Examining the culture of fatherhood in American children's literature: presence, interactions, and nurturing behaviors of fathers in Caldecott award winning picture books (1938-2002).

This research examines the cultural scenarios of American fatherhood by measuring father presence and interactions with children in Caldecott award winning picture books from 1938 to 2002. Comparisons of fathers and mothers are measured and examined for changes over time. Fathers are found to be present in relatively the same number of books as mothers yet are depicted as engaging in significantly fewer interactions with children than mothers. The 1960s were found to be a period of time in which changes occurred in the presence of fathers in the books, supporting prior research which posits a "shifting" pattern in fatherhood imagery. This study also finds concurrence with a "fluctuating" pattern of changes in the culture of fatherhood in media with regard to father interactions. Further, findings point to the possibility of a "new father" image occurring during the 1980s.

Keywords: fatherhood, culture of fatherhood, American culture, children's literature, children's culture 

**********

The purpose of this research is to examine the social constructions of fatherhood in contemporary American culture via a content analysis of children's literature. Specifically, the aim of this project is to contribute to the continuing dialogue in the "changing culture of fatherhood" line of inquiry. The culture of fatherhood refers to the "norms, values, beliefs, and expressive symbols pertaining to fatherhood" (LaRossa et al., 2000, p. 375). Family scholars interested in these constructs have sought to discover and document how fathers are portrayed in a variety of popular American media such as single-panel cartoons (Day & Mackey, 1986; LaRossa, Gordon, Wilson, Bairan, & Jaret, 1991, LaRossa, Jaret, Gadgil, & Wynn, 2000), magazine articles (Atkinson & Blackwelder, 1993), and television commercials (Coltrane & Allen 1994) via the methods of content analysis. The images of fathers in culture are of particular interest because they may "affect the social reality of fatherhood" (LaRossa et al., 2000) and possibly influence the expectations that participants in culture have for the role of a father.

LaRossa et al. (2000) explain that, taken together, the research involved in this line of inquiry supported that there has been a change in the culture of fatherhood but did not agree on how much of a change had occurred, when the change occurred, or how the change could be characterized in terms of linear shifts or oscillating fluctuations. A component of the complexity of this dialogue is due to the use of diverse media, differing points in time for analyses, contrasting approaches to the problem, and the use of a range of variables related to the (changing) roles of fathers. Table I provides a brief sketch of the major studies involved in examining the changing culture of fatherhood.

Of particular interest to social scientists are the intersections of cultural and possibly symbolic phenomena with historical events and patterns. For example, Atkinson and Blackwelder (1993) compare the cultural definition of fathering evidenced in magazine articles from 1900 to 1989 with fertility rates and married women's labor-force participation, finding an association between higher fertility rates and cultural definitions of fathers as providers.

Current culture-of-fatherhood researchers have provided clear rationales for why comics, magazine articles, and television commercials are meaningful sources of information about fathers in mainstream American culture. These media are easily accessible to a broad range of readers/consumers and are tools of communication driven by a sense of shared understandings and motivations to connect the reader to the message. The artifacts that have been examined are possible sources of role identity information for male parents and may provide a model of fathering behaviors that people may assume is the norm. These studies have taken an anthropological and historical approach and thus have called for further examination of a broader range of artifacts to build greater understanding of fathers in culture. Common features of the artifacts researched to date are that they are suited to people over the approximate age of eight to 10 in terms of cultural and cognitive accessibility. Comic strips, perhaps, have the broadest appeal but generally require the reader to have acquired somewhat sophisticated reading strategies in order to understand the meaning of the content in the frames.

In broadening the range of artifacts to examine fatherhood, then, a meaningful addition to our knowledge about the changing culture of fatherhood would be from the cultural perspective of younger people (those younger than eight). The intersections of children within culture are of interest to social scientists, educators, and families (Corsaro, 1997; Lesnik-Oberstein, 1994, 1998; Malaguzzi, 2000; Rinaldi, 2001; Super & Harkness, 1997; Whiting & Child, 1953).

One approach to understanding a child's experiences as part of a culture is to examine the artifacts with which children interact on a regular basis (Corsaro, 1997). Children's literature is an especially meaningful medium for exploration because of its widespread availability in the United States, its use as an instructional tool, and its enduring "shelf-life." Books available in public libraries, schools, and home collections are often decades old, which gives children (and researchers) the opportunity to interact with artifacts from several eras. Further, fathers may read books with their children and also use the books to "see themselves" in the role of parent. Picture books are specific textual artifacts made for the use of children under the age of eight. Picture books are a unique genre of children's literature that use pictures (as well as words) to tell a story (Anderson, 2002). Researchers in the social sciences and in the field of education have used picture books for children to examine such phenomena as sex roles and the portrayals of parenting (Child, Potter, & LeVine, 1946; Heller, 1985; Hillman, 1974, Jacklin & Mischel, 1973; Key, 1971; Weitzman Eifler, Hokada, & Ross, 1972). Heller's (1985) doctoral dissertation focused on the portrayals of fathers in picture books and magazines comparing the time periods of 1946 to 1955 versus 1973 to 1982. He found changes in how fathers are portrayed from the earlier time period (father as economic provider) to the later time period (father as active childcare provider). The current study differs from Heller's research in methodology (by using all Caldecott medal and honor books from 1938 to 2002) and seeks primarily to place findings within the current "changing culture of fatherhood" line of inquiry.

This research begins with several broad research questions intended to serve as springboards for a more specific set of questions developed within the process of the research. First, how are fathers portrayed in American picture books for children, and how has this changed over time? How do the portrayals of fathers compare to the portrayals of mothers? And finally, how does the analysis of American picture books provide insight into the social constructions of fatherhood and the changing culture of (American) fatherhood as it is presented to young people?

THEORETICAL FRAMEWORKS

Drawing on theories of cultural psychology (Ratner, 2002, who cites Vygotsky, 1997, 1998), meaning is both expressed and made by engagement with the artifacts of culture. Studies of cultural artifacts provide information about what a culture holds as valued as well as insight into how the participants of culture make meaning. According to cultural theories, there are many sources of information with which participants in culture might engage to construct meanings associated with the role of father. The present research places culture at the center of the inquiry, using a cultural/ecological model as a framework that emphasizes the relationship of the growing human to social and environmental factors in their immediate and extended settings (Super & Harkness, 1997). The components of a young child's ecology relevant to this study are their caretakers (in particular, their male parent) and the cultural aspect of picture books. Within a cultural/ecological framework, this research employs concepts from a symbolic interactionist perspective as it relates to family studies to explain why the line of inquiry is important and how the aspects of cultural artifacts might serve as tools or vehicles for the constructions of meanings related to fatherhood. According to LaRossa and Reitzes (1993), symbolic interactionist approaches to family studies are concerned with the importance of meanings, self-concept, and the influence of society. Meanings are thought to be the basis of how people interact and arise out of human interactions via an interpretive process. Thus, young children construct meanings by using the books as a source of information and, in turn, may react to others based on these meanings.

Reading books is an interpretive process, and therefore it is possible that there are multiple interpretations for any given picture book. It is beyond the scope of this research to discover the meanings that young people process when reading picture books; however, a content analysis of the images provided (in both the pictures and the text) can provide a basis for exploring how fathers are or are not portrayed in the artifacts.

With regard to self-concept, symbolic interactionism posits the self as developing via social interaction, and, in addition, developed self-concepts are thought to be motives for behaviors (LaRossa & Reitzes, 1993). The depictions of fathers in picture books may contribute to the formation of fatherhood expectations, identities, and motivations to act as a father is expected to act by readers of the picture books. Further, symbolic interactionism posits that, in the social realm of experience, people "work out the details of social structure" (LaRossa & Reitzes, 1993, p. 144). The significance of the present research lies in its ability to explore a component the broader culture concerning mainstream American fatherhood via the process of reading picture books. In this regard, symbolic interactionism shares common ideas with cultural/ecological approaches with regard to the role of social interaction in the development of self in culture.

Changing culture-of-fatherhood studies have examined aspects of fatherhood imagery such as the presence of fathers in media, how often fathers are mocked or portrayed as incompetent, and how fathers are characterized as nurturing or providing for children. This study of children's literature will examine these constructs as well as aspects of the theoretical construct of father involvement (Lamb, Pleck, Charnov, & Levine 1985). "Father involvement" includes the components of engagement (direct interaction with the child), accessibility (availability to the child), and responsibility for childcare (apart from direct care). The current examination of the portrayals of fathers in children's literature will focus on engagement (as measured by the number and type of interactions depicted in the books) and accessibility (as measured by the presence of fathers in the books). By using these concepts, it is possible to add dimension to the current changing culture of fatherhood studies.

THE APPROACH

CONTENT ANALYSIS

Common to the various techniques of analyzing content are that they are objective, systematic, employing rules of categorization, and replicable (see Bauer, 2000; Holsti, 1969; Weber, 1985). The objectivity referred to in the definitions of content analysis does not mean that the method produces a "single valid reading of the texts"; rather, validity of content analysis can be judged "in terms of its grounding in the materials and its congruence with the theory of the researcher, and in light of his or her research purpose" (Bauer, 2000, p. 133). With that in mind, this research aims to begin a dialogue about the content of American picture books for children with regard to the portrayals of families with specific attention to how fathers are depicted. This research will be as objective as possible (by using narrow operational definitions of variables) and systematic (by examining each book with identical criteria). Reliability will be examined by testing how well the content analysis can be replicated by other researchers. Still, this research aims to begin a dialogue about the content of the picture books rather than posit an absolute interpretation of the texts.

SELECTION OF APPROPRIATE CORPUS

The first step in the research process was to select an appropriate corpus of picture books with which to conduct the content analysis. Among the considerations in this task are that the books be published in the same time frames as the media that were examined in the foundational changing culture of fatherhood studies (1930s to 1990s). Additionally, it must be possible to systematically obtain books from each year between the 1930s and the 1990s in order to analyze at equal intervals from multiple points in time, such as was done by LaRossa et al. (2000). Further, the books must be suitable for young children (birth to age eight), commonly read by young people, and easily accessible to the researcher. Also, the books must be "American" in order to provide a comparison with the other American media examining the culture of fatherhood.

Taking these criteria into consideration, the media examined in this study are all of the Caldecott award winning (medal and honor) picture books from 1938 to 2000 (N = 271). These books meet the criteria desired for an appropriate corpus to examine for this study. Each year since 1938, the American Library Association (ALA) has awarded an illustrator of a newly published picture book the distinction of the Caldecott medal for quality illustration. Each year there are also several Caldecott honor awards given. Another criterion for the award is that the book be published in the United States and that the illustrator is a citizen or permanent resident of the U.S. This is the most prestigious award for picture books in the U.S. (Anderson, 2002). Because of the distinction, these books are widely available in children's sections of public libraries and in schools, which establishes that these books are accessible to many children. In fact, all of the books were obtained by the researcher from a public library that had a collection of all of the Caldecott books in a prominent location in the children's section of the library.

A limitation to using the Caldecott collection is that, while the books are easily accessible, this does not mean that the books are popular or well liked by young people. Still, after much consideration in the preliminary stages of the project, the Caldecott collection was deemed to be a set of books that best met all of the criteria. According to Anderson (2002), although receiving the Caldecott award does not mean that the book will be a "best seller," it does mean that every school and public library will have multiple copies, making the book readily accessible to children who patronize public libraries. In addition, several researchers of children's literature have also used the Caldecott collection as the basis of their work (Collins, Ingoldsby, & Dellmann, 1984; Heller, 1985; Kortenhaus & Demarest, 1993; Weitzman, Eifler, Hokada, & Ross, 1972), adding confidence in the Caldecott collection as an appropriate medium for social science investigation.

RESEARCHER'S ORIENTATION TO THE ARTIFACTS

Before a coding frame could be built, a necessary procedure was to spend time reading the books to get an idea of how the books worked by examining general elements of the texts. To begin, a list of all Caldecott award winning books (medal and honor) from 1938 to 2002 was obtained from the ALA web site. All books were available from a central branch of a county library system in the northeastern United States. The researcher was given permission to check out 100 of the books at a time to take home for several weeks.

Most of the books in the collection were of the same general length, with few exceptions. These books were examined further and were determined not to present an analytical problem because the interactions measured within the books were not higher than those books with fewer pages. The stories in the Caldecott collection were found to be distinctive and diverse, taking up a broad range of topics. The characters in the books include realistic humans, anthropomorphic animals, and realistic animals.

It was immediately apparent that about half of the books in the collection did not portray families at all, which would need to be a variable in the analysis. Further, it was discovered that there were some books that depicted a mother and a father, but the nature of the story was that these parents were not necessarily in the same family (as in the case of some collections of nursery rhymes). This was made a variable for coding.

Several books portray either a father or a mother who is the only parent in the book. In a few cases, the absence of a parent was explained (in the case of the death of one or the other of the parents), but in most cases the absence was not explained. Books with a father only or a mother only would also be an important variable in the analysis because it is the number of books that portray single parents that account for whether or not there are more books that portray mothers than fathers. Also, the portrayals of just one parent as the only visible companion or caretaker of a child in the story might give insight into how fathers and mothers are portrayed in child-caregiving activities.

The images of families in the Caldecott books were predominantly positive and straightforward. There were only a few books in which the depiction of either a mother or a father could be considered "negative" (harmful, not acting in the best interests of the child) compared to "positive" (nurturing and supportive). The coding frame would have to be constructed in such a way that positive and negative interactions are either considered or noted. This was accomplished by measuring general interactions, which is a value-free construct, as well as specific positive/nurturing parenting behaviors. Further, there was not a single book that mocked a parent or portrayed him or her as incompetent. This feature of the texts would mean that the measures for this study would need to be different from some of the foundational research that examined incompetence of fathers and mothers.

Finally, a feature of the corpus worth noting is that several illustrators had been given the award more than once. This was considered a possible bias in the corpus because these illustrators might have a particular style or way of perceiving either fathers or mothers. This could influence, for example, how many fathers are depicted or the types of activities that fathers engage in. After much thought, it was decided that the current research questions are concerned primarily with the contents of the artifacts rather than why these artifacts were created and by whom. Therefore, the influence of a few illustrators who may have books appearing more than once was not a factor in the current analysis. However, it would be a point worth examining in future work.

CONSTRUCTION OF A CODING FRAME

After a first read of most of the books, and drawing upon an exploratory pilot study of fathers in a general collection of children's literature (Quinn, 2000), a coding sheet and book with operational definitions and instructions for coding was developed. Each coding sheet provided space for recording general book information such as the title of the book, year of the award, names and sexes of authors and illustrators, publishing company, genre of the book (realistic, fantasy or special case such as religious story), character description (human, realistic animal, anthropomorphic animal), skin tone of the characters, and ethnicity of the characters. This information would be available should any of these variables surface as important to this or future research.

The coding sheet also provided space to record the key variables for the study: father present in the book (yes or no), mother present in the book (yes or no), father and mother in the same family (yes or no), number of depictions of interactions between father and child in the book, number of depictions of interactions between mother and child in the book, and the number of the following positive/nurturing parenting behaviors for both fathers and mothers: physical affection, verbal affection, teaching the child a skill, and playing with the child.

Presence of a father or mother in the book was defined as pictured or referenced in the text. This measure made possible a comparative analysis of the number of books that portray fathers, mothers, and both parents. After the data were recorded, this variable would provide an overall count of how many books depict a father, a mother, or both parents. This variable would also be grouped into mean number of books that depict fathers and mean number of books that depict mothers per half-decade time period in order to offer a comparison to the attention given to Father's Day and Mother's Day in the LaRossa et al.'s (2000) six-decade analysis, which also employed a half-decade analysis. The very beginning of the award period and the very end of the award period are partial half-decade periods due to the award being first granted in 1938 and the end of the data set being 2002. Therefore, the half-decade periods were defined as 1938-1939; 1940-1944; 1945-1949; ... 1995-1999; 2000-2002. Father presence in picture books is a comparable measure of the accessibility component of Lamb et al.'s (1985) father involvement model.

A set of positive/nurturing parenting behavior variables was constructed for the purpose of addressing the theoretical concepts of paternal nurturance (to compare with Atkinson & Blackwelder 1993 as well as Coltrane & Allan, 1994; and LaRossa et al., 2000) and paternal engagement (Lamb et al., 1985). These variables are designed to measure a father and mother's observable parenting behaviors with children by measuring general parental interactions with children in the books as well as specific types of interactions. Parental interaction with a child was coded as a continuous variable that represents the number of illustrations and textual references in which the parent (mother and father separately) "interacts" with the child. For example, an illustration of a parent giving food to a child; reading to a child; playing with, talking to, or working alongside a child (while talking or engaging in work together) is a direct interaction. Cases were also coded if an illustration did not accompany textual reference--for example, if the text of the book indicates that there was a conversation between a child and a parent but the book does not provide an illustration of the conversation--it would still be coded as an interaction. The following parenting behaviors were also coded individually after being counted as interactions: showing physical affection for the child (i.e., holding or hugging), verbal affection for the child (saying, "I love you" or praising a child for a job well done), teaching the child a skill, and playing with the child. These behaviors were found to encompass nearly all of the behaviors coded by Coltrane and Allan (1994) as well as LaRossa et al. (2000) in their measure of parental nurturance. The only behavior not measured in the picture books that was measured in the other media was "serving or caring for the child." This variable was coded in the pre-test of the picture books and found to be a potentially meaningful variable; however, the measure of inter-rater reliability was not found to be acceptable, and after much thought and consideration, it was determined that "care" was too general a term to hold up to the quantitative analysis and would need to be examined qualitatively in future research.

There was a maximum of one interaction coded per parent per page. (A preprinted scale was made available on the code sheet numbered from 1 to 10.) The code book also instructed that quantities over 10 should be noted in the margin of the code sheet by the actual number of interactions. If an illustration was spread across two pages (such as a centerfold), the illustration was to be treated as one page. Because most picture books are the same length, this approach would control for illustration styles that include many depictions on one page or those that clearly stretch across two pages. After interactions were coded, specific parenting behaviors were coded (physical affection, verbal affection, teaching the child a skill, and playing with the child). A maximum of one for each of the codes was allowed per page, following the same rules as the general interaction measure. It would be possible to count more than one of the codes per page; for example, if a father was depicted as playing with the child and saying "well done," this would be coded as one interaction, one instance of play with the child, and one instance of verbal praise/affection.

All parenting behavior variables were coded as continuous variables in the beginning. After running the first analysis, it was found that each of these variables had a very low mean. It became apparent that it was not a matter of determining how many times in a book a specific positive/nurturing parenting behavior was depicted but whether or not a book depicted these specific behaviors at all. Therefore, measuring the parenting behaviors via a dichotomous variable would be more appropriate. This would provide information about the number of books that depict either fathers or mothers engaging in specific positive/nurturing parenting behaviors. This was accomplished by using the compute function (continuous to dichotomous) on SPSS. Further, the positive/nurturing parenting behaviors could be combined into an aggregate positive/nurturing parenting behavior variable that would more closely resemble the measure of nurturance used in the prior research. This variable would provide information about the range of parenting behaviors depicted in books and could be used to compare number of books in which the range of father behaviors are more than, less than, or equal to mother behaviors. The aggregate parenting behavior variable was created on SPSS by summing the values of the specific parenting behavior variables, which produced a new variable with values ranging from 0 (book does not depict positive/nurturing parenting behavior) to 4 (book depicts four of the four positive/nurturing parenting behaviors coded). The comparisons of these variables for fathers and mothers could be analyzed overall as well as examined for changes over time.

Books that portray both a mother and a father were also coded to specify those books that portray parents in the same family (co-parents) or parents not in the same family. This variable would provide the ability to compare father and mother interactions based on whether or not the fathers and mothers were engaging in parenting as members of the same family. Using this distinction, it could be determined whether fathers or mothers are depicted as interacting more or less when they are in the same family, not in the same family, or the only parent in the book.

All coding was completed by the primary researcher, who read and coded each Caldecott award and honor book from 1938 to 2002 (N = 271) and entered the data into SPSS for analysis. Each book was entered as a file, and all variables were assigned a column in the data file. Several weeks later, a random sample of 140 books was coded by independent researchers for the purpose of checking the reliability of the first coding and to call attention to variables that might need to be examined further. The reliability coders were given a brief training session by the primary researcher and equipped with a coding book and coding sheets. The training session involved a discussion of the general purpose of the research (to examine the depictions of families in children's picture books) and details of how to code the books for the study. The reliability coders completed coding of several award and non-award winning books during their training session and then went to the public library or to an office in the university that held a temporary collection of the award winning books to read the books and complete the reliability checks. The reliability coding sheets were entered into SPSS files and checked against the primary researcher's coding to determine the degree of reliability. Percent agreement for the variables of father present in a book and mother present in a book was 100%. The continuous variables of father interaction and mother interaction with child were tested using an alpha model and found to be .90 and .91, respectively. The dichotomous variables of parenting behaviors were tested using Kappa, and all were found to be significant at the .01 level. No changes were made to the primary researcher's coding values; these were used in the final analysis.

RESULTS

FATHER PRESENCE COMPARED TO MOTHER PRESENCE IN THE BOOKS

The number of books in the population of Caldecott award winning books from 1938 to 2002 is N = 271. Out of the 271 books, there are 155 (57% of the collection) that depict a parent (either a father, mother, or both parents) and 116 books that do not depict a parent at all (43% of the collection). There are 128 books that depict a father (47.23% of the collection, or 82.6% of the books that depict a parent or parents) and 138 books that depict a mother (50.92% of the collection, or 89.03% of the books that depict a parent or parents). In the 65 years examined, there are 10 fewer books that depict a father than those that depict a mother. This is because there are 10 fewer books that portray a father as the only parent in the book (n = 17) than depict a mother as the only parent in the book (n = 27).

FATHER PRESENCE: CHANGE OVER TIME?

The numbers of books that depict fathers and mothers by half-decade is presented in Table 2. Before examining how the presence of fathers has changed over time, it is important to consider how the presence of families in the award books might have changed over time. Recall that nearly half of the books in the collection do not portray either a father or a mother. We must be careful to consider that changes in how many fathers are present in books might also be related to how many books have families in them. For the purpose of this analysis the term family refers to nuclear families that include either a mother or a father. Table 2 also provides information about how many awards were given per half-decade and how many of these books depict either a father or a mother (or both). The books that depict either a father or a mother or both parents are called "family books" because they depict a family (compared to books in which there is no family in the book). If there is a rise or decline in the number of books that portray fathers, it may be due to similar patterns related to the number of books that were granted an award or the number of books that portray families. In order to control for this, the number of books that portray families will be calculated as a percentage of all books and tested to determine whether any time periods are significantly different from the others. The number of books that portray a father and a mother will be presented in both absolute value and as a percentage of all "family books."

The Levene test reveals that the variances of family books per half-decade are not equal (13.279, p = .000). Based on an ANOVA with Games-Howell post hoc tests, the only half-decade time period that stands in contrast to any of the others is the late 1930s (with regard to the number of books that portray families). Specifically, the number of books with families in them during the late 1930s were found to have significant differences with the early 1950s (mean difference = .50, p = .001), the late 1950s (mean difference = .50, p = .017), the early 1970s (mean difference = .6818, p = .000), the late 1970s (mean difference = .5294, p = .027), the early 1980s (mean difference = .55, p = .006), and the late 1990s (mean difference = .3913, p = .049). In the early years of the award (1930s), there were high numbers of books that portrayed families; these "family books" decreased until the early 1970s, when they were at the low end of the data set.

With regard to how many books portray fathers, the early 1960s was a low period, when fathers were portrayed in just 60% of the family books, in contrast to the late 1950s and late 1960s when fathers were portrayed in 90% of family books. However, an ANOVA with post hoc tests (Games-Howell) does not reveal significant differences between any time periods with regard to the percentages of fathers portrayed in the books, F(13, 141) = .75, p = .71. With regard to how many books portray mothers as a percentage of family books, the late 1960s was the low point with mothers present in 60% of all books that depict a family. ANOVA with Games-Howell post hoc tests did not reveal any significant differences between time periods for mothers, F(13, 141) = 1.17,p = .305.

In terms of comparing fathers to mothers, the data reveal that the percentage of books that portray fathers are found to be the lowest in the data set during the early 1960s, while in contrast mothers are in 100% of the books that portray families during that time period. In the half-decade following that, the percentage of books that portray mothers are found to be lowest in the data set while the percentage of books that portray fathers is at its highest.

When absolute values are examined, we see that in the early 1960s fathers were in four fewer books than mothers and that in the late 1960s fathers were in three more books than mothers. This difference in number of books that portray fathers and mothers is caused by the number of books that portray a father as the only parent compared to the number of books that portray a mother as the only parent. In the early 1960s, there were no books that portray a father as the only parent and four books that depict mother only. In the late 1960s, there were four books that depict a father as the only parent and just one that depicts a mother as the only parent. For fathers, the late 1960s was the period of time in which there was the greatest number of books that depict a father as the only parent. For mothers, the early 1960s was one of the half-decades with the highest number of these books (during the early 1990s, there were also four books that depict a mother as the only parent). In addition to distinguishing this time period as worthy of note with regard to the analysis, this finding points out that a half-decade analysis is appropriate to uncover such possible fluctuating patterns. If the decade were analyzed as one period of time, these differences might have appeared to cancel each other out.

Of importance in this analysis is that rarely are the absences of fathers or mothers explained within the text of the picture books. Therefore, a book that depicts a father as the only parent in the book is not necessarily depicting a "single father" (father who parents his child without a mother). In most books, the reader can only make assumptions about the whereabouts of the "other parent." Still, what is interesting is that a book that depicts either one parent or another shows that parent within the realm of the child's world. So what can be said about these books is not that "single fathers" appeared to be depicted in increasing amounts after the 1960s but that fathers were depicted alone with their children more often after the 1960s.

Clearly, the 1960s were a time period in which we see a change in the number of books that depict fathers as the only parent in the books and in the number of books that depict mothers as the only parent in the books. This finding leads to the question of whether or not the time period was possibly a "turning point" in the culture of fatherhood with regard to the presence of fathers in children's literature (a change in direction, indicating a shift in the presence of fathers in award-winning books) or whether or not the 1960s were a decade in which father (and mother) depictions "fluctuated" (changed for a brief period of time without long-term change). In order to examine this question, the number of books depicting either a father as the only parent or a mother as the only parent during the time period before 1965 (1938-1964; 27 years; 76 "family books") was compared with the time period after 1965 (1965-2002; 37 years; 79 "family books").

Table 2 shows that from 1938 to 1964 there were 12 books that depict a mother as the only parent and four books that depict a father as the only parent (a difference of eight books). In the time period of 1965 to 2002, there were 15 books that depict a mother as the only parent and 13 that depict a father as the only parent (a difference of two books). Chi square analysis reveals that the difference between the number of books that depict a father as the only parent in the earlier time period (1938-1964) and the later time period (1965-2002) is statistically significant, [chi square](2, n = 17) = 4.77, p = .03. The same analysis was performed to check whether there was a significant difference between the time periods for books that depict mothers only, and it was not found to be significant, [chi square](2, n = 27) = .33, p = .56. Based on this finding, it appears that the 1960s was indeed an exceptional point in time for the culture of fatherhood in American award-winning picture books, after which we saw an increase in the presence of fathers in books due to an increase in books that portray a father as the only parent. This analysis points to the possibility of the 1960s being a turning point, at least with regard to an increase in the presence of single fathers in Caldecott award winning picture books. However, presence is just one aspect of the changing culture of fatherhood.

The changing-culture-of-fatherhood research is also interested in points in time in which fathers and mothers are given "equal attention." Table 2 shows that, during the entire decade of the 1970s, the number of books that depict fathers and the number of books that depict mothers is equal. This is of interest because the foundational research related to the changing culture of fatherhood has also pointed to the decade of the 1970s as one of equity in gender in artifacts (LaRossa et al., 2000).

FATHER INTERACTIONS: VARIATIONS ACCORDING TO FAMILY STRUCTURE

Overall, mean father interactions per book are 1.59 (SD = 1.83) compared to 2.37 (SD = 2.40) for mothers. These overall means include all books with fathers and all books with mothers (including books with a father only and a mother only). These overall means were found to differ significantly at the .001 level. Significant differences were found between fathers' means and mothers' means in books with parents in the same family (p = .001) and books in which parents are not in the same family (p = .04).

Mean mother interactions are higher than father interactions, regardless of whether the book portrays a father and a mother in the same family or not. However, of note is the finding that fathers are depicted as interacting more with children when they are the only parent in the story (M = 2.18, SD = 2.60, n = 17) compared to when there are two parents in the same family (M = 1.41, SD = 1.54, n = 86) or when there are two parents who are not in the same family (M = 1.84, SD = 2.10, n = 25). It is reasonable that if a book depicts one parent, that parent would be depicted as interacting more than if there were another parent in the book. However, what is interesting here is that, while this is true for fathers, it is not true for mothers. Mothers were found to be depicted as interacting in fairly equal amounts whether or not they were the lone parent in the story (M = 2.33, SD = 3, n = 27) or a co-parent with a father (M = 2.29, SD = 2.21, n = 86).

With regard to changes in interactions over time, a one-way ANOVA did not reveal that there were significant differences between half-decade periods (for father interactions: F(13, 114) = .84, p = .62; and for mother interactions: F(13,124) = .37, p = .976.

Although no statistical differences were found between half-decade periods, the data can still be examined for low and high points in terms of father and mother interactions to determine the possible shape of changes in interactions over time. Table 3 reveals the highest mean for mother interactions is in the late 1940s, a time when father interactions are also high. However, the highest mean father interactions occur in the late 1980s, the only time period when father interactions are higher than mother interactions. The lowest interaction means for both fathers and mothers occur in the early 1960s. The time period for which father interactions and mother interactions are the most similar is the early 1990s, when the means differ by the least amount (.05). Also of note is the largest difference between mother interactions and father interactions, which occurs in the 2000s, when mother interactions exceed that of fathers by two per book. These uneven patterns in the depictions of the interactions of fathers and mothers might be described as "fluctuating" rather than as shifting, because there does not appear to be a trend established after any particular point in time.

PARENTING BEHAVIOR

Table 4 shows that fathers are portrayed in more books than mothers as verbally giving affection, teaching, and playing with children while mothers are depicted in more books than fathers as providing more physical affection for children. When the specific parenting behavior variables are combined into an aggregate variable that represents the range of diverse parenting behaviors for fathers and mothers per book, overall findings are that there are 14 books in which fathers are depicted as engaging in more types of nurturing behaviors than mothers and 15 books in which mothers are depicted as engaging in more types of nurturing behaviors than fathers. There are 10 books in which parents engage in an equal amount of positive/nurturing parenting behaviors (this does not include the number of books in which both parents are depicted as engaging in zero nurturing behaviors).

Recall that what is being measured is how many of the different behaviors are depicted in the books and not how many times these behaviors were depicted in the books. Caution must be taken when interpreting these results. It is especially challenging to analyze the possible changes in this variable over time because there are so few books in each time period that portray parents as engaging in positive/nurturing parenting behaviors. Therefore, this question is approached by looking at the midpoint of the data set for the three possible scenarios (father depicted as exhibiting more types of the defined parenting behaviors than mother, mother more than father, father and mother equal). This will be compared to the natural midpoint of the data set to determine whether the books are spread evenly throughout the data set or if books portraying the specific scenarios are located at particular points in time.

With regard to time periods, the natural midpoint of the data set is the decade of the late 1960s. There are just as many time periods before this point as after (half-decades). With regard to number of books, the midpoint of the data set occurs in 1968, which is in the same half-decade. The midpoint across the data set for books that feature fathers as engaging in more types of nurturing behaviors than mothers is located in 1950, indicating that this type of book was more common in the early period of the data set. Similarly, for books that feature mothers more than fathers according to the same criteria, the midpoint is 1955. However, the midpoint for books that depict fathers and mothers equally in terms of the range of parenting behaviors is 1991. From this we see that books in the early decades of the data set feature one parent as engaging in more types of parenting behaviors than the other (either the mother or the father), while the later decades of the data set feature both parents equally. It cannot be determined whether or not the pattern is a fluctuation or a shift, due to the characteristics of the data and the analysis, but what can be determined is that the change in depictions of parenting behaviors toward equal depictions has occurred recently.

DISCUSSION

INSIGHT INTO THE SOCIAL CONSTRUCTIONS OF FATHERHOOD

The findings of this content analysis support the assumption that Caldecott books provide depictions of fathers and mothers. The basis for this research is that a child's-eye view of the culture of fatherhood is presented to young people via the artifacts of children's literature. Based on the present data, the readers of these books are presented with the several norms or expectations for the roles of fathers. The most salient expectation is that both fathers and mothers are present in the lives of young children. By reading Caldecott award winning books, children may also expect that fathers interact with children less than mothers but that they interact with children more often when mothers are not present. With regard to behaviors of fathers, a norm represented in these texts is that fathers provide verbal affection, teach children skills, and play with children, but they provide less physical affection to children than mothers. Because the images of fathers have changed over time, readers of Caldecott books published after the 1960s may perceive fathers to be as interactive as mothers and more integral to the daily lives of children, whereas readers of books published prior to the 1960s may have perceived fathers as less interactive and perhaps more "distant" to their domain.

These are the norms/scripts with which children and possibly fathers (as readers of these texts) have engaged over the past six decades to form ideas and expectations related to the culture of (mainstream American) fatherhood. These ideas provide a portion of the conceptual material that readers of the texts "work with" to build shared understandings of the customs, expectations, and roles of male parents.

In general, fathers are depicted as being less interactive with children than mothers, yet, because only Caldecott books were used in this analysis, it cannot be assumed that this difference can be applied to non-award winning books. Expressed as a percentage of mean mother interaction, overall fathers are depicted as interacting with children 67% as much as mothers. This concurs with the research that indicates that real-life fathers engage with children less than mothers (Dowd, 2000; Pleck, 1997). However, in award-winning American picture books, the fictional fathers appear to interact with their children more than research estimates of the amount of time real-life fathers interact with their children, which has been found to be 43.5% as much as mothers (as reported by Pleck). This observation is not meant to suggest that the books mirror or attempt to mirror any social reality but rather that the images contribute to cultural ideas that may or may not approximate reality to the reader.

There are differences in both fathers' and mothers' direct interactions depending on whether the book portrays a mother and father who are in the same family, in different families, or the lone parent in the books. Fathers are depicted as interacting with children more when they are the only parent in the book. This may contribute to a cultural idea that fathers are expected to interact with children less than mothers, and in addition that they interact with children more often when a mother is not present. This concurs with research by Brayfield (1993) cited in Atkinson and Blackwelder (1993, p. 976) that men's time spent in childcare has been found to occur when the mother is unavailable.

Although fathers may be portrayed as less interactive, they are portrayed as engaging in just as many positive/nurturing parenting behaviors as mothers. In fact, fathers are found to be depicted as providing verbal affection, teaching children, and playing with children in more books than mothers. This finding sheds new light on the portrayals and roles of fathers, specifically that there are behaviors that fathers may be socially associated with other than earning money for their families. However, these categories of parental behavior may be biased theoretically toward male behaviors. For example, making judgment upon the child (by offering praise or affection) can be thought of as an act of power, and teaching children is an instrumental role. Further, active play with children has been socially sanctioned as a male behavior in Western cultures compared to non-Western cultures (Roopnarine, Hooper, Ahmeduzzaman, & Pollack, 1993).

THE CONTINUING DIALOGUE: HOW DOES THIS RESEARCH INFORM THE CHANGING-CULTURE-OF-FATHERHOOD LINE OF INQUIRY?

The changing culture of fatherhood research is concerned primarily with the timing, shape, and duration of changes in how fathers are portrayed in a variety of media. These "changes" relate to the constructs of presence (how many fathers are portrayed) and behaviors (how fathers are portrayed) and are generally compared to measures of mothers according to the same criteria. Another dimension of the inquiries into the culture of fatherhood involves the intersections of historical and cultural phenomena.

Comparison to Mothers. Fathers and mothers are found to appear in relatively the same amount of books in this collection, with mothers in slightly more books. This finding indicates that readers of these books will be exposed to relatively equal amounts of information about male and female parents. This finding concurs with LaRossa et al.'s (2000) finding that, overall, Mother's Day received more attention than Father's Day in comic strips. In award-winning American children's literature, mothers are given more attention than fathers--however, only slightly more attention.

Timing and Shape. The current study reveals a variation in the presence of fathers in the books from 1965 to 1969. This increase in father presence (especially as it is relative to mother presence, which had a decrease), indicates that the 1960s may have been a turning point in the culture of fatherhood as it is presented to young children. The difference in the presence of fathers and mothers can be accounted for by an increase in the number of books that depict a father as the only parent in the book. This turn can be characterized as a "shift" because analysis reveals that there is a significant difference in the number of books that portray a father as the only parent before 1965 and after 1965.

This concurs with Day and Mackey's (1986) findings with regard to the shape of change of images of fathers in media. While the 1960s can easily be identified as a turning point with regard to depictions of father presence, there is not a single period of time that can be clearly articulated as significant with regard to the depictions of father interactions. Because of this, while it appears that there has been a shift in father presence in the Caldecott books, it also appears that there have been "fluctuations" in the depictions of father interactions. The periods of time that figure prominently with regard to fluctuations in interactions are the 1940s, 1960s, 1980s, and 1990s. Therefore, this research supports both a shift during the 1960s and fluctuations throughout the past 60 years.

The data also reveal that the decade of the 1970s was the only time period in the data set during which fathers and mothers were in an equal number of books (in all other decades, mothers were in more books than fathers). This finding concurs with LaRossa et al.'s (2000) finding that fathers and mothers were referenced equally in comic strips in the 1970s. This indicates that gender parity in parenting was evidenced in more than one form of media during the 1970s.

Several features of the corpus of Caldecott award winning books combine to paint a picture of a "new father" image in the 1980s, which was found by other researchers (Atkinson & Blackwelder, 1993; LaRossa et al., 2000). There is a high mean of father interactions found in the Caldecott award winning books during the late 1980s as well as an increase in the number of books that depict the father as the only parent in the story. There is also a prevalence of books that depict fathers engaging in a variety of nurturing behaviors in equal amounts to mothers after 1991, demonstrating that the "new father" image found in the Caldecott collection appears to endure. However, caution must be taken when considering the concept of a "new father" for several reasons. First, while it seems that father interactions are high during the half-decade of the late 1980s, the high means can be attributed to a single book depicting a father as the only parent that has the highest number of father interactions in the collection (Owl Moon by Jane Yolen, illustrated by John Schoenherr, award year 1988). It is clear, then, that one book in a half-decade can skew the data. That is not to say that that one book is not sufficient to act as a cultural indicator. In fact, one book can indeed act as a cultural indicator (as evidenced by the practice in Western culture of identifying masterworks in a literary canon that come to represent a time period or an aspect of culture within that time period). Still, if there was an emergence of a "new father" in these books occurring in the late 1980s and early 1990s, we might be asking ourselves what happened to him. Father interactions have appeared to drop beginning in the late 1990s and in the first few years of the new millennium. To examine this issue further, we must first consider what the "new father" is compared to (i.e., who is the "old" father?). Is/was there an "old father?" Is it possible that the "new father" is not a new idea? For example, Table 3 shows there are similarities in mean interactions of fathers in the books between the late 1980s/early 1990s and the late 1940s/early 1950s. Also of note is the similar pattern of high means of father interactions (that lasts approximately 10 years) followed by lows in father interactions (seen in the late 1950s and late 1990s). This patterning suggests that there might be peak points in the history of the Caldecott award in which fathers are depicted as interacting highly with children (the early 1950s and early 1990). These peak periods are followed by swift drops in the depictions of father interactions. A similar patterning does not exist for the depictions of mothers. It appears, then, as if the books in the collection get saturated with high father interactions at certain points of time, after which we might predict that the images of highly interactive fathers will be low. It remains to be seen whether the lows in the depictions of fathers will continue to be low as they were during the decades of the 1960s and 1970s and throughout the next decades. Another cautionary point worthy of mentioning in this regard is that books need not portray fathers as interacting with their children to be of interest in the history of the culture of fatherhood. There are several dimensions to the portrayals of fathers in books such as the qualities of their behaviors (nurturing, playful, instrumental, etc.) and also whether or not the book takes up a father as the topic of the book. This aspect is beyond the scope of the current article but worthy of investigation.

Intersections with History and Culture. The examination of media artifacts such as children's literature allow for interpretations in the context of broader historical and cultural phenomena. Atkinson and Blackwelder (1993) found a relationship between what they called a "definitional shift" in fathering (from providing to nurturing) and the fertility rate in the United States. Their research included data from the early 1900s until the 1980s. They found from the 1940s to the 1950s, when fertility rates increased compared to prior decades (the post-war "baby boom"), magazine articles began to take up more topics related to fathers as providers. When fertility decreased in the decades after the 1950s (until the 1970s), provider themes in magazine articles decreased, favoring more nurturing roles of fathers. Since the 1970s, the fertility rate in the U.S. has remained flat. Atkinson and Blackwelder's data show nearly a complete parallel relationship between the fertility rate and the providing role of fathers in magazine articles. They also considered a possible relationship between this definitional shift in father images and women's labor-force participation but did not find a strong relationship between the linear increase in women's labor-force participation and oscillations in the topics of magazine articles.

The data from Caldecott award winning picture books could also be compared to societal trends such as fertility and women's labor-force participation; however, it is difficult to draw a relationship between these variables. Broadly speaking, there has been an increase in the number of books that depict fathers and mothers as "equal" in the range of parenting behaviors they exhibit, which might be parallel to the increase in women's labor-force participation. However, the increases in the numbers of books portraying parents "equally" has not been linear as has women's labor-force participation. All other variables (number of books portraying fathers and/or mothers, and father/mother interactions) show fluctuations that do not correspond with the trends in fertility (which increased until the 1950s and decreased until the 1970s, staying flat after that point). Certainly, labor-force participation and fertility rates are just two ways of looking at historical phenomena that might have a relationship to cultural artifacts representing families. Also, there are broader aspects of the cultural landscape that drive these trends, such as involvement in armed conflicts, world economic factors, globalization, and healthcare technologies (which impact family life through reproductive choices). Historical and cultural analysis is problematic in that there are many aspects to history and, as a consequence, many versions of the historical record. A limit to the current research is that these issues are not taken up fully; however, the work is intended to spark dialogue in this regard.

IMPLICATIONS

The findings of this research have implications for young children and those who care for them by providing documentation of children's culture, which places childhood as a social category worthy of examination. Young children are given voice via the acknowledgement of the artifacts that are produced in culture for their use. This is particularly important because young children are often marginalized due to their lack of physical, economic, and political power relative to adults (Corsaro, 1997). Research that focuses energy on the perspectives of young people helps to make their concerns and viewpoints more visible to the world at large.

With regard to family studies and the changing-culture-of-fatherhood line of inquiry, the current research of Caldecott award winning books supports the suggestions that there has been a shift in the portrayals of (the presence of) fathers as well as "fluctuations" in the images of fathers (interactions with children). The data point to the 1960s as an important time in the history of fatherhood and suggest that there appears to be an image of a "new father" that occurs in children's literature in recent decades.

Limits of this work include the use of an adult as primary researcher. This work provides information about children's artifacts and therefore gives a glimpse at what might be a child's vantage point in American culture. However, because an adult was the researcher, we cannot assume that children interpret the images in the same way that an adult does. interpretation is the mechanism with which the children will use the images to form ideas and ideals about fatherhood and is a major component in the process of forming identity. This limitation is addressed in the current work by structuring the research questions around the examination of only what is visible or observable within the artifacts of children's literature rather than examining the meanings of the artifacts to children.

Other limits are that the use of a specific collection of award-winning books does not provide a comprehensive understanding of how fathers are portrayed in all books for children. However, it was necessary to begin the line of inquiry into children's literature with a defined set of texts such as the Caldecott award winning books to provide a reasonable comparison to prior research that examines the changing culture of fatherhood. Using American award-winning books carries with it the benefit of also providing a comparison to award-winning books from other countries that grant awards. Further, this research of Caldecott award winning books also provides a baseline for future studies, which should examine best-selling books, public library collections, and adolescent literature.

FUTURE DIRECTIONS

It was originally thought that, if periods of time could be identified as standing out compared to others, a qualitative approach could help to answer questions related to how fathers are presented in these artifacts. However, given that so many time periods hold potential interest, it would not be prudent to look at these time periods exclusively, without an understanding of what happened during the time periods in between them. Rather than directing future research on time periods, a focus on specific variables and specific categories of books would yield more meaningful interpretations of the data. My recommendation for the next step in this process is to examine all books that depict positive/nurturing parenting behaviors of both fathers and mothers that were identified in this research (n = 41). This examination should go further than the present analysis and construct a new coding frame for a more indepth description of the books. I also recommend that all books that portray one parent in the story (either fathers as the only parent, n = 17, or mothers as the only parent, n = 27) be examined, for these are the books that account for the difference between the number of fathers and mothers found in these books. Future work should consider these questions, aligned with theories of cultural psychology, working toward the empowerment of children's perspectives by including children in the research work. This approach should ask readers what they think about the images and how the images might inform their ideas about fathers.

Correspondence concerning this article should be addressed to Suzanne M. Flannery Quinn, University of South Florida, 4202 E. Fowler Avenue, EDU 162, Tampa, FL 33620. Electronic mail: squinn@ tempest.coedu.usf.edu.

REFERENCES

Atkinson, M., & Blackwelder, S. (1993). Fathering in the 20th Century. Journal of Marriage and the Family, 55, 975-986.

Anderson, N. (2002). Elementary children's literature. Boston, MA: Allyn & Bacon.

Bauer, M. (2000). Classical content analysis: A review. In M. Bauer & G. Gaskell (Eds.), Qualitative researching with text, image, and sound (pp. 131-151). London: Sage.

Child, I.L., Potter, E.H., & Levine, E.M. (1946). Children's textbooks and personality development: An exploration in the social psychology of education. Psychological Monographs, 60, 1-54.

Collins, L.J., Ingoldsby, B.B., & Dellmann, M.M. (1984). Sex-role stereotyping in children's literature: A change from the past. Childhood Education, 60, 278-285.

Coltrane, S., & Allen, K. (1994). "New" fathers and old stereotypes: Representations of masculinity in 1980s television advertising. Masculinities, 2, 43-66.

Corsaro, W. (1997). The sociology of childhood. Thousand Oaks, CA: Pine Forge Press.

Day, R., & Lamb, M. (2004). Conceptualizing and measuring father involvement: Pathways, problems, and progress. In R. Day & M. Lamb (Eds.), Conceptualizing and measuring father involvement (pp. 1-15). Mahwah, NJ: Lawrence Erlbaum Associates.

Day, R., & Mackey, W. (1986). The role image of the American father: An examination of media myth. Journal of Comparative Family Studies, 3, 371-388.

Dowd, N. (2000). Redefining fatherhood. New York: New York University Press.

Grauerholz, E., & Pescosolidio, B.A. (1989). Gender representations in children's literature: 1900-1984. Gender and Society, 3(1), 113-125.

Heller, C. (1985). The image of father in the American family as depicted in children's picture books: 1946-1955 vs. 1973-1982 (sex role, literature, parent). (Educat.D. Dissertation, Columbia University Teachers College). Dissertation Abstracts International, 46(09), 2555.

Hillman, J.S. (1974). An analysis of male and female roles in two periods of children's literature. Journal of Educational Research, 8, 84-88.

Holsti, O. (1969). Content analysis for the social sciences and humanities. Reading, MA: Addison-Wesley.

Jacklin, C.N., & Mischel, H.N. (1973). As the twig is bent--Sex role stereotyping in early readers. School Psychology Digest, 2, 30-39.

Kortenhaus, C.M., & Demerest, J. (1993). Gender role stereotyping in children's literature: An update. Sex Roles, 28(3/4), 219-232.

Lamb, M., Pleck, J., Charnov, E., & Levine, J. (1985). Paternal behavior in humans. American Zoologist, 25, 883-894.

LaRossa, R., Jaret, C., Gadgil, M., & Wynn, G.R. (2000). The changing culture of fatherhood in comic-strip families: A six-decade analysis. Journal of Marriage and the Family, 62, 375-387.

LaRossa, R., Gordon, B., Wilson, R., Bairan, A., & Jaret, C. (1991). The fluctuating image of the 20th century American father. Journal of Marriage and the Family, 53, 987-997.

LaRossa, R., & Reitzes, D. (1993). Symbolic interactionism and family studies. In P. Boss, W. Doherty, R. LaRossa, W. Schumm, & S. Steinmetz (Eds.), Sourcebook of family theories and methods (pp. 135-163). New York: Plenum.

Lesnik-Oberstein, K. (1994). Children's literature: Criticism and the fictional child. Oxford: Clarendon Press.

Lesnik-Oberstein, K. (1998). Childhood and textuality: Culture, history, literature. In K. Lesnik-Oberstein (Ed.), Children in culture (pp. 1-28). London: MacMillan Press.

Malaguzzi, L. (2000). A charter of rights. In The hundred languages of children: Catalogue of the exhibit (pp. 214-215). Reggio Emilia, Italy: Reggio Children.

Pleck, J. (1997). Paternal involvement: Levels, sources and consequences. In M.E. Lamb (Ed.), The role of the father in child development (3rd ed., pp. 66-103). New York: John Wiley & Sons.

Quinn, S. (2000). Fathers in children's literature. Paper presented at the National Council on Family Relations Annual Conference. Minneapolis, Minnesota, November.

Ratner, C. (2000). Cultural psychology theory and method. New York: Kluwer Academic.

Rinaldi, C. (2001). Infant-toddler centers and preschools as places of culture. In Project Zero and Reggio Children Making learning visible: Children as individual and group learners. Reggio Emilia, Italy, Reggio Children.

Roopnarine, J.L., Hooper, F.H., Ahmeduzzaman, M., & Pollack, B. (1993). Gentle play partners: Mother-child and father-child play in New Delhi, India. In K.B. MacDonald (Ed.), Parent-child play: Descriptions and implications (pp. 287-304). Albany: State University of New York Press.

Super, C., & Harkness, S. (1997). The cultural structuring of child development. In J. Berry, P. Dasen, & T. Saraswathi (Eds.), Handbook of cross-cultural psychology: Basic processes and human development (pp. 1-39). Needham, MA: Allyn & Bacon.

Weber, R. (1985). Basic content analysis. London: Sage.

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

Whiting, J., & Child, I. (1953). Child training and personality. New Haven: Yale University Press.

Yolen, J. [author], & Schoenherr, J. [illustrator] (1987). Owl moon. New York: Philomel.

SUZANNE M. FLANNERY QUINN

University of South Florida

Table 1
Summary of Changing Culture of Fatherhood Research

Researchers (year)       Media Examined/                Magnitude
                         Years Examined                 of change

Day                      Cartoons in Saturday           Significant
  & Mackey (1986)        Evening Post
                         1922-1968 with
                         1971-1978
LaRossa et al. (1991)    Cartoons in Saturday           Significant
                         Evening Post
                         19,242,832,364,044
Atkinson                 Magazine Articles              Significant
  & Blackwelder (1993)   Middle years of 1980s-
                         1990s
Coltrane                 Classic TV                     Minimal
  & Allan (1994)         Advertisements
                         1950s and 1980s
LaRossa et al (2000)     Syndicated Comic Strips        Significant
                         1940-1999

Researchers (year)       Timing and reason              Shape
                                                        of change

Day                      1970s                          Shift
  & Mackey (1986)        Due to gender parity in
                         incompetence

LaRossa et al. (1991)    1930s/1940s and 1970s          Fluctuation
                         Due to gender parity in
                         incompetence
Atkinson                 1970s critical                 Fluctuation
  & Blackwelder (1993)   for image of "new father"

Coltrane                 Only slight changes; women     Not enough
  & Allan (1994)         portrayed more like men (as    change to
                         worker rather than parent)     hypothesize
                                                        a shape
LaRossa et al (2000)     1970s                          Fluctuation
                         Due to gender parity in
                         incompetence;
                         1980s
                         Dramatic increase in
                         nurturance and supportive
                         characterization of fathers

Table 2
Number of Caldecott Award Winning Books that Depict Fathers
or Mothers per Half Decade 1938-2002

Half decade period     38-39   40-44   45-49   50-54   55-59
Number of                9      20      27      30      20
  award books
Number of                9      16      16      15      10
  family books
Percentage of award    100      80      59.3    50      50
  books that are
  family books
Number of books          1       2       0       1       0
  father as only
  parent
Number of books          2       2       1       2       1
  mother as only
  parent
Difference in number    -1       0      -1      -1      -1
  of books that
  depict
  a father and a
  mother
Percentage of family    77.8    87.5    93.8    86.7    90
  books that depict
  fathers
Percentage of family    88.9    87.5   100      93     100
  books that depict
  mothers

Half decade period     60-64   65-69   70-74   75-79   80-84
Number of               16      16      22      17      20
  award books
Number of               10      10       7       8       9
  family books
Percentage of award     62.5    62.5    31.8    47      45
  books that are
  family books
Number of books          0       4       1       1       1
  father as only
  parent
Number of books          4       1       1       1       2
  mother as only
  parent
Difference in number    -4       3       0       0      -1
  of books that
  depict
  a father and a
  mother
Percentage of family    60      90      85.7    87.5    77.8
  books that depict
  fathers
Percentage of family   100      60      85.7    87.5    88.9
  books that depict
  mothers

Half decade period     85-89   90-94   95-99   00-02   Total
Number of               18      20      23      13     271
  award books
Number of               11      14      14       6     155
  family books
Percentage of award     61.1    70.0    60.9    46.2    57.2
  books that are
  family books
Number of books          2       2       2       0      17
  father as only
  parent
Number of books          3       4       3       0      27
  mother as only
  parent
Difference in number    -1      -2      -1       0     -10
  of books that
  depict
  a father and a
  mother
Percentage of family    72.7    71.4    78.7   100      82.6
  books that depict
  fathers
Percentage of family    81.8    85.7    85.7   100      89
  books that depict
  mothers

Table 3
Mean Interactions of Fathers and Mothers with Children
per Book by Half Decade 1938-2002

Half-decade      38-39   40-44   45-49   50-54   55-59
  period

Mean father       1.43    1.93    2.13     2      0.89
  interactions
Mean mother       2.75    2.71    3.13     2.5    1.9
  interactions
Difference       -1.32   -0.79   -0.99    -0.5   -1.01

Half-decade      60-64   65-69   70-74   75-79   80-84
  period

Mean father       0.67    1.44    1.67    1.43    1.57
  interactions
Mean mother       1.5     2       1.83    1.86    2.63
  interactions
Difference       -0.83   -0.56   -0.17   -0.43   -1.1

Half-decade      85-89   90-94   95-99   00-02   Overall
  period                                           mean

Mean father       2.25    2.20    0.73    0.83    1.59
  interactions
Mean mother       2       2.25    2.42    2.83    2.37
  interactions
Difference        0.25   -0.05   -1.69   -2      -0.78

Table 4
Number of Books that Portray Specific Positive/Nurturing
Parenting Behaviors

Number of books that depict:   Fathers   Mothers

Physical affection               18        23
Verbal affection                 11        10
Teaching                          8         7
Play                             11         8

Table 5
Distribution of Caldecott Award Winning Books that Depict Fathers
and Mothers as Engaging in Positive/Nurturing Parenting Behavior
per Half Decade 1938-2002

Half-decade      38-39  40-44  45-49  50-54  55-59  60-64  65-79
  period

Fathers and        0      0      1      1      0      0      0
  mothers equal
Fathers more       0      2      4      3      1      1      0
  than mothers
Mothers more       1      1      2      2      2      0      0
  than fathers

Half-decade      70-74  75-79  80-84  85-89  90-94  95-99  00-02
  period

Fathers and        0      1      0      0      4      2      0
  mothers equal
Fathers more       0      0      0      2      0      1      0
  than mothers
Mothers more       1      1      0      1      0      0      4
  than fathers