed using illustrations as the unit of analysis.

This review of the literature from previous studies provided the groundwork for the five following hypotheses of our study:

Hypothesis 1: There will be little or no difference in the total number of female and male characters in the illustrations.

Hypothesis 2: Girls will be overrepresented in depiction of use of household artifacts while boys will be overrepresented in depiction of use of production artifacts.

Hypothesis 3: Behavioral characteristics will be less gender matched than was found in previous research.

Hypothesis 4: Girls and boys will be equally as likely to be portrayed indoors.

Hypothesis 5: There will be a marked increase, over previous studies, of the number of books which show diverse cultures.

As many of our hypotheses are replication studies of prior research, analysis of the trends over time shown in the original study will be compared to our findings.

Data and Methodology


The goal of this study was to update prior research and look for changes in the trends that have been noted by other research teams. For the purpose of comparison all 56 winners and honorees of the Caldecott Award from 1992-2004 were chosen. (See Appendix A for a complete list of Books)

According to the Children’s Book Council (2005) there are roughly 5,000 children’s books published every year, and while the Caldecott Award winners and honorees may not be a large percent of that population, we agree with previous researchers that they do represent the most balanced and consistent sub-sample available. By selecting these recognized books, we are able to avoid biases that arise from the huge marketing campaigns that would accompany books with ties to popular movies or T.V. shows, continuations of a popular book series, or the size and prestige of the particular publisher.

Weitzman (1972), Davis (1987) and Kortenhaus (1993) all pulled sample books from the general population to be sure that Caldecott books were representative of all children’s books of the same time period. While minor differences between award winners and non-winners existed for all three of these studies, for each of them the Caldecott awardees were found to be slightly less sexist.

The bulk of our research focused on the visual presentation of gender to children, and the Caldecott winners and honorees allowed us to look at a wide range of books selected primarily because of the quality of the illustration. According to the Association for Library Service to Children (2005) the rules set by the Caldecott committee explain that, "A picture book for children is distinguished from other books with illustrations, is one that essentially provides the child with a visual experience. A picture book has a collective unity of story-line, theme, or concept, developed through the series of pictures of which the book is comprised." Award winners and honorees are selected on the basis of "excellence of execution in the artistic technique employed," and "excellence of pictorial interpretation of story, theme, or concept; of appropriateness of style of illustration to the story, theme or concept; of delineation of plot, theme, characters, setting mood or information through the pictures." (See Appendix B for Complete Rules and Regulations of the Caldecott Awards)

Much of our research used illustrations as the unit of analysis (n=1403). We considered each complete picture as one illustration. In cases where the picture covered two pages the illustration was counted only one time. In cases where there were multiple, distinct pictures on a page each picture was counted as one illustration.

For the analysis of manifest content, all illustrations were coded for: Number of females, number of males, race of central characters, and gender of central characters. Our study also collected data for gendered and ambiguous anthropomorphism (attribution of human motivation, characteristics, or behavior to inanimate objects, animals, or natural phenomena). We believe that clearly gendered animals can have as strong a socializing impact on children as do human characters.

Latent content such as diversity, activity level, location and artifact use was confirmed by a system of double coding. Each illustration was evaluated by two independent coders; in the event of disagreement on one of these variables the variable was discussed by the research team to determine the appropriate code.

The gender of the author and/or illustrator was determined by either biographical data sheets, book jacket flaps, or by request to the publisher. In cases where the gender was not identifiable, the book was not included in the analysis of this hypothesis.


For "gender" we limited our analysis to those characters whose gender was immediately identifiable, and of sufficient size in the illustration to be recognizable. In cases where the gender of the character was not clear from the illustration we confirmed it with the text of the book. Non-human characters were coded for gender using the same guidelines, with the addition of an ambiguous category. Non-human, non-anthropomorphized, characters (i.e. pets, barnyard animals, forest critters) were not counted for the purposes of this study.

"Central character" was determined to mean the character that dominated the illustration, based on focus of activity or size of character, and does not reflect any value assessment of the characters present. "Secondary character" was used in cases where there were two characters that shared the same visual weight. As these two categories were not being compared to each other, no protocol was established for which of these characters would be listed as central.


From Crabb and Bielawski’s (1994) categories of artifact use, we chose the two that have consistently shown large gender representation gaps. (1) Household artifacts, which were defined as human-made objects used to produce effects in the home, including artifacts used in food preparation, cleaning, repair, family care, and home manufacture. (2) Production artifacts were defined as objects used to produce effects outside the household, including artifacts used in construction, agriculture, transportation, and all other work outside the home.

In order to use an aspect of the behavioral characteristics of past studies, we chose to exclude those variable values which have had the most criticism of operationalization or acquisition method, and narrowed our categories to those which have shown the only marked and consistent gendered difference.

In addition to the two remaining variables; active and passively-active, we added a third value of passive, to account for those times when a character is not exhibiting any movement or activity. When limited to the two previous categories, a character that was sewing had the same activity value as one who was sleeping. We believed that it would be important to show these differences, and expect to find that boys will no longer dominate active roles, girls will no longer be relegated to actively-passive roles and that there will be a balance of illustrations showing no activity at all.

While this deviation from prior studies of behavioral traits will make comparison to other studies impossible, we believe that it will be useful in two ways. First it will remove issues of historical and discipline (psychology versus sociology) specific interpretation from this and future studies. And second, this reduction will make this study feasible to future researchers with smaller research teams. The three categories were defined as:

Active: gross motor (large muscle) physical activity, work, play.

Passively active: fine motor (small muscle) activity; alert, attentive, activity, but with minimal physical movement (e.g., reading, talking, walking, holding something).

Passive: (no muscle) activity; required no coordination or effort (standing, sleeping, thinking, daydreaming, watching TV).

"Location" was defined as the place where the central character is located. Two potential locations were coded.

Inside – within human made confines; house, school, room, in a fenced yard, in car.

Outdoors – beyond the confines of human made walls.


While Oskamp (1996) did not give an operational definition of culturally diverse in their research, we believed that this new area of analysis in the study of picture books deserved a deeper look. We divided the variable into three easily distinguishable categories;

Socially Conscious: Shows awareness of social issues in the world such as pollution, poverty, crime, etc.

Racially/Culturally Conscious: Portrays different cultural customs, holidays, religions etc.; celebrates differences

Melting Pot: Portrays characters of different races or groups as a homogenous group; focuses on similarities.

Each book was not limited to one category of sensitivity to diversity although multiple themes were seldom used.

It is also important to note that some of these books were, as Creany (1995) pointed out, retellings of familiar stories or folk tales such as Casey at the Bat, Rapunzel, and John Henry. While we realize that the level of gender and cultural sensitivity did not exist at the time these stories were created, we chose not to exclude them from our sample. We believe that every image that a child is exposed to will have an impact on how they view the world in which they live, and should be included in a study of this nature.



According to the U.S. Census Bureau (2005), in 2000, females made up roughly 51% of the total US population. It is reasonable, then, to expect children’s picture books to depict the visibility of males and females accurately. Based on US demographic data and the literature review which suggests that female representation in children’s picture books is increasing, we not only expected a near equitable proportion of male and female main characters, but also that females would make up at least 50% of all gendered characters.

Although our findings show that less than half of the book titles (n=22) gave any indication of the gender of the central character, the majority of these were male- oriented. 16 contained either male names or terminology used to refer to males such as grandfather or man; in contrast, only 6 titles referred to females. A closer examination also revealed that males were more than twice as likely to be portrayed as main characters. Of the 56 books coded, there were 2 ABC books that contained no characters at all and 10 additional books that contained characters but did not focus on any one character as central. Graph 1 illustrates that of the remaining 44 books, 29 featured a male main character (2 of these books featured two males and 1 featured a male and an ambiguous character sharing the role), while only 9 featured a female main character. These findings do not support our hypothesis, and goes against the work of Oskamp, who found that the gender disparity in main characters between the years 1986 to 1991 had evened out to near equity (9 females and 10 males). One possible explanation for the obvious decrease in female main characters is that the Oskamp study focused on Caldecott winners and honors in a six-year period while we focused on Caldecott winners and honors in a 13-year period. As a result, the Oskamp study had only 20 books in their sample size whereas we examined 56 books. This is almost triple the amount of books and, although it certainly does not explain the disparity in the findings completely, it is a notable difference that may have had an impact on our results.

Our findings also indicate that females only make up about a third of all gendered characters. For the purpose of clarity, we chose not to include characters in illustrations where there were more than 30 characters, as it was extremely difficult to determine the genders with certainty. After making this adjustment, there were 2,538 gendered characters (human and anthropomorphized nonhumans). Of those, 32.5% (n=826) were females and 67.5% (n=1712) were males. Compared with the Oskamp findings the number of female characters has dropped 11.2% since 1991. Table 1 summarizes the changes over time since the Weitzman study (1972). Again, our findings go against the general results of previous research, which suggested that female visibility was increasing. One possible thought on this discrepancy in findings is that we noticed a high number of non-anthropomorphized nonhumans and ambiguously gendered humans in our study. This led us to wonder whether authors and illustrators were resorting to these "safe" depictions of characters in an attempt to avoid the sensitive gender issue. It would have been interesting to examine the Caldecott award and honors books prior to 1992 to see if they portray the same proportion of non-anthropomorphized nonhumans and ambiguously gendered humans, and if the decrease in gendered characters (most noticeably female characters) had any relationship to the number of these non-gendered characters. Unfortunately, the previous research in our literature review did not examine this variable, which left us with no data for comparison.


Table 1: Number of Females portrayed

[a] [b] [c] [d]

1967-1971 1972-1979 1980-1985 1986-1991 1992-2004

Total number 685 1315 1084 1189 826

% female 19.1 32.9 42.2 43.7 32.5

[a] From Weitzman, Eifler, Hokada, & Ross (1972) and Williams, Vernon, Williams, & Malecha (1987).

[b] From Williams et al. (1987).

[c] From Williams et al. (1987).

[d] From Okamp, Kaufman, Wolterbeek, and Atchison (1996)



We were also interested in the possible relationship between the gender of the author and illustrator and the visibility of female characters. Although it is true that the author has the most influence on the gender of the main character, the illustrator has a certain amount of power in deciding how many females and males to include in the background and/or as additional/peripheral characters. In 38 of the 56 books, the author was the illustrator. For these reasons, we felt it was important to examine female visibility and its relationship to the gender of not only the author, but the illustrator as well. Of the 56 Caldecott award and honor books published between the years 1992 and 2004, the genders of the authors were slightly uneven with males predominating; 36 males and 20 females. The illustrators’ gender showed an even greater disparity, with 42 males and 14 females. We expected to see no difference in the visibility of female characters, based solely on the gender of the illustrator and author. However, our findings did not support this hypothesis. The most noticeable difference when examining female visibility by author’s gender was in the count of illustrations that included 1-5 female characters. Of the 480 illustrations with a female author, 53.3% (n=256) featured 1-5 female characters, while in contrast only 26.8% (n=255) of the 950 illustrations with a male author contained 1-5 female characters. When taking the illustrator’s gender into consideration, the most noticeable disparity existed at the no female character count and the 1-5 female character(s) count. Of the 369 illustrations done by a female illustrator, 54.5% (n=201) featured no female character and 45.3% (n=167) featured 1-5 female characters. In comparison, of the 1,061 illustrations done by a male illustrator, 66.4% (n=704) contained 0 female characters and 32.4% (n=344) contained 1-5 female characters. Both female visibility by author’s gender (p<.000) and female visibility by illustrator’s gender (p<.000) are statistically significant. Graphs 2 and 3 summarize the findings for female visibility by both the author and illustrator’s gender.


Actual counts of female and male characters revealed that an under representation of females exists, yet that still leaves much of the bigger picture out. Simply counting the number of female and male characters is not an effective way to capture a true understanding of whether females and males are being represented more equitably. In an attempt to get a more thorough understanding of what children actually learn about gender roles through picture books, we also examined the artifact use, the activity level, and the location of characters central to each illustration.

Prior research by Crabb indicates that although females are no longer invisible, they are still primarily limited to using domestic artifacts such as cleaning, cooking, and sewing supplies. Out of the 88 characters shown using a domestic or production artifact (the latter being items such as construction equipment and shovels), 37 were using domestic artifacts and 51 were using production artifacts. As hypothesized and supported by previous studies, females were overrepresented in the use of domestic artifacts, while males were overrepresented in the use of production artifacts. More specifically, 77.8% (n=28) of the females were depicted using a domestic artifact and only 22.2% (n=8) were depicted using a production artifact. In contrast, 82.7% (n=43) of the males were shown using a production artifact, while only 17.3% (n=9) were shown using a domestic artifact [refer to Graph 4]. Although the sample size that we obtained is small, there is still statistical significance (p<.000) that this observed relationship is real.

In the past, males were often portrayed as doing, while females were delineated to merely being. The findings from the Oskamp study (1986 -1991) indicate that this is changing and that these stereotypical portrayals are beginning to fade. Based on this, we hypothesized that the activity level of characters would be relatively equal for both males and females. Of the 1,065 characters central to the illustrations, 344 were females and 721 were males. Graph 5 summarizes the findings and indicates that a disparity does exist in the activity level of male and female characters, but this disparity is small. Male characters were slightly more often portrayed as active (16.1%) than female characters (10.6%) and slightly less often portrayed as passive (25.8%) than female characters (32.3%). Our results indicate that the gap in the portrayal of female and male activity level is beginning to close (p<.017).

The Oskamp study showed a marked change over the years concerning the locations of characters based on gender. Traditionally, males were depicted outside having adventures, while females were isolated behind the safety of walls. By the early 1990s, males were being portrayed indoors more often and there was an increase in the percentage of females portrayed outside. We suspected that this trend would continue and that there would be a relatively even distribution of males and females depicted in both inside and outside locations. Of the 914 central characters (141 were eliminated from this aspect of the study due to undistinguishable location), 298 were females and 616 were males. 50.7% (n=151) of the females were located outside and 49.3% (n=147) were located indoors. In comparison, 61.2% (n=377) of the males were located outside and 38.8% (n=239) were located indoors [see Graph 6]. Although the data suggests that males are still portrayed outdoors more often and females are portrayed indoors more often (p<.003), it appears that this gap of location portrayal is closing.


Although the focus of our study was on gender representation, we were also interested to examine whether there was an increase in books that featured culturally/racially diverse themes. Out of the 56 books; 36 featured no particular theme, 13 featured a racially/culturally conscious theme, 3 featured a melting pot theme, 2 featured a socially conscious theme, and the remaining 2 featured a mixture of both racially/culturally conscious and melting pot. Oskamp et. Al (1997) observed that only 8 of their 22 (36%) books contained culturally sensitive themes. Oskamp did not differentiate between those books that focused on celebrating differences or commonalities, so for purposes of comparison we combined our two dimensions of racially/culturally aware themes. We were disappointed to find that the hypothesis about there being a greater visibility of nontraditional cultures was not supported in this study. We found that only 18 of the 56 (32%) books that we coded for theme indicated cultural or racial sensitivity of any sort. (See Appendices C, D and E for coding sheets and codebook)


Our research continues a body of research studying the stereotyping of gender within children's picture books from beginning in the 1970s, more specifically, what do children learn about gender from preschool literature? The areas investigated specific to our question were the matters of gender visibility, stereotypical role characteristics of females and males, and an inquiry into the expansion of racial and cultural diversity taking place. It is important that these types of studies continue in order to review the social conditioning of the young (such as gender stereotype reinforcement or breaking away from the markers that dictate 'male' or 'female' behaviors) that occurs within the most universal of training devices- children's books.

Gender stereotyping operates to limit a child's prospective development and promise, and books will help shape their mind-set not only about themselves but also about others around them. Historically, the most influential and successful way to communicate cultural values and attitudes is through storytelling, and as we are predominately a literate society, this method now includes children's books. Because of the powerful influence of children's books, it is clear that the characters portrayed in this literature shape the notions that children hold pertaining to customary social roles and ideals, giving them a guide to how boys or girls (or mommy and daddy) are expected to behave, or by indication of visibility, how valuable they may (or may not be) within the social context.
A review of prior research showed a hopeful trend in the visibility of females in central roles. From a low of 11% in the 1960s (Weitzman, 1972) the portrayal of female characters rose steadily through the 1980s (Oskamp, 1996) to a high of 45%. Unfortunately, this trend seems to have peaked, and may now be on the decline. Our research showed significant ground lost in the struggle for visibility.

Acceptable use of artifacts is one are where an amplification of the existing coding tools could likely yield valuable information. Our results, using simply domestic versus production, showed clear rules being set for future generations. Future studies might consider expanding this variable to include, gendered accessories (parasols, jewelry, dolls, slingshots, marbles, etc.) sporting equipment, intellectual stimuli, or any number of other items that might show gendered limitations .It would also be a valuable addition to future studies to show how artifact usage is modeled for children by the behavior of an adult within the children’s books, as "one medium which may both express, and model social representations of material culture … and gender" (Crabb, 1994). In order to educate young children that women are equally as competent as their counterparts, it is essential that the roles and definitions thereof within children’s literature be updated to reflect this reality, which in turn will provide true opportunities to realize their human potential. Girls need to be exposed to images of strong women who are willing and able to use production artifacts alongside their male counterparts.

It was our intention to assess the aspect of culturally sensitive themes within the Caldecott books from 1992-2004, this is a relatively new direction in children’s literature research, and encompasses areas concerning diversity of race, religions, and life in other nations. Our research concluded that 34% of the literature in our sample possessed such sensitivity themes. Although the global economy and influence is effectively bringing the outside world closer to us, there was a lack of evidence revealing that children's literature is responding to this turn of events in kind. An awareness of others unlike them is an invaluable quality for children to possess. in order to foster healthy relations and to break existing prejudices of gender, race, and ethnicity. This quality should absolutely be edified and internalized by younger generations. Because of the importance of awareness of differing cultures and populations, it would be beneficial to conduct further studies of such themes.

There are also other genre of literature worth applying this level of critical analysis to: preschool materials, elementary fundamental readers and early education textbooks- science and math books in particular

The issue of language is also ripe for study, as gender stereotypes tend to go from the overt message to covert implication, the use of ‘differential language’ is an important area as it lends itself to the more subtle approach concerning that of gender role determination. There is an awesome power wielded by those authors and illustrators, enmeshed in the world of children’s literature to influence the attitudes and behavior of the future. Books aimed at the most vulnerable and impressionable members of our population "can serve as a vehicle to perpetuate or abandon stereotypes" (Turner-Bowker, 1996), and their collective responsibility could not be greater. (For additional, subjective observations, see Appendix F)
In summary, this study showed that the presentation of gender to children in picture books is not representative of the world in which we live. If we want our children, especially girls, to grow up with a realistic view of the world, it is imperative that we change the ways that females are portrayed in picture books. Children form rigid gendered stereotypes by the age of five (Turner-Bowker, 1996), so the images that they are exposed to during this impressionable time are instrumental in determining what those stereotypes will be. In order to continue on the path of equality between the sexes, we need to have the literature our children are exposed to reflect this equality.

            *I would like to acknowledge the other researchers involved in this project, and to thank Dr. Daniel Sullivan for his guidance throughout this research.



Ashley, D., Orenstein, D. M. (1998). Sociological Theory: Classical Statements. Needham Heights: Allyn and Bacon

Association for Library Service to Children (ALSC). Terms and criteria for the Caldecott Award. Retrieved March 12, 2005. http://www.ala.org/ala/alsc/ awardsscholarships/literaryawds/caldecottmedal/caldecottterms/ caldecottterms.htm

Children’s Book Council. Retrieved February 17, 2005. http://www.cbcbooks.org/html/ illustrating.html

Clark, R., Lennon, R., Morris, L. (1993). Of Caldecotts and Kings: Gendered images in recent American children’s book by black and non-black illustrators. Gender and Society, Vol. 7, No. 2 (June, 1993), pp. 227-245

Crabb, P. B., Bielawski, D. (1994). The social representation of material, culture, and gender in children’s books. Sex Roles: A Journal of Research, Jan, 1994

Creany, Ann D. (1995). The appearance of gender in award-winning children’s books. Eyes on the Future: Converging Images, Ideas and Instruction. Chicago, IL: International Visual Literacy Association.

Davis, Albert J. (1984). Sex Differentiated Behaviors in Nonsexist Picture Books. Sex Roles: A Journal of Research, vol.11, No. ½, (August 1984), pp. 1-15

Gooden, Angela M.,Gooden, Mark A.(2001). Gender representation in notable children’s books: (1995-1999). Sex Roles: A Journal of Research, July 2001

Kortenhaus, Carole M. (1993) Gender role stereotyping in children’s literature: An update. Sex Roles: A Journal of Research, Feb

Oskamp, S. Kaufman, K., Wolterbeek, L.A. (1996) Gender role portrayals in preschool picture books. Journal of Social Behavior & Personality; Special Issue Vol. 11 Issue 5, p.27,13, 2 chart

Turner-Bowker, Diane M. (1996). Gender stereotyped descriptions in children’s picture books does "Curious Jane" exist in the literature? Sex Roles: A Journal of Research, Oct, 1996

U.S. Census Bureau. (2000) Retrieved March 12, 2005. http://www.census.gov/ population /pop-profile/2000/chap20.pdf

Weitzman, L.J., Eifler, D., Hokada, E., & Ross C. (1972). Sex-role socialization in picture books for preschool children. American Journal of Sociology, vol. 77, pp. 1125-1150.

Williams, J.A., Jr., Vernon, J.A., Williams, M.C., & Malecha, K. (1987). Sex-role socialization in picture books: An update. Social Science Quarterly, vol. 68, pp. 148-156

Appendix A

List of books

2004 Medal Winner:

The Man Who Walked Between the Towers by Mordicai Gerstein (Roaring Brook Press/Millbrook Press)

Honor Books:

· Ella Sarah Gets Dressed by Margaret Chodos-Irvine (Harcourt, Inc.)

· What Do You Do with a Tail Like This? illustrated and written by Steve Jenkins and Robin Page. (Houghton Mifflin Company)

· Don't Let the Pigeon Drive the Bus by Mo Willems. (Hyperion)

2003 Medal Winner:
My Friend Rabbit by Eric Rohmann (Roaring Brook Press/Millbrook Press)

Honor Books:

· The Spider and the Fly illustrated by Tony DiTerlizzi, written by Mary Howitt (Simon & Schuster Books for Young Readers)

· Hondo & Fabian by Peter McCarty (Henry Holt & Co.)

· Noah's Ark by Jerry Pinkney (SeaStar Books, a division of North-South Books Inc.)

2002 Medal Winner:
The Three Pigs by David Wiesner (Clarion/Houghton Mifflin)
Honor Books:

· The Dinosaurs of Waterhouse Hawkins illustrated by Brian Selznick, written by Barbara Kerley (Scholastic)

· Martin's Big Words: the Life of Dr. Martin Luther King, Jr. illustrated by Bryan Collier, written by Doreen Rappaport (Jump at the Sun/Hyperion)

· The Stray Dog by Marc Simont (HarperCollins)

2001 Medal Winner:

So You Want to Be President? Illustrated by David Small, written by Judith St. George (Philomel)
Honor Books:

· Casey at the Bat illustrated by Christopher Bing, written by Ernest Thayer (Handprint)

· Click, Clack, Moo: Cows that Type illustrated by Betsy Lewin, written by Doreen Cronin (Simon & Schuster)

· Olivia by Ian Falconer (Atheneum)

2000 Medal Winner:
Joseph Had a Little Overcoat Simms Taback (Viking)
Honor Books:

· A Child's Calendar illustrated by Trina Schart Hyman
Text: John Updike (Holiday House)

· Sector 7 by David Wiesner (Clarion Books)

· When Sophie Gets Angry-Really, Really Angry by Molly Bang (Scholastic)

· The Ugly Duckling illustrated by Jerry Pinkney Text: Hans Christian Andersen, adapted by Jerry Pinkney (Morrow)

1999 Medal Winner: Snowflake Bentley, Illustrated by Mary Azarian, text by Jacqueline Briggs Martin (Houghton)

Honor Books:

· Duke Ellington: The Piano Prince and the Orchestra illustrated by Brian Pinkney
Text: Andrea Davis Pinkney(Hyperion)

· No, David! by David Shannon (Scholastic)

· Snow by Uri Shulevitz (Farrar)

· Tibet Through the Red Box by Peter Sis (Frances Foster)

1998 Medal Winner: Rapunzel by Paul O. Zelinsky (Dutton)
Honor Books:

· The Gardener illustrated by David Small
Text: Sarah Stewart (Farrar)

· Harlem illustrated by Christopher Myers
Text: Walter Dean Myers (Scholastic)

· There Was an Old Lady Who Swallowed a Fly by Simms Taback (Viking)

1997 Medal Winner: Golem by David Wisniewski (Clarion)
Honor Books:

· Hush! A Thai Lullaby illustrated by Holly Meade; text: Minfong Ho (Melanie Kroupa/Orchard Books)

· The Graphic Alphabet by David Pelletier (Orchard Books)

· The Paperboy by Dav Pilkey (Richard Jackson/Orchard Books)

· Starry Messenger by Peter Sís (Frances Foster Books/Farrar Straus Giroux)

1996 Medal Winner: Officer Buckle and Gloria by Peggy Rathmann (Putnam)
Honor Books:

· Alphabet City by Stephen T. Johnson (Viking)

· Zin! Zin! Zin! a Violin, illustrated by Marjorie Priceman; text: Lloyd Moss (Simon & Schuster)

· The Faithful Friend, illustrated by Brian Pinkney; text: Robert D. San Souci (Simon & Schuster)

· Tops & Bottoms, adapted and illustrated by Janet Stevens (Harcourt)

1995 Medal Winner: Smoky Night, illustrated by David Diaz; text: Eve Bunting (Harcourt)
Honor Books:

· John Henry, illustrated by Jerry Pinkney; text: Julius Lester (Dial)

· Swamp Angel, illustrated by Paul O. Zelinsky; text: Anne Issacs (Dutton)

· Time Flies by Eric Rohmann (Crown)

1994 Medal Winner: Grandfather's Journey by Allen Say; text: edited by Walter Lorraine (Houghton)
Honor Books:

· Peppe the Lamplighter, illustrated by Ted Lewin; text: Elisa Bartone (Lothrop)

· In the Small, Small Pond by Denise Fleming (Holt)

· Raven: A Trickster Tale from the Pacific Northwest by Gerald McDermott (Harcourt)

· Owen by Kevin Henkes (Greenwillow)

· Yo! Yes? illustrated by Chris Raschka; text: edited by Richard Jackson (Orchard)

1993 Medal Winner: Mirette on the High Wire by Emily Arnold McCully (Putnam)
Honor Books:

· The Stinky Cheese Man and Other Fairly Stupid Tales, illustrated by Lane Smith; text: Jon Scieszka (Viking)

· Seven Blind Mice by Ed Young (Philomel Books)

· Working Cotton, illustrated by Carole Byard; text: Sherley Anne Williams (Harcourt)

1992 Medal Winner: Tuesday by David Wiesner
(Clarion Books)
Honor Book:

· Tar Beach by Faith Ringgold (Crown Publishers, Inc., a Random House Co.)

Appendix B

Terms and Criteria

Randolph Caldecott Medal


1. The Medal shall be awarded annually to the artist of the most distinguished American picture book for children published in English in the United States during the preceding year. There are no limitations as to the character of the picture book except that the illustrations be original work. Honor Books may be named. These shall be books that are also truly distinguished.

2. The Award is restricted to artists who are citizens or residents of the United States. 

3. The committee in its deliberations is to consider only the books eligible for the award, as specified in the terms.


1. A "picture book for children" as distinguished from other books with illustrations, is one that essentially provides the child with a visual experience. A picture book has a collective unity of story-line, theme, or concept, developed through the series of pictures of which the book is comprised.

2. A "picture book for children" is one for which children are a potential audience. The book displays respect for children's understandings, abilities, and appreciations. Children are defined as persons of ages up to and including fourteen and picture books for this entire age range are to be considered. 

3. "Distinguished" is defined as 

· marked by eminence and distinction: noted for significant achievement

· marked by excellence in quality

· marked by conspicuous excellence or eminence

· individually distinct 

4. The artist is the illustrator or co-illustrators. The artist may be awarded the medal posthumously.

5. "Original Work" means that illustrations reprinted or compiled from other sources are not eligible. 

6. "American picture book in the United States" specifies that books originally published in other countries are not eligible. 

7. "Published . . . in the preceding year" means that the book has a publication date in that year, was available for purchase in that year, and has a copyright date no later than that year. A book might have a copyright date prior to the year under consideration but, for various reasons, was not published until the year under consideration. If a book is published prior to its year of copyright as stated in the book, it shall be considered in its year of copyright as stated in the book. The intent of the definition is that every
book be eligible for consideration, but that no book be considered in more than one year.

8. "Resident" specifies that author has established and maintained residence in the United States as distinct from being a casual or occasional visitor.

9. The term, "only the books eligible for the Award," specifies that the committee is not to consider the entire body of the work by an artist or whether the artist has previously won the award. The committee's decision is to be made following deliberation about books of the specified calendar year.

10.  "In English" means that the committee considers only books published in English. This requirement DOES NOT limit the use of words or phrases in another language where appropriate in context.


1. In identifying a distinguished picture in a book for children, 

a. Committee members need to consider: 

· Excellence of execution in the artistic technique employed;

· Excellence of pictorial interpretation of story, theme, or concept; of appropriateness of style of illustration to the story, theme or concept; of delineation of plot, theme, characters, setting mood or information through the pictures. 

b. Committee members must consider excellence of presentation in recognition of a child audience. 

2. The only limitation to graphic form is that the form must be one which may be used in a picture book. The book must be a self-contained entity, not dependent on other media (i.e., sound or film equipment) for its enjoyment.

3. Each book is to be considered as a picture book. The committee is to make its decision primarily on the illustration, but other components of a book are to be considered especially when they make a book less effective as a children's picture book. Such other components might include the written text, the overall design of the book, etc.  

Note: The committee should keep in mind that the award is for distinguished illustrations in a picture book and for excellence of pictorial presentation for children. The award is not for didactic intent or for popularity.


Appendix C

Book Coding Sheet



Number of Pages

Number of Illustrations


Author’s Name

Author’s Race

Author’s Gender


Illustrator’s Name

Illustrator’s Race

Illustrator’s Gender

Main Character

Main Character’s Race

Main Character’s Gender

Artifact Use

Activity Level

Secondary Character

Secondary Character’s Race

Secondary Character’s Gender

Artifact Use

Activity Level

Book Theme


Appendix D

Illustration Coding Sheet

Book Title

Number of Female Characters

Number of Male Characters

Number of Ambiguous Characters

Number of Female Nonhuman Characters

Number of Male Nonhuman Characters

Number of Ambiguous Nonhuman Characters


Number of Nonwhite Female Characters

Number of Nonwhite Female Characters

Number of Ambiguous Nonwhite Characters

Central Character

Gender of Character

Location of Character

Artifact Use

Activity Level

Secondary Character

Gender of Character

Location of Character

Artifact Use

Activity Level




Appendix E


Illustrator’s Gender

0 – Female

1 – Male

2 – DK (ambiguous)

Character’s gender

0 – Female

1 – Male

2 – DK (ambiguous)


0 – Nonwhite

1 – White

2 – DK

Book Theme

0 – Socially Conscious: Shows awareness of social issues in the world such as pollution, poverty, crime, etc.

1 – Racially/Culturally Conscious: Portrays different cultural customs, holidays, etc.; celebrates differences.

2 – Melting Pot: Portrays characters of different races as a homogenous group; focuses on similarities.

3 – N/A


0 – Outside

1 – Inside

2 – N/A


0 – Domestic

1 – Production

2 – N/A

Activity Level

0 – Active: Gross motor (large muscle) activity, physical activity, hard labor, play.

1 – Passively Active: Fine motor (small muscle) activity; work that does not require hard physical labor (e.g. sewing, drawing, etc.).

2 – Passive: Very little or no activity; (e.g. reading, talking, daydreaming, sleeping, lounging).

Appendix F

Subjective Thematic Notations of Cultural/Racial Diversity, Gender Breaking or Underlying Reinforcement

Although the focus of our study was on gender representation, we were also interested to see if there was an increase in books (as well as individual illustrations) that featured other themes, such as cultural and/or racial diversity, the emergence of gender breaking or reinforcement, and also particular notations that deserved further explanation.

In regards to approaching racial diversity (in which some also served a historical educational purpose), the books "Working Cotton", "Harlem", and "Martin’s Big Words" are all works that feature non-white characters (central or non-central), that in addition, gave insight to the African-American experience and offering an alternative perspective concerning individuals unlike themselves. This realization is delivered to the young reader by the uncomplicated images of whole families experiencing hard labor in the cotton fields of the past, the identification of eminent historical figures-such as Langston Hughes and Dr. Martin Luther King, Jr. - who are central (with many others) to the cultural and social lives of many African-Americans, it feeds to a young mind "Black life from a Black perspective."

The book Yo! Yes?, is a narrative of interaction between white and African-American male characters; a simple tale espousing positively crossing an unstated racial separation and establish a friendship that is beneficial to both characters- the boys are smiling, have formed a bond, and demonstrate unity by appearing arm-in-arm in an illustration.

Along with racial diversity, there were two books that were written with the intent of affording the young reader an introduction to a higher level of cultural and ethnic consciousness.

The book Tibet through the Red Box, featured not only non-white characters throughout the text, but featured the Tibetan cultural aspect of generational storytelling, and highlighted its illustrations with the Tibetan terrain of mountains and fantastic mazes.

The aspects of cultural diversity are evident within the book Joseph Had an Overcoat. It should also be noted that this text also teaches the reader about the moral lesson of being prudent, non-materialistic, and how to put together something out of nothing. The character, Joseph, is identified throughout the book as being a Jewish peasant- the symbols abound in the illustrations of the menorah and the Jewish proverbs that hang in his simple home. Another aspect of Joseph’s Yiddish culture is the significant display of music throughout the book; dancing, singing, and the playing of instruments are all key elements in his life, as is the community and family. The character of Joseph also breaks away from the gender stereotype of sewing (generally reserved to be primarily female task) by making many different useful items out of his overcoat, such as a jacket, a tie, a handkerchief, and finally a button, only to then write a book about his experience so that others may follow his example.

As a supplement to the hard data accumulated from this study concerning female and male visibility, location, and artifact use, it is conducive to scrutinize the underlying patterns found within the books in relation to gender stereotype reinforcement and the efforts to expand the notion of gender cemented to interrupt the accepted social philosophy of being "male" or "female." This expansion is seen in the book Seven Blind Mice, where the power of thought and decision shifts from male to female, as opposed to the male functioning as principal problem solver. The first six characters are identified as being male, and it is not until the seventh character of the lone female that the many part of the puzzle are put together and made whole (the puzzle turns out to be an elephant) - it is she that is the most active and exploratory, has the clearest thought processes, and is the bearer of the final truth.

The illustrations found in the book Owen, portray a twist on the notion of gender, both emotionally and materially. The parents are both involved with household chores, such as laundry which is typically rendered exclusively a female activity, and also work in partnership to solve the problem of kindergarten-bound Owen’s deep attachment to his beloved (and well-worn) blanket, "Fuzzy." The father (in one illustration clad in pink pajamas adorned with hearts) is shown comforting Owen and allows his son to show tearful emotion. This is a definitive step away from the adage of "boys don’t cry."

It must also be noted that there were a number of the books coded that reinforced gender stereotypes and expectations, both subtle and otherwise. While the story of Officer Buckle and Gloria (written by a female author), hails the overall importance of friendship, it is apparent that the nature of Officer Buckle’s character is to be the chief possessor of power between the two personalities. The canine subordinate Gloria, is female, and acts upon every command given to her by her male master, Officer Buckle. When it is discovered that she is the more revered of the two, causing the power dynamic to be threatened, he stops the safety lectures that he and Gloria had been giving to school children.

The book Casey at Bat, which is a historical perspective about baseball, displayed crowd illustrations that were heavily numbered with men, and as few as 3 or 4 women. All the active characters were men; and in fact, many were portrayed using baseball bats, etc., which were non-codeable as production but highly gendered as an object. The text employed such language as "his teeth were clenched in hate…he pounds with cruel violence his bat upon the plate." This is highly aggressive behavior and description of the "All-American" activity of baseball is also commonly associated with sports and masculinity.

In the book, The Man Who Walked Between the Towers, the author attempts to reveal the timeless lessons about courage, liberty, and the importance of taking the necessary gambles in order to achieve and be the best person one can aspire to be. One gamble that the author did not take in his book however, was the opportunity to pierce the concept of traditional gender roles and stereotypes. Although a policewoman makes the initial call about what is happening at the twin towers; she is the only female on the force and primarily used for "back-up" purposes, and quickly recedes into the background. (This is also mentioned in the conclusion, I think…..) In a very disturbing illustration where the man who walked between the towers stands before the judge- there is a lone woman in the background with her cleavage distinctly depicted and smiling broadly, while the judge is exemplified as being almost jolly (as though he is Santa Claus wearing a judicial robe), looking on in a paternal capacity- wielding his gavel and dispensing even-handed justice.

The final book for gender analysis, So You Want to Be President, was published by Scholastic Inc., and written by a female author. While this book feeds the young reader an abundance of information about former U.S. Presidents, it also reflects and reinforces the male dominated nature of U.S. politics. Making it extremely clear that the opportunity to gain the highest position of political power is available exclusively to men, included in the criteria for eligibility to become president are such phrases as "it would help if your name was James." This is a perplexing book as it reflects many conservative perspectives regarding gender in politics while at the same time containing illustrations with non-traditional representations of gendered appearance and behavior. In one case, more than a few of the former Presidents are shown cleaning clothes on a washboard. In another illustration, a number of former presidents appear in skirted cheerleading attire – which, while not necessarily central to the content of the book, is worth citing as an example of explicit breach of gender norms.