a different population and a somewhat different story

6 downloads 0 Views 634KB Size Report
112 females enrolled in psychology or computer science cours- ... of the images were male, while 34% were ..... Francis Crick (reported together), Bill Nye.
THE DRAW A SCIENTIST TEST:

A DIFFERENT POPULATION AND A SOMEWHAT DIFFERENT STORY MARK D. TRIOMAS Departmentof Psychology MississippiState University

TuAcy B.

HENLEY

Department of Psychology and Special Education Texas A & M University-Commerce CATHERINE M. SNELL

Department of Psychology MississippiState University This study examined Draw-a-Scientist-Test (DAST) images solicited from 212 undergraduate students for the presence of traditional gender stereotypes. Participants were 100 males and 112 females enrolled in psychology or computer science courses with a mean age of 21.02 years. A standard multiple regression generated a model that accounts for the' variability in the sexes of drawings consistent with past findings. The focus of this research, however, was in comparing the results of our sample (college students) 'vith previous studies that have that have used the DAST with much younger (e.g., elementary-aged) students. Results were strikingly similar, suggesting either that gender stereotypes are widely persistent even among college science majors, or that the DAST may not be a particularly sensitive measure despite its wide use.

The Draw a Scientist Test (DAST) was first utilized by Chambers (1983) to examine stereotypic views of scientists among school children. Chambers' initial study examined the strength and presence of "modem sanitized,' and "older mythic," stereotypic images-of the scientist in 4,807

children's (ages 5 to II years) drawings that were collected from 1966 to 1977. Chambers assessed the presence of'lab coats, eyeglasses, facial hair, symbols of research, symbols of knowledge, technological aspects, and captions that were believed to represent a stereotypical view

Author Note We thank Donna Reese for access to the computer science department.,We also wish to thank her and Lisa Henderson for the recruitment of participants from computer science. Correspondence concerning this article should be addressed to Tracy Henley, Department of Psychology and Special Education, Texas A & M University - Commerce, Commerce, Texas 75429. E-mail: [email protected].

140

College Student DAST Images...!/141 of scientists. Chambers reported that 49% of this sample consisted of girls, but participants produced only 28 drawings of female scientists (0.56%). All of these female scientists were drawn by female participants. Subsequently, Fort and Varney (1989) gathered 1,654 drawings of scientists through a national contest for 2nd through 12th grade school students. Fort and Varney received 135 drawings of female scientists. They reported that, "the 8% depicted by our respondents is close to reflecting reality," (p. 9) given their estimate that women then made up roughly 6% of the scientific and engineering workforce at that time. A careful analysis of DAST images reported by Newton and Newton's (1992) survey of 1,143 primary school children (ages 4-11) suggests children draw more females at a younger age, but by the sixth school year, males were drawn by 83% of the participants. Following up, Newton and Newton (1998) surveyed 1000 children from reception (the UK equivalent of preschool/kindergarten) to grade 6 and, "...concluded that there were few significant changes to primary pupils' conceptions of science and scientist...' (p. 1148). Over time, the number of female scientists drawn has slightly increased. This may be due to changes in social perceptions, or to refinements in the measure. For example, Matthews (1994) had 132 children from years 7, 8, and 10 generate two different drawings, out of this total, 66% of the images were male, while 34% were female. Likewise, Maoldomhnaigh and Mho-

lain (1990) considered the effects of test administration instructions, eventually changing the DAST instructions to "Draw a Man or Woman Scientist:' Stated thusly, 367 children (299 females and 68 males) between the ages of 11 and 16 years provided drawings. Although boys in this sample almost exclusively drew males, 49% of the girls drew a female. Brosnan (1999) modified the task by re-framing it as a, "draw-a-computer-user-test." For Brosnan, whose sample consisted of 395 children ages 5 to 11 years, males performed similar to males who complete the standard DAST. Interestingly, 70% of the females drew a female computer user, while only 4% of males drew a female computer user. Variations of the DAST have been utilized in the U.S. and Canada (e.g., Parsons, 1997), Ireland (e.g., Maoldomhnaigh & Hunt, 1990), Finland (e.g., Raty, 1997), England (e.g., Brosnan, 1999), Korea (e.g., Song & Kim, 1999); and Taiwan (e.g., She, 1998) with similar results. A recurrent finding in the DAST literature is that scientists are stereotyped as being male by girls and this may serve as a limiting factor in their self-efficacy toward becomifig scientists. Maoldomhnaigh and Mholain (1990) observed, "If boys are at all possessed of the image of the female scientist, it is too deeply hidden to be uncovered by the stratagems used in this study," (p. 72) and Chambers (1983) speculated, "...that, when asked to, 'draw a scientist,' even scientists themselves utilize the standard image" (p. 256). At times, the DAST literature has been used to emphasize the need for teaching reforms with respect to gender bias (e.g.,

142 / College Student Journal

Kahle, et. al, 1993; Newton & Newton, 1998). Perhaps not surprisingly, media has also been seen as perpetuating this male scientist stereotype. Steinke and Long (1995) noted that twice as many male scientists were represented as female scientists on the four major television science programs. Having noted that, at least some dramatic perceptual changes on matters of gender and science have occurred in recent decades. For example, in an article titled The College-Student Image of

the Scientist(Beardslee & O'Dowd, 1961), male college students were asked about the attractiveness of various scientific disciplines, while female college students were asked about their preferences for having men of science for husbands. The purpose of this study is twofold. First, it is to examine DAST drawings produced by college students for the presence of traditionally interpreted stereotypes. That is, to attempt to replicate the basic findings of the DAST as gender stereotypes are presumptively dynamic, and as such, subject to change over time. Secondly, our goal is to explore the similarities and differences between a college-aged sample with the previous results obtained from school children. Method Participants

The participants were 212 undergraduate students enrolled in psychology or computer science service courses at Mississippi State University. The sample consisted of 47 freshmen (22.2%), 48 sophomores (22.6%), 62 juniors (29.2%), and 55 seniors (26%), who voluntarily par-

ticipated, or participated for class credit (age m = 21.02, sd= 3.14, range = 18-31). The sample was 52.8 % female (n = 112) and 47.2 % male (n = 100). Participants represented 41 different majors with 62.9% (n = 132) having majors in scientific disciplines, 34.3% (n = 72) having majors in nonscientific disciplines, and 2.8% (n = 6) were undeclared students. Participants were tested in extant courses or experimental groups. Testing took place in classrooms in either the psychology department (n = 131, 61.8%) or the computer science department (n = 81, 38.2%). Materials

The study consisted of completing two forms. The first form was a standard size sheet of paper with the following instructions printed one-half inch from the top margin: You have up to 15 minutes to complete this portion of your participation. It would be reasonable to utilize the entire 15 minutes. Please do not talk to the other people who are participating in this study, as they are busy working as well. In the space below, please DRAW A SCIENTIST. The second form was a brief questionnaire used to ascertain age, year in school, major, participant's sex, sex of drawing, whether the participant drew a famous scientist (and his/her identity), identity of the participant's favorite scientist (and reason for being such), type of scientist drawn, number of instructors for the current semester in the department tested, current number of male and female instructors in the department tested, and sex of the last

College Student DAST Images...!/143

person with whom the participant had a substanative conversation. Procedures

All surveys were administered by one female experimenter or one male experimenter. During class-time administrations (n = 3), the instructors were not present in two classes, and a female instructor was present in one. The instructors' sexes were 1 male and 2 female. Participants were told the study consisted of 2 tasks and their instructions were on the forms. Participants were also told they would get no clarification about the instructions and no participant asked for clarification. After the 15 minutes for drawing elapsed, participants were asked to cease drawing and turn their drawings face down. Participants then received the questionnaire. After completing the questionnaire, participants were debriefed. One participant withdrew his drawing after debriefing, but allowed his questionnaire to be used. Chambers' (1983) original scoring system was used for all but one of the stereotypical science indicators. Lab coats were scored for presence (without regard to color), eyeglasses were defined as two lenses (with or without arms), and facial hair was any facial hair growth (e.g., moustaches, beards, jaw length or longer sideburns, 5 o'clock shadow). Symbols of research were scientific instruments and laboratory equipment (e.g., microscopes, pencils in pockets, test tubes, Bunsen burners, clip boards). Symbols of knowledge were items that represent having knowledge (e.g., books, filing cabinets, blackboards, posters with the periodic

table, a double helix, or other figures and tables). Symbols of technology were any products of science (e.g., computers, telephones). Captions were scored differently than the traditional method, as they were scored just for presence or absence, and not for content. The operational criteria for technical aspects of the drawings were defined a priori by consensus between the researchers. Drawings generated by 17 graduate students (not used in this study) were jointly scored by the raters in order for training. All drawings were independently rated by 2 clinical psychology graduate students. Results Sex of Scientists The reported sexes of the drawings were 82% male (n = 173), 12.3% female (n = 26), and 5.7% nonspecified (n = 12). The drawings indicated as nonspecified included stick people, a drawing with members of both sexes, and a drawing of a cartoon dog. Female participants generated 2 nonspecified (1.8%), 92 male (82.1%), and 18 female (16.1%) drawings, while male participants drew 10 nonspecified (10.1%), 80 male (81.8%), and 8 female (8.1%) drawings. An ANOVA found a significant difference between sex of drawing and participants' sex (F = 8.123, p < .005). Types of Scientists Participants reported the types of scientist drawn as chemists (55.2%, n = 117), biologists (13.7%, n = 29), psychologists (4.7%, n = 10), physicists (4.2%, n = 9), computer scientists (2.8%, n = 6), nonspecific (8.5%, n = 18), an astronomer (n

-------------------

144 / College Student Journal

=1, .5%), a natural scientist (n = 1, .5%), and 11 participants failed to report the type of scientist drawn (5.2%). The 10 remaining cases (4.7%) are believed to be sarcastic, or noncompliant, responses and include 3 mad/deranged scientists (1.4%), 3 nerd/geek scientists (1.4%), and 4 other comic responses (1.9%) with one each, "Elvis," "pothead" "dumb," and an "electric" scientist (that is, one who was being electrocuted). Analysis of Drawings Interobserver agreement was calculated on 17 items for 211 drawings. The raters' separate data were compared to detect rater differences. Initial agreement was 3,381 out of 3,587 chances to agree (94.3%). The 206 disagreements were independently rated a second time by the same raters and agreement was 3,533 (98.5%). The remaining 54 disagreements were resolved by conference and the majority of those involved lab coats, symbols of knowledge, facial hair, and captions. Lab coats occurred in 54.5% (n = 115) of the drawings. Scientists wore eyeglasses in 145 drawings (68.4%) and had facial hair in 44 drawings (20.8%). Evidence of research was present in 148 drawings (69.8%). Symbols of knowledge were present in 27 drawings (12.7%), and symbols of technology were present in 21 drawings (9.9%). Captions were present in 71 (33.5%) drawings. The mean number of 'stereotypical' indicators present is 2.92 per drawing, which compares to Chambers (1983) reported 3.05 for 4th grade children and 3.26 for 5th grade children. An ANOVA detected a significant difference between

class year and lab coats (F = 3.31, p < .021). All other indicators were not significantly different between class years. While 41 drawings (19.3%) had smoke coming from apparatuses, in 2 drawings the scientists (1%) were smoking tobacco (or possibly marijuana), and 1 (0.5%) had smoke coming from the scientist's ears. Computers were present in 10 drawings (4.7%). A name tag was present on 29 (13.7%) scientists, with 8 being illegible and the others representing either wellknown fictional or nonfictional scientists, including 5 local computer science and psychology faculty members, Kevorkian and Einstein were among others named. Most DAST scientists had a traditional hairstyle (n = 117, 55.2%). However, 44 had wild hair (20.8%), 38 were bald (17.9%), 4 (1.9%) had nontraditional hair (e.g., males with long hair, or bald females - excluding stick figures), and 8 (4.2%) were undeterminable due to being covered by hats/scarfs, or being headless. Earrings were visible on 3 males scientists (1.4%) and no female scientists. The scientist was represented by a simplistic stick person in 12 drawings (5.7%), or an elaborate and more "artistic" stick person in 9 drawings (4.2%). FavoriteScientist The favorite scientist reported was Albert Einstein (n = 105,49.5%). Given the principle result, it was curious that Marie Curie was the second favorite (n = 7, 3.3%). The third favorites were instructors with whom the participant had classes (n = 6, 2.8%), which included 2 current psychology department faculty members, 2 computer science department faculty members (n = 3), and a former middle school

College Student DAST Images.../145

science teacher. Next were Isaac Newton and George Washington Carver (n = 5, 2.4% each). StephenHawkings, Benjamin Franklin, and Sigmund Freud were listed (n = 4, 1.9% each), and Thomas Edison and Dr. Frankenstein/mad scientist were rated sixth most popular (n = 3, 1.4% for each). Emmet Brown (from the Back to the Future Movies), James Watson and Francis Crick (reported together), Bill Nye (from the television show Bill.Nye the Science Guy), Alexander Graham Bell, Galileo Galilei, Richard Feynman, and Johannes Kepler were each listed twice (0.9% each). Twenty-three different scientists were mentioned once (0.5% each) and includ-. ed, among others, Robert Oppenheimer, Charles Darwin, Erik Erikson, Michael Faraday, Carl Sagan, Ivan Pavlov, Karen Horney, Dmitri Mendeleev, and Jean Piaget. Nonexistent scientists, such as The Professor (from Gilligan'sIsland), Dexter (from the cartoon, Dexter's Laboratory), and Brain (from the cartoon Pinky and the Brain), along with the participant (self as favorite), a participant's aunt, a participant's uncle, and a participant's father were also mentioned once. Twenty-nine participants did not report a favorite scientist and an additional 13 failed to report a reason for their selection. The 170 reported reasons for being a participant's favorite scientist can be classified into 7 categories: the scientist's theories or intellectual contributions (n = 66, 31.1%); being the only scientist the participant could think of (n = 41, 19.3%); the scientist was famous or funny (n = 22, 10.4%); the scientist was smart, brilliant,

or a genius (n = 20, 9.4%); the scientist's hair or appearance (n = 11, 5.2%); the scientist was the participant's instructor (n = 2, .9%); and 8 participants (3.8%) cited other reasons that do not readily fit a classification scheme. Multiple Regression Using Mahalanobis distance to detect outliers (Mertler & Vannatta, 2002), three cases were identified as extreme. These cases were inspected and determined to be older students, or students with a heavy class load whose instructors were all the same sex. After due deliberation all of these cases were included in the analysis. An evaluation of linearity revealed no concerns. However, drawings reported as having a nonspecific sex were excluded from analysis, because most people are male or female and the nonspecific cases may be instances of reactivity. A standard multiple regression analysis was conducted to evaluate the prediction of DAST image sex from 10 variables (age, year, participant sex, favorite scientist, female instructors, male instructors, conversation, experimenter sex, science, and Einstein). Regression results indicate that the overall model significantly predicts DAST image sex, R1 = .133, R2adj =.075, F (10, 151) = 2.31, p < .015. The model accounts for 13.3% of the variance in sex of DAST images. A summary of regression coefficients is presented in Table 1 and indicates that only three (participant sex, favorite scientist, and experimenter sex) of the ten variables significantly contributed to the model. The 95% confidence interval slopes are participant sex .004 to

146 / College Student Journal Table 1 SAnmftY

it SmrltmeoY,s ReVesion An al$sk PYeodkt-ig DASTImage Sex II

a

AZe

-.0173

-.115

YeUr

-.0113

Sex

.10so

Favofite

I

2.1700

F1 MI

.0500 .02S0

Converation

.0490

.109 .077 ~.101*

i .325

-.320 2.051 2.12S 1.319 1.106 .995 -2.095 1.579

3749

-.009

.042

.202

-.026 .165

.035

.207

.171

.139 .271

.099 .072

.107 .090

.321

.059

.0s1

-.132 Epeimenter -.1250 Science .0890 .144 .022 .127 P17At WY,-) EinsWen 2047 MWJA ------------W." 11t71? Note: Age - age of participmnt, Ytar - year in school, Sex - sex of pmticipant, Favoiite - sex oi &volite scientist, El -nuMber of Lfmale instractols in the depulment teltd for the c'lTent seester, MI - number of mate instnictors in the department ttzted for the clnuent Semester, Conveā€¢sation - sx oa the last person with whom the paticipant had a gennine convelsation, Expeiimenter -sea ofI)AST adminiistrator, Science -in a science =jor, Einstein -Einstein as favolite scientist.*p < .05.

.212, favorite scientist .015 to .418, and administrator -.243 to -.007, and do not contain the value zero. Discussion One could anticipate more complex representations in drawings as an individual develops a more complex knowledge of the scientist, which may account for the increase in stereotypic indicators from first through fifth grade school children reported by Chambers (1983). However, this sample of college undergraduate students' drawings exhibited similar stereotypic characteristics to those of Chambers' fourth and fifth grade children. Symington and Spurling (1990) challenged the traditional interpretation of the DAST and speculated that children's drawings may

represent a perceived public stereotype and not a personal belief or perception. It appears that the drawings generated by this sample support that idea. Indeed, the DAST is a tool initially used

to reveal a child's attitudes and beliefs about science, and the DAST has traditionally been interpreted the same way for children in any grade. The drawings generated by these undergraduate students are similar to those of children that appear in the DAST literature, but at some level this sample cannot have the same beliefs and attitudes that are inferred by traditional DAST interpretation, because 62.9% are majors committed to a college science degree. Thus, it appears that some changes in attitudes and beliefs that occur between kindergarten and college, may not be well

College Student DAST Images...!/147 72% of the students enrolled in doctoral captured by the DAST. Specifically, we submit that many fac- programs were women. It seems reasontors go into a drawing and some cognitive able to expect that a stereotype of the idea(I)s may be expressed in drawings, scientist as being male might then not exist while others may not be. In part this expres- in the field of psychology. Yet, even psysion could be limited by such matters as chologists are subject to societal biases subjective artistic ability, sex of task admin- and those women whose major was psyistrator, task demand, and motivation. chology (n = 52) predominately drew male Chambers (1983) claimed the DAST was scientists (n = 43). These women are future useful in identifying children's attitudes scientists, but traditional interpretation because it doesn't rely on a verbal response would indicate that most (82.7%) believe and can be utilized at an early age. Others that scientists are still best illustrated by (e.g., She, 1998) have simply asked sub- men. Although we will conclude by observjects their views. With college students, either approach is potentially transparent ing that Symington and Spurling's (1990) to the thoughtful subject. But, as these data public stereotype account is certainly a feaillustrate, that was not a problem with this sible explanation for these women's apparent dissonance, it should be noted sample. that, stick people were drawn by about 10% of the participants and it is doubtful Conclusions even 10% of undergraduate students that age a mean In theory, this sample with scientists are stick figures. In short, believe of 21.02 years ranging from 18 to 31 years old, represents the generation after Cham- perhaps another result of this study is to bers' (1983) sample of 5 to 11 years old question the effectiveness of the DAST as children from 1966 to 1977. The results a tool for truly understanding perceptions can been seen as suggesting a shortcom- of scientists. ing in the educational system, because drawings generated by this sample are References remarkably similar to those of Chambers' Wicherski, M., & Kohout, J. L. M., R., Auguste, study. From a perspective across time, Office. Research APA (1999). exposure to years of science education has http://research.apa.org/mes96contents.html#c ontents seemingly had no effect on these participants. Or, as a direct comparison between Beardslee, D. C., & O'Dowd, D. D. (1961). The college student image of scientists. Science, the two samples suggests, college students 133, 997-1001. think no differently about science and genBrosnan, M. J. (1999). A new methodology, an old der than elementary school children. story? Gender differences in the "draw-a-comThe APA Research Office (Auguste, puter-user" test. European Journal of Wicherski, & Kohout, 1999) reported that Psychology and Education, 14, 375-385. 72% of master's degree recipients in psychology were women. They also reported

148 / College Student Journal Chambers, D. W. (1983). Stereotyped images of the scientist: The draw-a-scientist-test. Science Education, 67, 255-265. Fort, D. C., & Varney, H. L. (1989). How students see the scientist: Mostly male, mostly white, and mostly benevolent. Sci&nce and Children, 26, 8-13. Kahle, J. B., Parker, L. H., Rennie, L. J., & Riley, D. (1993). Gender differences in science education: Building a model. Educational Psychologist, 28, 379-404. Maoldomhnaigh, M. C., & Hunt, A. (1988). Some factors affecting the image of the scientist drawn by older primary school pupils. Research in Science and TechnologicalEducation, 6, 159-166. Maoldomhnaigh, M. C., & Mholain (1990). The perceived expectation of administrator as a factor affecting the sex of scientists drawn by early adolescent girls. Research in Science and Technological Education, 8, 69-75. Matthews, B. (1994). What does a chemist look like? Education in Chemistry, September, 127-129 Mertler, C. A., &Vannatta, R. A. (2002). Advanced multivariate statistical methods: Practical applicationand interpretation.(2nd ed.). Los Angeles, CA: Pyrczak Publishing. Newton, D. P., & Newton, L. D. (1992). Young children's perception of science and the scientist. International Journal of Science Education, 14, 331-341.

Newton, L. D., & Newton, D. P. (1998). Primary chilren's conceptions of science and the scientist: Is the impact on the National Curriculum breaking down the stereotype? International Journalof Science Education, 20, 1137-1149. Parsons, E. C. (1997). Black high school females' images of the scientist: Expression of culture. Journal of Research in Science Teaching, 7, 745-768. Raty, H., & Snellman, L. (1997). Children's images of an intelligent person. Journal of Social Behavior and Personality,12, 773-784. She, H.-C. (1998). Gender and grade level differences in Taiwan students' stereotypes of science and scientists. Research in Science and TechnologicalEducation, 16, 125-135. Song, J., & Kim, K.-S. (1999). How Korean students see scientists: The images of the scientist. InternationalJournal of Science Education, 21, 957-977. Steinke, J., & Long, M. (1995). A lab of her own? Portrayalsoffemale characterson children's educationalscience programs.(Report No. ED 384 937; CS 508 980). Paper resented at the Annual Meeting of the International Communication Association, 45th, Albuquerque, NM, May 25-29, 1995. (ERIC Document Reproduction Service, No. ED384937). Symington, D., & Spurling, H. (1990). The 'draw a scientist test': Interpreting the data. Research in Science and Technological Education, 8, 75-77.

COPYRIGHT INFORMATION

TITLE: The Draw a Scientist Test: A Different Population and a Somewhat Different Story SOURCE: Coll Stud J 40 no1 Mr 2006 WN: 0606003829017 The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited.

Copyright 1982-2006 The H.W. Wilson Company.

All rights reserved.