"Education, Faculty of"@en . "Curriculum and Pedagogy (EDCP), Department of"@en . "DSpace"@en . "UBCV"@en . "Imms, Wesley David"@en . "2009-03-25T19:51:48Z"@en . "1997"@en . "Master of Arts - MA"@en . "University of British Columbia"@en . "The purpose of this study was to determine if gender\r\ndifferences existed in rates of participation and academic\r\nachievement in art education.\r\nParametric and non-parametric statistical analyses\r\nwere conducted on data representing the final art grades of\r\n2,231 students from 59 countries, assessed by the\r\nInternational Baccalaureate Organization during the 1995/96\r\nschool year.\r\nStatistically significant results indicated that rates\r\nof academic achievement, rates of participation, and\r\nchoices of syllabus were gender-oriented in this\r\npopulation. Boys, in comparison with girls, were found to\r\nbe less likely to undertake art, to choose less\r\nacademically oriented syllabuses and to be out-performed\r\nacademically in art.\r\nIt was concluded that boys displayed a lassitude\r\ntowards art education that was consistent with a more\r\ngeneralized educational trend, currently the focus of neo-masculinist\r\ndiscussion.\r\nImplications of the findings of this study were\r\ndiscussed in reference to boys' level of visual literacy,\r\nthe relevancy of art curriculum to boys' specific\r\neducational needs, the extent of a \"feminine\" stereotype of\r\nart, and factors within art education which impact on how\r\nboys determine \"masculinity\".\r\nImplementation of \"relational\" research was urged to\r\ninvestigate the impact of art education on boys' formation\r\nof concepts of masculinity, and the potential role of art\r\neducation in neo-masculinist discussion currently exploring\r\ntheories of multiple masculinities."@en . "https://circle.library.ubc.ca/rest/handle/2429/6511?expand=metadata"@en . "3892910 bytes"@en . "application/pdf"@en . "ANALYSIS OF GENDER DIFFERENCES IN ART EDUCATION: RATES OF PARTICIPATION AND ACADEMIC ACHIEVEMENT IN INTERNATIONAL BACCALAUREATE ART AND DESIGN EDUCATION. by WESLEY DAVID IMMS B.Ed., University of South Australia, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN THE FACULTY OF GRADUATE STUDIES (Department of Curriculum Studies) We accept this thesis as conforming to the requiredvstandard THE UNIVERSITY OF BRITISH COLUMBIA September 1997 \u00C2\u00A9 Wesley David Imms, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The University of British Columbia Vancouver, Canada DE-6 (2/88) Abstract 11 The purpose of this study was to determine i f gender differences existed in rates of participation and academic achievement in art education. Parametric and non-parametric st a t i s t i c a l analyses were conducted on data representing the final art grades of 2,231 students from 59 countries, assessed by the International Baccalaureate Organization during the 1995/96 school year. Statistically significant results indicated that rates of academic achievement, rates of participation, and choices of syllabus were gender-oriented in this population. Boys, in comparison with g i r l s , were found to be less likely to undertake art, to choose less academically oriented syllabuses and to be out-performed academically in art. It was concluded that boys displayed a lassitude towards art education that was consistent with a more generalized educational trend, currently the focus of neo-masculinist discussion. Implications of the findings of this study were discussed in reference to boys' level of visual literacy, the relevancy of art curriculum to boys' specific X l l educational needs, the extent of a \"feminine\" stereotype of art, and factors within art education which impact on how boys determine \"masculinity\". Implementation of \"relational\" research was urged to investigate the impact of art education on boys' formation of concepts of masculinity, and the potential role of art education in neo-masculinist discussion currently exploring theories of multiple masculinities. i v Table of Gpntents Abstract i i Table of Contents * i v L i s t of Tables v i L i s t of Figures . v i i Acknowledgments v i i i Chapter I. Statement of the Problem 1.1 Introduction 1 1.2 Rationale for the Study 2 1.3 Research Questions 6 Chapter I I . Literature Review 2.1 H i s t o r i c a l Background. 9 2.2 Schooling and Gender 12 2.3 Boys and Schooling. 15 2.4 Boys' P a r t i c i p a t i o n and Academic Achievement in Art Education ..17 Chapter I I I . Method 3.1 Data Source i 21 3.2 Design of the Study 25 3.3 Procedure 29 3.4 Data Analysis 32 V 3.5 Limitations of the Study 32 Chapter IV. Results and Conclusions 4.1 Part i c i p a t i o n 35 4.2 Achievement 40 4.3 Summary of Findings 49 Chapter V. Discussion 5.1 Implications of Findings for Art Education 53 5.2 Suggested Further Research 61 References 64 Appendices. Appendix A; Countries Represented i n the Data 73 Appendix B; Description of the I.B.D. Syllabus 74 Appendix C; Assessment C r i t e r i a of the I.B.D. Art Syllabus 76 Appendix D; Calculation of Familywise Error Rate....82 Appendix E; Calculation of Ef f e c t Size... 83 Appendix F; Correspondence with the International Baccalaureate Organisation 84 v i List of Tables Table 1; Relative Frequencies of Girls' and Boys' Participation in I.B.O. and I.B.D. Art 35 Table 2; Frequency of Student Participation in Art, by Gender 38 Table 3; Contingency Table of Relative Frequency of Student Participation in Art by Gender 39 Table 4; Distribution of Grades (Gender and Syllabus Combined). ... 40 Table 5; Summary of Means; Grades, by Syllabus Level and Gender 43 Table 6; Summary Table for 2x3 (Gender x Syllabus) Randomized Groups ANOVA 46 List of Figures v i i Figure 1; Diagram of the Curriculum Structure of the International Baccalaureate Diploma 22 Figure 2; Frequency Polygon of Participation Levels, Gender by Syllabus 37 Figure 3; Histogram of Distribution of Grades, Gender and Syllabus Combined 41 Figure 4; Bar Chart of Distribution of Grades by Syllabus 42 Figure 5; Polygon of Frequency of Award of Each Grade, Value by Gender. 44 Figure 6; Bar Chart of Boys' and Girls' Mean Art Scores, by Syllabus 45 Figure 7; Plot of Significant Interaction Effect; Grades, by Syllabus Level and Gender 48 V11X Acknowledgments This project is one of many achievements that have stemmed from two years of family adventure. Without Christine's generosity, support, enthusiasm and valued advice; and Tyson and Phoebe's wonderful acceptance of a different way of l i f e , i t would never have come to fruition. I would like to acknowledge the Assessment Centre of the International Baccalaureate, Cardiff, Wales; and i t ' s Chief Examiner for art/design, Dr. Doug Boughton, for access to the raw data used in this study. From the Faculty of Education, University of British Columbia I would like to thank my committee members - Dr. Ron MacGregor and Dr. Don Allison for their generous advice and supervision. I would also like to thank Dr. Graeme Chalmers for his assistance and continued interest in this project. I am especially grateful to Dr. Anna Kindler, Chair of my thesis committee for her generous allocation of time to this project, the wise counsel she provided and most importantly, her friendship and support over the past two years. I look forward to continuing to learn under her mentorship. 1 Chapter I. Statement of the Problem 1.1 Introduction. A decade ago on a warm spring afternoon in a southern Australian school, I sat in a staff meeting listening with interest to a discussion concerning the students being awarded the annual academic prizes. Coming at the end of my f i r s t year of teaching, i t was a new experience for me, but a colleague sitting beside me whispered that a similar conversation occurred each year. The names being reviewed were not the concern; each was worthy of the honour. Rather, i t was the fact that they were virtually a l l girls - again. \"Where are the boys?\" the bewildered Principal asked. More than ten years later, Pat Clarke, an o f f i c i a l in the British Columbia Teachers Federation, asked exactly the same question. \"Where are the boys?\" (Clarke, 1997). While reviewing that province's high school academic awards he noticed that boys were conspicuously absent. In discussing this phenomena, he lamented a seeming lack of purpose displayed by boys and posited that, in trying to be \"manly\", boys were surrendering to a n i h i l i s t i c male sub-culture characterized by poor academic achievement and avoidance of social or educational challenge. 2 My experiences as an art educator inclined me to agree with Clarke. Since that staff meeting early in my teaching career, I became conscious of consistent gender differences in my art classes; boys seemed less likely than gir l s to choose art and when they did, were not as academically successful. Over the years my observation of boys' avoidance and poor academic performance in art was so consistent i t was seen by me to constitute a worrying trend. But was this trend a generalized phenomena? If i t was, did i t constitute an educational concern? 1.2 Rationale for the Study Unfortunately, the queries raised above could not be answered from current literature. Very l i t t l e has been written about boys' participation and achievement in art education, principally because gender discussion in this f i e l d had traditionally followed limited topics of conversation within which boys' issues rarely featured (Packard and Zimmerman, 1977; Pariser and Zimmerman, 1990). This has been, however, characteristic of education as a whole. Since the early 1970s i t had been convincingly argued that curriculum was boy-oriented; the construction of curriculum, the focus of texts, pedagogical strategies and academic streaming have been centered on masculinist 3 meritocracy (Askew and Ross, 1988). Consequently i t was research on girls and their problems that warranted particular attention. Until recently no rationale had been established that justified discussion specific to males. As a result very l i t t l e is now known about boys' experiences generally in education (Bushweller, 1995). Why Should We Look at Boys' Schooling? Within the last few years a re-examination of conventional parameters of gender discussion has occurred, prompted by evidence of increased social, gender-interaction and schooling problems of boys. Some worrying social trends have emerged. Boys have been shown to be four times as likely as girls to commit suicide, be murdered by high school age and twice as likely to be victims of assault or robbery (Bushweller, 1995). They now form the majority (up to 75%) of behavioural and emotional disorder diagnoses (Soderman and Phillips, 1986), the majority of truancies, sexual misconduct and assaults (Jackson and Salisbury, 1996). Boys' impact on the schooling of girls has increasingly become an issue. Inappropriate behaviour by some boys in class is observed to demand disproportionate amounts of teachers time and limit g i r l s ' academic achievement (Sadker and Sadker, 1994). In addition, some 4 boys' sexist attitudes are considered to affect g i r l s ' levels of self-esteem and limit their personal and social potential (Gray, 1987). Perhaps the most significant - and recent - trend has been the identification of some worrying academic problems for boys. They now form the majority (up to 75%) of special education enrollments in both the U.S.A. (Bushweller, 1996) and the United Kingdom (Sbderman and Phillips, 1986). They are disproportionately represented in retention rates between classes (Lee and Bryck, 1986) and have lower academic success and lower expectations than girls of future career prospects (Pascal and Bertram, 1995). Clarke (1997) notes that in a number of school districts in the Canadian province of British Columbia, boys in the 1995/96 school year were 5 times as likely as girls to drop out of school; they achieved 80% of the failing grades, 80% of the suspensions, only 20% of academic honour awards and one third of d i s t r i c t and provincial scholarships. Clarke (1997) and Duffy (1996) also note the decrease in male enrollments in undergraduate courses from 60% in the 1970s to 40% in the 1990s. Facts such as these highlight a perceived c r i s i s with boys. Boys appear to be giving up on their schooling, rapidly becoming what Clarke (1997) refers to as a new 5 \"social underclass\", or what the Economist magazine (Men, tomorrow's second sex, 1997).terms the \"newest at-risk section of our community.\" These concerns have resulted in the emergence of a rationale that argues boys problems are becoming so c r i t i c a l , the paucity of research that exists must be immediately addressed. Subsequently, an increasing body of writing has concentrated largely on subject-specific research. Academic subjects such as English (Martino, 1995; Nilan, 1995; Jackson and Salisbury, 1995; Gilbert and Gilbert, 1995) and Social Science (Smith, 1995) have begun to correlate boys' social, gender-interaction and schooling problems with perceptions of what boys consider \"manly\" behaviour. Art education has been largely under-represented in this discussion and w i l l continue to be limited in i t s ability to offer subject-specific dialogue on this issue until i t can establish a construct describing the nature of boys' participation in the visual arts. The questions posed in the introduction, therefore, are timely and relevant. The observed phenomena w i l l be of educational concern as they w i l l indicate the degree to which boys' lassitude towards education, as outlined in recent gender discussion, applies to art. This w i l l be achieved through 6 documentation of trends within a large international population of boys' participation levels, academic achievement and syllabus selections. This study w i l l provide foundation data which further studies might u t i l i z e to build a discussion concerning the interplay of boys, art and masculinity. 1.3 Research Questions The research question for this study asked: In the International Baccalaureate population, are boys' levels of participation and academic achievement, and their choice of syllabuses, significantly different to girls'? In addressing this query, specific s t a t i s t i c a l tests were used to explore the effects of the dependent variables of grade1 and participation on the independent variables of gender and syllabus level. Two question clusters were utilised. The f i r s t cluster aimed at determining i f gender differences existed in participation in art, and i f those rates of participation were consistent between a l l International Baccalaureate Diploma (hereafter I.B.D.) art/design syllabus levels. The second cluster focused on achieved 1 Within this context, \"grade\" refers to the academic mark assigned to students work u t i l i s i n g normal assessment procedures. 7 grade, attempting to determine i f gender differences existed in academic results in I.B.D. art, and whether those academic results were consistent in a l l syllabus levels. Question cluster 1\u00E2\u0080\u00A2 In the I.B.D., is boys' participation in art significantly different to that of girls'? Are these findings consistent across syllabus levels? 1.1 Does significant difference exist in boys' overall participation in the I.B.O. program compared to g i r l s ' overall participation in the I.B.O. program? 1.2 What is the participation rate of students undertaking I.B.O. art compared to overall student participation in the I.B.O. program? 1.3 What is the proportion of boys within the I.B.O. program who study art, compared to the proportion of girls within the I.B.O. program who study art? 1.4 Is the rate of participation in art of boys significantly different than for girls? 1.5 Is the rate of participation in art for a l l students (boys and girls) significantly different between the three syllabuses utilised in the International Baccalaureate Art/Design program (610, 611 or 612)? 8 1.6 Is the rate of participation in art in any of the three (610, 611 or 612) syllabuses significantly different for boys than for girls? Question cluster 2. In the International Baccalaureate, is boys' academic achievement in art significantly different to that of girls? Are these findings consistent across the 610, 611 or 612 syllabus levels? 2.1 Is achievement in art, as measured by standard grading practices, significantly lower for boys than for girls? 2.2 Is achievement, as measured by standard grading practices for a l l students (boys and girls) significantly different between any of the three syllabuses? 2.3 Is achievement in art in any of the three syllabuses, as measured by standard grading practices, significantly different for boys than for girls? 9 Chapter II. Literature Review. 2.1 Historical Background. Traditional Histories of Gender in Education Traditionally, histories of gender debate in education have concentrated on the three decades of transformative socio-political changes witnessed during the 1970s, 1980s and 1990s in many Western societies (Askew and Ross, 1988; Gray, 1987). During th i s time a gender-oriented power imbalance in schools has been identified and continues to be addressed. Of concern has been an inequality of educational opportunity, an inequality of employment opportunity, sexist orientations of teachers curriculum and texts, and sex-role stereotyping; a l l to the disadvantage of g i r l s . Subsequent research focused on the following key themes of discussion; the advantages of co-education to provide more equitable educational and career opportunities (Dale, 1969, 1971, 1974; Austin, 1977a, 1977b), strategies to encourage gi r l s ' participation in traditionally male-oriented academic subjects (Byrne, 1978), the empowerment of girls to lead (Graham, 1974), to aspire to professional careers (Tidball and Kistiakowsky, 1976) and the negating of sex-role stereotyping through identification and 10 modification of sexist texts and teaching practices (Taylor, 1981). As a result g i r l s ' subject choices were extensively examined and boys domination of the classroom and schooling environment was effectively challenged (Weiner, 1985). Feminism and Gender in Education The source of the gender-equity related strategies which culminated from these research agendas have been widely attributed to radical feminist ideology of the early 1970s (Gray, 1987; Soerensen, 1992; Smith, 1995). Women were committed to addressing poor academic achievement by gi r l s , restrictive role stereotypes, lower career expectations, female submisiveness and a lack of assertive s k i l l s by girls in the classroom. By the 1980s significant legislation in many countries had resulted in the focus of gender work in education concentrating on the empowerment of girls through an \"equal opportunities\" approach (Jackson and Salisbury, 1996). The rewards for girls from these efforts have been so significant that Anne Soerensen, a feminist academic, wrote in 1992 \"...we are now witnessing a breakdown of the former educational pyramid. Women as a whole, but especially the young generation, have achieved a more equal position and 11 they have moved into many former male dominated subjects and areas.\" (p.2 01) . Such has been the domination of the debate and the advances made by feminists, the term 'gender issues' in education is how often seen as synonymous with women' s issues (Sacca, 1989). The Emergence of Neo-Masculinism2. Some contemporary interpretations of the history of gender work in education now recognize that, with the predominance of feminist ideology, an educational paradigm too specific to allow gender work with boys was created (Connell, 1989; Martino, 1995; Jackson and Salisbury, 1996). While acknowledging a considerable debt to feminist struggles since the 1970s, neo-masculinist authors describe the historical domination of the feminist perspective in gender discussion as debilitating for the development of a dialogue concerning male identity (Smith, 1995). It has not allowed room for men to discuss issues related to boys (Jackson and Salisbury, 1996). The resulting discussion, emerging only in the last few years, has been to examine the interplay of feminism and the emerging neo-masculinity 2 The term \"neo-masculinism\" and \"neo-masculinity\" have not received wide usage in gender literature but are considered useful within this discussion as representing those writers whose philosophy endorses a re-interpretation of masculinity from the traditional \"unitary\" to a more pluralist \"multiple\" definition. 12 movement in terms of a re-interpretation of definitions of masculinity. Martino (1995) recognized how, historically, feminist domination of discussion in gender led to the utilization of a \"unitary\" definition of masculinity which stereotyped males as patriarchal and hegemonic. Jackson and Salisbury (1996) observed a similar historical trend, and questioned how boys could be encouraged to take responsibility for their own change when this definition allowed them such l i t t l e room to move, and awarded such l i t t l e credit for efforts boys make to question perceptions of masculinity. Connell (1989) examined the debilitating effect of feminist ideology on boys' classroom discussion of gender issues, resulting in a subsequent suppression of a male-led mandate to effect masculine change. 3.2 Schooling and Gender The neo-masculinist perspective has gained an element of acknowledgment within gender debate. Although some feminist authors, such as Pallotta-Chiarolli (1990), Burgess (1990) and Bailey (1996) appeared reticent to abandon ideologies that had gained so many advantages with such hard work, more generalist feminist opinions coincided with neo-masculinist theories to create a post-modern paradigm within gender debate in education (Soerensen, 13 1992). This perspective questioned, after years of f r u i t f u l service, the continued appropriateness of \"equality of opportunity\" measures. Anne Soerensen (1992) argued a re-examination of the historical development of gender work in schools would only further feminist interests and f a c i l i t a t e future development in that same direction. She posited that, in retrospect, radical and liberal feminism appeared to have developed ideologies that were not pluralist. In contrast post-feminism i s , from a theoretical viewpoint challenging the former concepts, maintaining that we have to leave the essential thinking and absolute strategies of the 197 0s in order to open up a f i e l d of research to plurality and a constructive uncertainty, (p. 201) Soerensen (1992), Reay (1990b), Kruse (1992) and others have re-focused, to some extent, feminist debate on boys. Central to post-feminist philosophy is a recognition of the need to include a greater understanding of what makes boys behave as they do (Reay, 1990b). These researchers recognize boys schooling as a previously ignored variable in gender equity debate; \"...with the advent of neo-masculinism feminist perspectives have been forced to 14 reconsider their most fundamental question: Who are the losers and who are the winners?\" (Soerensen, 1992; p, 209) However traditional themes in gender discussion s t i l l constitute a majority of writing. Single-gender schooling and segregated classes have enjoyed a renaissance as a arena to discuss g i r l s ' improved academic performance (Jiminez and Lockheed, 1989), a venue for interventionist pedagogical strategies with boys (Gray, 1987; Kruse, 1992; Reay, 1990b), and a site for anti-sexist, anti-gender stereotyping pedagogies (Riordan, 1990; Burgess, 1990). Within art education, the most recent gender publication in North America - Gender issues in art education: Content, contexts and strategies (Collins & Sandell, 1996) - continues to focus on concerns with gir l s schooling in art education; topics such as \"Art Criticism from a Feminist Point of View\" (Congdon, 1996), \"Teaching Feminist Art and Social Activism\" (Wryick, 1996) and \"Feminist Interventions in Teaching Art History\" (Attenborough, 1996) point to where the emphasis lies in art education. While the editors of this work claim a wider sphere of contexts are now being represented within gender discussion in art education, these contexts remain limited to \"...the recent and dramatic evolution of 15 concerns related to women, art and education.\" (Collins and Sandell, 1996; p.xi). This exclusion of masculine issues within art education is inconsistent with generalised trends in education. Gender research is now being seen as emerging from two decades dominated by feminist oriented \"equality of opportunity\" discussion, into a phase of discussion more representative of contemporary theories of pluralism (Soerensen, 1992). A pluralist paradigm has been created within the topic of schooling and gender discussion, part of which acknowledges the unique problems boys face in today's educational environment, and the relational aspects of those problems to g i r l s ' needs. 2.3 Boys and Schooling Within this new construct of gender discussion in education, studies such as this one obtain their validity. Establishing a sense of boys academic participation and achievement is a necessary precursor for further investigation of the complex causes of boys' actions in schools (Martino, 1993). This study w i l l , through the vehicle of boys' levels of literacy in art education, point to directions the f i e l d ought to take i f pluralism is to be achieved. Without this knowledge art education, with i t s 16 limited history of research specific to boys, is handicapped in its ability to participate in current discussion concerning boys and schooling. The nature of this discussion identifies the complex socio-cultural influences on boys' perceptions of masculinity as implicit to their levels of literacy. Alloway, Davies, Gilbert and Gilbert (1996) summarised the consensus of opinions of many authors when they stated \"...we can't understand boys' literacy problems unless we understand boys' developing sense of masculinity and how literacy f i t s within this.\" (p. 6). The parameters for contemporary discussion concerning boys and their schooling are established within the concept that boys' academic performance is primarily influenced by social, rather than psycho-biological, effects. Contemporary research in this f i e l d is beginning to concentrate primarily on the interplay of boys' perceptions of gender and academic worth. Connell (1989) presented the theory of \"multiple masculinities\". Noting that in the 1980s a variety of types of masculinity existed - \"cool guys\", \"swots\" or \"wimps\" - into which boys were streamed according to pre-existing mores constructed by the school, society and youth sub-cultures. Often this streaming was fought by the youths but such was the influence of peer pressure and the 17 social constructs of school and home, their efforts were seen to be in vain (Jackson & Salisbury, 1996; Smith, 1995). Connell's analysis of boys struggling within limiting interpretations of masculinity has become the pivotal theme of neo-masculinist discussion. Martino (1994, 1995) and Nilan (1995) u t i l i z e de-construction/re-constructionalist pedagogies to investigate limiting gender stereotypes within English education and offer similar analyses in relation to instruction in Social Science (Smith, 1995). Mac an Ghaill (1994) also used the theory of multiple masculinities to explore their potential to negate boys' marginalization of sexual (homosexual) ethnic and disadvantaged masculinities. 2.3 Boys' Participation and Academic Achievement in Art Education. Literature concerning boys' participation and achievement in art is very limited. Art education's traditionally narrow perspective when researching gender-related topics (Helgadottir, 1991) has resulted in data on this topic being limited to appendant information in larger, more generalized studies. Unfortunately, the majority of these have concentrated on limited subject areas - usually mathematics, English, and science (MacCann, 18 1995) - with l i t t l e i f any information being gathered and discussed relevant to art. Standardized tests, available through national s t a t i s t i c a l archives, have been shown to be inconsistent in their results; girls perform poorly compared to boys on standardized tests while outperforming them in the classroom (Sadker, Sadker and Steindam, 1989). MacCann (1995) conducted a comprehensive analysis of examination and assessment results of the New South Wales (Australia) Higher Schools Certificate. A section, specific to art, with a sample of n = 8,478 found a disproportionate representation of girls in art (68%). Academically, gir l s outperformed boys in a l l areas of the art assessment, with the exception of the 90th and 95th percentiles in the examination section where boys marginally outperformed g i r l s . Overall, boys showed a wider standard deviation; significantly so in the lower end of the distributions. Gender differences in participation in art education were included as inconclusive appendant information in studies concerning subject preferences in co-educational and single-gender schools by Omerod (1975), Lee and Bryck (1986), Trickett, Castro, Trickett and Schaffner (1982), and Stables (1990). Of interest from these studies was a consistent observation of polarization tendencies of the 19 subject categories within which art was represented. By polarization, i t is meant the tendency for gender differences in subject choice to be accentuated in co-educational schools (Department of Education and Science, 1975). No studies were located that specifically investigated gendered differences in choice of art as a subject of study. This could be seen to form part of the \"vacuum of knowledge\" noted by Pariser and Zimmerman (1990), who called for \"...traditional research to undergird [the] conjectures...\" (p.5) evident in art education research in gender issues. In summary, i t is conceded that a study examining participation and achievement in relation to gender w i l l inevitably and unavoidably be challenged to discuss i t s results in terms of a diversity of socio-political and socio-cultural issues. These issues, such as sexism, gender-stereotyping or equality of opportunity have historically been discussed within education from a predominately feminist perspective. With a growing awareness of the compounding problems of boys in contemporary schooling, the traditional feminist constructs of gender are being viewed - both by some feminists and other academics - as too limiting iri their interpretations of masculinity. As a result, post-feminism and neo-20 masculinism are advocating a more pluralist approach towards issues of gender. This approach allows studies such as this to work from the rationale that boys may well be disadvantaged by factors within schooling. The identification of these factors, and the development of strategies to address their impact on boys' inappropriate behaviour and sexist gender-concepts w i l l ultimately serve the needs of both boys and g i r l s . 21 Chapter III. Method. For the purpose of this study, a large sample was required of male and female students from a population of art students with a homogeneous syllabus and assessment method. 3.1 Data Source. A primary source of data which f u l f i l l e d these cr i t e r i a was obtained through the International Baccalaureate Organization, Cardiff, Wales (hereafter I.B.O.). The I.B.O. provided scores of 2,231 male and female students from 59 countries (Appendix A) who were examined in their art/design program during the 1995/96 school year. This number represented the f u l l I.B.O. art/design population assessed in the Northern Hemisphere for that year. The data was obtained through negotiations with the I.B.O. The data was compiled by I.B.O. staff from a l i s t of requirements supplied by the researcher (see Appendix F). It was supplied in computer printout format for transcription to data f i l e and s t a t i s t i c a l computer analysis by the author. 22 The I.B.O. Program. The I.B.O. offers a 2 year program culminating i n the award of an I.B. Diploma (I.B.D.). The I.B.D. i s a pre-t e r t i a r y award, internationally recognized as a university entrance q u a l i f i c a t i o n . Strong results i n the I.B.D. are seen to ensure entrance to major un i v e r s i t i e s worldwide (Anderson, 1994). Group ! LANGUAGE B / A2 Group 3 LANGUAGE A l Group 5 MATHEMATICS Extended Essay Theory of Knowledge Creativity, Action & Service Group 1 EXPERIMENTAL SCIENCES Group 4 INDIVIDUALS & SOCIETIES Group 6 THE ARTS / SIXTH SUBJECT One more subject from groups I -4 or Advanced Mathematics SI. or Art / Design, Music, Classical Languages, Computing Studies a School Based Svllabu Figure 1. Diagram of the curriculum structure of the International Baccalaureate Diploma. 23 The program constitutes a comprehensive, multi-disciplinary course of study, operating within a school-based curriculum development model and u t i l i z i n g moderated, criteria-based assessment procedures. Students study 6 subjects over a two year period - 3 at the Higher Level, or H.L., and 3 at the Subsidiary Level or S.L. (see Figure 1). The levels of study are differentiated by contact hours - 240 for the H.L., 120 for the S.L. - and also by curriculum content which is seen as encompassing a greater commitment and range of s k i l l s development within the H.L. In addition to studying the six subjects, f u l l diploma students are expected to complete an extended essay, follow a course in the theory of knowledge, and undertake extra-curricular activities. It is also possible for students to study one or two individual I.B.D. subjects, for which they can be awarded individual certificates (Chalmers, 1988). Art within the I.B.D. Art/Design is included within the sixth group together with 6 other options. The aims of the art/design component of the I.B.D. encompass the development of high quality, personally meaningful, aesthetic, imaginative and creative s k i l l s with a pluralist emphasis on the developing of visual literacy s k i l l s (Chalmers, 1988). 24 Its curriculum is offered as three syllabuses; the H.L. with a studio/research curriculum, the S.L. (studio) or the S.L. (research); of which the students choose one. The H.L. syllabus encompasses a combination of studio and research components. Students are required to develop a personal focus to visual art production through the disciplines of criticism, art production and art history (Blaikie, 1994). Appendix B provides an outline of the course and the c r i t e r i a for the studio and research components. Assessment procedures for I.B^D. art/design. An externally assessed, portfolio display/interview method of assessment is utilized. An external examiner - usually a practicing art educator - conducts an interview with each student, assessing the work through discussion, review of the research workbook, and examination of the portfolio. Students may present a research workbook, or a portfolio of studio work for assessment at either of the subsidiary 612 or 611 syllabuses. Appendix C provides the c r i t e r i a on which assessments are made. Grades are awarded on a 1 to 5 ordinal scale for each of the 6 c r i t e r i a with 5 being the highest possible score. Scores are weighted by the I.B.O. (see Appendix C) to give proportional credit to \"imaginative and creative thinking\" 25 (studio component) and \"independent research\" (research component). A final grade award on a 0 (failure) to 7 (excellent) ordinal scale is produced. 3.2 Design of the study; Because a l l data was obtained from existing f i l e s of past art students within the I.B.O. program, a non-experimental, ex post facto design was utilized. The intention was to address the research questions u t i l i z i n g both descriptive and inferential statistics. I n i t i a l l y , the subjects represented by the data were treated as a population on the basis that they represented the entire collection of events in which we were interested (Howell, 1997, p2). This allowed a descriptive analysis of data to answer some research questions using distribution tables, means and standard deviations. Secondly, gender differences in rates of participation and academic achievement scores were tested for significance u t i l i z i n g the inferential statistic models of chi-square (x 2 ) and factorial analysis of variance (ANOVA). Differing inferential statistic models were considered necessary for analysis of significance of data on participation and achievement. 26 Choice of Statistical Models Participation. The data concerning participation rates was presented as the frequency of participation occurring within the categories of gender (boys and girls) and syllabus (610, 611, 612). The appropriate non-parametric s t a t i s t i c a l test for these conditions was considered to be an independent-samples chi-square test (Howell, 1997). This test allowed questions to be answered that concerned associations or relationships of 2 or more populations based on frequencies of observations in categories (Schumacher and McMillan, 1993). When analysis of participation data concerning one independent variable was required, the appropriate stati s t i c was considered to be the goodness-of-fit chi-square test. Achievement. The data on achievement grades was presented as the means of factors with two or more levels. Traditionally, comparison of means has been undertaken u t i l i z i n g a variety of parametric s t a t i s t i c a l tests. Of those relevant to the types of variables being analyzed, the use of multiple independent sample t tests, or multiple one-way ANOVAs were considered inappropriate due to the 27 higher probability of familywise error 3 (a^ = .15), (Appendix D). Factorial analysis, on the other hand, allowed for analysis of more than two independent variables, each with more than two levels with no danger of familywise error. The 2-way ANOVA required for this study enabled a l l three questions to be answered in one test; two concerning main effects, and one concerning interactions. Statistical Designs Design of the chi-square. An independent-samples, 2 x 3 (gender x syllabus) chi-square design was utilized for tests conducted with 2 factors with more than two levels. The dependent variable was frequency of enrollment. The independent variables were gender - with i t ' s two levels, boys and girls - and syllabus, which contained three levels - 610 = high; 611 = subsidiary syllabus, studio based; and 612 = subsidiary syllabus, research based. Goodness-of-fit chi-square tests were conducted for analyses concerning only one factor. Assumptions. Each observation (subject) was represented in only one category, independent sampling was assured, and sample sizes were well above the minimum 3 \"Familywise error\" describes the tendency for incorrect findings to be compacted by repeated testing. 28 required of a 5 (Glenburg, 1987). The assumptions were considered to be met. Design of the ANOVA. A 2 x 3 (gender x syllabus) factorial analysis of variance (ANOVA), randomized groups design, was utilized. The dependent variable was grade represented on a 0 (fail) to 7 (high) ordinal scale. The independent variables were gender, with two levels (male and female) and syllabus with three levels (610 = high; 611 = subsidiary syllabus, studio based; and 612 = subsidiary syllabus, research based). Tukey's Honestly Significant Different post-hoc comparisons (hereafter Tukey's HSD) were implemented on the independent variable of syllabus level to identify syllabus effect. Assumptions. Statistical conventions have traditionally required the meeting of data, sample and population assumptions before ANOVA results can be considered valid (Cone and Foster, 1996). Data assumptions required the use of interval or ratio scales of measurement- The data for this study was not interval or ratio; rather, i t was presented on an eight point ordinal scale. However, recent literature has convincingly argued the robustness of ANOVA when u t i l i z i n g ordinal data (Howell, 1997; Boneau, 1960 cited in Howell, 29 1997; Box, 1953, 1954a, 1954b cited in Howell, 1997; Bradley, 1964 cited in Howell, 1997). This assumption was considered met. Sampling assumptions required the use of independent sampling. Independent sampling requires that each subject be represented only once within the data. This assumption was considered met. Population assumptions required normal distributions and comparable variances (Glenburg, 1986). Both distributions were reasonably symmetrical in distribution and variances reasonably homogeneous. Furthermore these requirements have been somewhat nu l l i f i e d by recent literature which has convincingly argued ANOVA's robustness to non-conformity and non-normality (Boneau, 1960; cited in Howell, 1997). These assumptions were considered to be met. 3.3 Procedure. Variables Within the Obtained Data Set. The I.B.O. supplied, for each student, data concerning gender, final academic grade in art, syllabus level iri which the art score was obtained - given as both a category and syllabus number - the participant's school code number, the schooling type (singler-gender of co-educational), the 30 total students (by gender) in each school enrolled in the I.B.O. program and the country in which the school was situated. Variables chosen. The independent variables chosen from this l i s t for this study were gender, syllabus level, and total of students (by gender) studying in the I.B.O. program. These were selected on the grounds that the primary objective of this study was an investigation of boys participation and achievement in art. Supplementary variables were rejected i f perceived to confound the st a t i s t i c a l v i a b i l i t y of this objective. Variables rejected. Although some interest existed in possible differences in art scores between countries, the variable of country was eventually rejected. This was done because 1) the I.B.O. data contained a large degree of representation of ex-patriate students studying in international schools that could negate the true 'international' classification of this variable; 2) widely disparate sample numbers (see Appendix A) between countries would make st a t i s t i c a l conclusions suspect, and 3) a 2 x 3 x 59 (gender x syllabus x country) ANOVA would not be able to successfully compute higher order interactions with the existence of empty cells in some syllabuses from some countries. 31 The variable of schooling type was of considerable interest; the representation of boys and their achievement in single-gender schools in comparison to co-educational schools was significant in light of current neo-masculinity discussion. However, with only one boys' single-gender school in the data 4 the validity of any findings was questionable. The variable representing the three specific syllabuses (610,611,612) was chosen in favour of the variable representing the two categorical levels (High or Subsidiary). The choice of this variable allowed an interpretation of trends in boys' participation in studio-oriented or research-oriented art education curriculum, and boys 1 academic achievement in those orientations relative to g i r l s . If required, descriptive statistics could s t i l l be utilized to examine trends concerning the categorical levels by simply combining the 611 and 612 data. Treatment of 'N' and 'P' grades. A number of scores were presented in the data as 'N' (failed to adequately meet criteria) and 'P.' (pending; absent from assessment, failed to meet c r i t e r i a ) . On 4 The I.B.O. data incorrectly classified at least one all-boys school as 'mixed', leading to the conclusion that more errors in the data might exist. A reasonable sample might be possible for future studies. 32 advice from the I.B.O. these scores were interpreted as a f a i l , and awarded an '0' rating. 3.4 Analysis of Data. Data was analyzed using the Statistical Package for the Social Sciences (SPSS). Alpha Level. Traditionally, convention in statistics in the Social Sciences has accepted an alpha level of a = .05 as an appropriate level of significance (Howell, 1997). It allows a low probability (.05) of Type I errors (a) while retaining a reasonable power (1-B) to detect Type II errors (6). This alpha level was adopted for this study. Power Analysis. With a two-tailed test, an a = .05, an effect size (Appendix E) of d = .40, and a sample size of >n ~ 750, Cohen's (1988) Power Table indicated the study demonstrated a power of >.99, or a >99% probability of correctly rejecting a false null hypothesis. 3.5 Limitations of the study Generalization of the Results; Results from the descriptive s t a t i s t i c a l analyses were applicable only to the population of students assessed in 33 the Northern Hemisphere I.B.O. art program in 1995/96. Inferential s t a t i s t i c a l analysis of participation (chi-square) and achievement (ANOVA) was generalized only to students in the I.B.O. program in other years including those from the Southern Hemisphere population. Any s t a t i s t i c a l l y significant results could not be assumed to be indicative of the whole art education population; principally because the data collection did not f u l f i l l the required assumptions for randomized sampling to allow generalization. In as much as the purpose of this study was to explore i f the phenomena concerning the researcher's classroom observations of boys participation and achievement rates was more generalized, limiting the conclusions to the 2,231 I.B.O. art students more than f u l f i l l e d this objective. However, i t should be recognized that the I.B-O. data provided a comprehensive source of art education data. The sample size was very large. Data was obtained from art students in a range of school grades (10, 11 and 12). It represented students from over 200 schools - private and public, co-educational and single-gender, church and non-church. The schools represented a wide range of countries - 59 in a l l - from each of the five continents. The data was obtained using homogeneous methods; identical 34 curriculum and assessment c r i t e r i a were used and moderation procedures were utilized to assure homogeneity in assessment procedures. As such, findings from the I.B.O. data can be interpreted as representative of a wide, art education population. Results from this study could be viewed as indicative of important trends in art education. Causation The s t a t i s t i c a l analysis from this study was not appropriate to be used to interpret causation of identified trends. It identified a situation; further specific quantitative or qualitative studies are needed to investigate why any trends identified by this study exist. 35 Chapter IV. Results and Conclusions. 4.1 Participation. Considerable gender differences concerning student participation were detected in both the overall I.B.O. program and within the art program (Table 1). Boys and girls were not evenly represented in the overall I.B.O. assessments for the 1995/96 school year; of the 24,252 students assessed, gir l s constituted 53.45% Table 1. Relative Frequency of Girls' and Boys' Participation in I.B.O. and I.B.D. in Art. Boys G i r l s Enrollment n % n % I.B.O. 11,289 .4655 12,963 .5345* Art (I.B.D.) 793 .3554 1,438 .6445** * x 2 = (1/ N = 24,252) = 115.54 p<.005 ** x 2 = (If N - 2,231) = 186.47 JK.005 36 (12,963) and boys 46.55% (11,289) (x 2 = [1, N = 24,252] = 115.54 p<.005). In answering question 1.1, i t was concluded that a significant difference existed in the overall representation of boys in the I.B.O. program, compared to the overall representation of g i r l s . Question 1.2 and 1.3 asked what proportion of the overall I.B.O. population undertook art and the proportion represented in that figure of girls and boys. It was found that 2,231 or 9.20% chose art as the optional subject within the sixth group. 1,438 (11.09%) of g i r l s chose art, while 793 (7.02%) of boys chose art. Question 1.4 asked i f the differences in enrollments of girls was significantly different to the enrollment of boys. Girls were found to significantly over-represent boys in art (x 2 = [1, N = 2,231] = 186.47 p<.005). They accounted for 1,438 or 64.5% of art enrollments compared to boys' 793 or 35.5% of art enrollments. Overall participation decreased significantly between the higher and lower syllabuses (see Figure 2). 1,257 students (56.3%) were assessed in the higher 610 syllabus, 778 (34.9%) and 196 (8.8%) were assessed in the lower 611 and 612 syllabuses respectively. In answering question 1.5 - does participation differ within the syllabuses -37 significance difference was found ( x 2 = [2, N = 2,231] 759.25 rj<.005) . 1000 Gender Boys Girls 610 611 612 Syllabus levels Figure 2. Frequency Polygon of Participation Levels; Gender x Syllabus. Question 1.6 asked i f significant gender differences occurred in any of the three syllabuses. In addition to an overall (boys and gi r l s combined) decrease in numbers across the three syllabuses, significant gender differences were also detected in participation rates. Figure 2 and Table 2 illustrates the significant drop in participation rates between the syllabuses by both genders (x^ 38 [2,N=2,231] = 17.38 p_<.005) and boys significant under-representation in each syllabus (610 syllabus, x 2 [1,N=1,257] = 142.35 p<.005; 611 syllabus, x 2 [1,N=778] =42.58 p_<.005; 612 syllabus, x 2 [1,N=196] = 8.16 p<.005). An interaction between syllabuses was also evident - boys were more likely to study in the 611 and 612 syllabuses. Table 2. Frequency of Student Participation in Art, by Gender. syllabus Gender n 610* 611** 612*** Male 793 417 298 78 Female 1,438 840 480 118 * X2= [1* N = 1,257] = 142.35, p<.005 ** X2- [1/ N = 778] =42.58, p<.005 ***X2= [1, N = 196] = 8.16, p<.005 This can be seen in Table 3, where the relative (%) frequency of boys participation compared to girl s was found to be greater in the lower, 611 and 612 syllabuses. In 39 comparison to g i r l s , boys showed a trend of lower proportional participation in the higher syllabus, a significant trend towards proportionately higher participation in the lower syllabuses, and a tendency to study the studio-oriented rather than research-oriented lower syllabus. Table 3. Contingency Table of Relative Frequency of Student Participation in Art, by Gender. syllabus Gender 610 611 612 Male .526 .376 .098 Female .584 .334 .082 Overall, gi r l s clearly over-represented boys in participation rates in this assessment of the I.B.O. art program by nearly 2 to 1. In addition, gendered enrollments differed significantly between syllabuses, leading to the conclusion that not only did fewer boys 40 study art, their choice (proportionately) of which syllabus to study differed from that of g i r l s . 4.2 Achievement. Disparate gender differences were also found in achievement levels in art within the I.B.O. program. Table 4. Distribution of Grades, (Gender and Syllabus Combined). Score Frequency % Cumulative % 0 120 5.4 5.4 1 13 0.6 6.0 2 54 2.4 8.4 3 156 7.0 15.4 4 436 19.5 34.9 5 607 27.2 62.1 6 594 26.6 88.7 7 251 11.3 100.0 When combined, the final grades for a l l students in a l l syllabuses (N = 2,231) were distributed around a 41 population mean of u, = 4.791, and a population standard deviation of a = 1.675. A slight bimodal distribution of scores was seen in the distribution of grades (Table 4) with the proportion of students scoring \"0\" higher than students scoring \"1\" and \"2\". This trend was not significant (measure of kurtosis = 1.466). A mean of M = 4.791, median of 5.00 and mode of 5.00 resulted in a negative skew to the distribution. This 700 6001 500 400 300 200 100 0 \u00E2\u0080\u00A2 1 1 1 _ . -| * I -Std. Dev = 1.67 Mean = 4.8 N = 2231.00 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 grade Combined for gender, syllabus. Normal distribution curve superimposed. Figure 3. Histogram of Distribution of Grades, Gender and Syllabus Combined. skew was not seen as significant (measure of skewedness = -1.188). The resultant histogram of distribution of grades 42 (Figure 3) indicated that a normal d i s t r i b u t i o n of grades could be assumed. Grades were i n i t i a l l y analyzed by syllabus, with gender combined, to determine the pattern of d i s t r i b u t i o n i n individual courses and to compare scores between syllabuses. The bar chart of the d i s t r i b u t i o n of grades by syllabuses (Figure 4) indicated roughly normal dis t r i b u t i o n s i n each of the syllabuses with the trend of a 610 611 612 Figure 4. Bar Chart of D i s t r i b u t i o n of Grades, by Syllabus. negative skew more pronounced i n the 610 syllabus. It also i l l u s t r a t e d a trend of consistently decreasing mean scores 43 from the 610 to the 612 syllabuses, the mean of scores being M = 5.09, M = 4.42, M = 4.37 for the 610, 611 and 612 syllabuses respectively. Variances increased as means decreased between the syllabuses, with standard deviations being found of SD = 1.53, SD = 1.77, SD = 1.86 for the 610, 611 and 612 syllabuses respectively. Gender differences were apparent in overall art scores when examined with the syllabuses combined. Girls, with a mean of M = 4.87 (SD = 1.64) out-performed boys in overall art scores. Boys scored an overall mean of M = 4.66 (SD = 1.73). Table 5. Summary of means. Grades, by Syllabus Level and Gender. Males Females i 1 Syllabus n M n M 1 . - V 610 417 4.91 840 5.18 611 298 4.49 480 4.37 612 78 3.94 118 4.65 When examined by syllabus, girls out-performed boys in two of the three syllabuses. Girls showed a higher mean 44 score than boys i n the 610 syllabus and 612 syllabus. Boys showed a higher mean score than g i r l s i n the 611 syllabus (see Table 5.). A study of d i s t r i b u t i o n of actual scores values by gender (see Figure 5) indicated a d i s p a r i t y of a l l o c a t i o n of grades i n t h i s d i s t r i b u t i o n . Boys appeared to be more l i k e l y to score lower grades - at the 2/3 l e v e l - and g i r l s more l i k e l y to score higher grades at the 6/7 l e v e l . 30 20 10 c: .15 There existed a .15 (15%) probability that one of the significant results had incorrectly rejected a null hypothesis (that i s , committed a Type I error). 83 Appendix E. Calculating Effect Size A pr i o r i power analysis for this stiudy required estimation of a population standard deviation (o), and effect size (d). Population standard deviation. Based on the limiting range of an 8 point scale for grade, a a = 1.50 was considered appropriate. Effect size. Howell (1996) stipulated that an approximation of effect size was preferable to arbitrarily assigning one of Cohen's (1988) three levels of d. It was considered that on a score range of 0 - 7 a difference of \u00C2\u00B1.6 between boys' and g i r l s ' mean scores would represent a considerable difference in grades. A Cohen's d = .40 was calculated using the formula, d = uL - JU? o This effect size conformed roughly to Cohens' (1988) \"middle\" size; larger effect sizes would have increased the danger of Type II errors, while a smaller effect size would have allowed for significant results to be found that would not, in practice, be taken seriously. Appendix F. Correspondence with the International Baccalaureate Organisation. "@en . "Thesis/Dissertation"@en . "1997-11"@en . "10.14288/1.0055190"@en . "eng"@en . "Curriculum Studies"@en . "Vancouver : University of British Columbia Library"@en . "University of British Columbia"@en . "For non-commercial purposes only, such as research, private study and education. Additional conditions apply, see Terms of Use https://open.library.ubc.ca/terms_of_use."@en . "Graduate"@en . "Analysis of gender differences in art education : rates of participation and academic achievement in international baccalaureate art and design education"@en . "Text"@en . "http://hdl.handle.net/2429/6511"@en .