You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/234626169

The Draw a Scientist Test: A Different Population and a Somewhat Different


Story

Article  in  College student journal · January 2006

CITATIONS READS

30 464

3 authors, including:

Mark D. Thomas Tracy Henley


Albany State University Texas A&M University-Commerce
13 PUBLICATIONS   67 CITATIONS    79 PUBLICATIONS   1,221 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

William James View project

Prehistory of Psychology View project

All content following this page was uploaded by Mark D. Thomas on 24 February 2015.

The user has requested enhancement of the downloaded file.


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 cours-
es 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 sam-
ple (college students) 'vith previous studies that have that have
used the DAST with much younger (e.g., elementary-aged) stu-
dents. Results were strikingly similar, suggesting either that
gender stereotypes are widely persistent even among college sci-
ence majors, or that the DAST may not be a particularly sensitive
measure despite its wide use.

The Draw a Scientist Test (DAST) was children's (ages 5 to II years) drawings
first utilized by Chambers (1983) to exam- that were collected from 1966 to 1977.
ine stereotypic views of scientists among Chambers assessed the presence of'lab
school children. Chambers' initial study coats, eyeglasses, facial hair, symbols of
examined the strength and presence of research, symbols of knowledge, techno-
"modem sanitized,' and "older mythic," logical aspects, and captions that were
stereotypic images-of the scientist in 4,807 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. Correspon-
dence 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:
thenley@tamu-commerce.edu.

140
College Student DAST Images...!/141

of scientists. Chambers reported that 49% lain (1990) considered the effects of test
of this sample consisted of girls, but par- administration instructions, eventually
ticipants produced only 28 drawings of changing the DAST instructions to "Draw
female scientists (0.56%). All of these a Man or Woman Scientist:' Stated thus-
female scientists were drawn by female ly, 367 children (299 females and 68 males)
participants. between the ages of 11 and 16 years pro-
Subsequently, Fort and Varney (1989) vided drawings. Although boys in this
gathered 1,654 drawings of scientists sample almost exclusively drew males,
through a national contest for 2nd through 49% of the girls drew a female. Brosnan
12th grade school students. Fort and Var- (1999) modified the task by re-framing it
ney received 135 drawings of female as a, "draw-a-computer-user-test." For
scientists. They reported that, "the 8% Brosnan, whose sample consisted of 395
depicted by our respondents is close to children ages 5 to 11 years, males per-
reflecting reality," (p. 9) given their esti- formed similar to males who complete the
mate that women then made up roughly standard DAST. Interestingly, 70% of the
6% of the scientific and engineering work- females drew a female computer user,
force at that time. while only 4% of males drew a female
A careful analysis of DAST images computer user.
reported by Newton and Newton's (1992) Variations of the DAST have been uti-
survey of 1,143 primary school children lized in the U.S. and Canada (e.g., Parsons,
(ages 4-11) suggests children draw more 1997), Ireland (e.g., Maoldomhnaigh &
females at a younger age, but by the sixth Hunt, 1990), Finland (e.g., Raty, 1997),
school year, males were drawn by 83% of England (e.g., Brosnan, 1999), Korea (e.g.,
the participants. Following up, Newton and Song & Kim, 1999); and Taiwan (e.g., She,
Newton (1998) surveyed 1000 children 1998) with similar results. A recurrent find-
from reception (the UK equivalent of ing in the DAST literature is that scientists
preschool/kindergarten) to grade 6 and, are stereotyped as being male by girls and
"...concluded that there were few signifi- this may serve as a limiting factor in their
cant changes to primary pupils' self-efficacy toward becomifig scientists.
conceptions of science and scientist...' (p. Maoldomhnaigh and Mholain (1990)
1148). observed, "If boys are at all possessed of
Over time, the number of female sci- the image of the female scientist, it is too
entists drawn has slightly increased. This deeply hidden to be uncovered by the strat-
may be due to changes in social percep- agems used in this study," (p. 72) and
tions, or to refinements in the measure. For Chambers (1983) speculated, "...that, when
example, Matthews (1994) had 132 chil- asked to, 'draw a scientist,' even scientists
dren from years 7, 8, and 10 generate two themselves utilize the standard image" (p.
different drawings, out of this total, 66% 256).
of the images were male, while 34% were At times, the DAST literature has been
female. used to emphasize the need for teaching
Likewise, Maoldomhnaigh and Mho- reforms with respect to gender bias (e.g.,
142 / College Student Journal

Kahle, et. al, 1993; Newton & Newton, ticipated, or participated for class credit
1998). Perhaps not surprisingly, media has (age m = 21.02, sd= 3.14, range = 18-31).
also been seen as perpetuating this male The sample was 52.8 % female (n = 112)
scientist stereotype. Steinke and Long and 47.2 % male (n = 100). Participants
(1995) noted that twice as many male sci- represented 41 different majors with 62.9%
entists were represented as female (n = 132) having majors in scientific dis-
scientists on the four major television sci- ciplines, 34.3% (n = 72) having majors in
ence programs. Having noted that, at least nonscientific disciplines, and 2.8% (n = 6)
some dramatic perceptual changes on mat- were undeclared students. Participants
ters of gender and science have occurred were tested in extant courses or experi-
in recent decades. For example, in an arti- mental groups. Testing took place in
cle titled The College-Student Image of classrooms in either the psychology depart-
the Scientist(Beardslee & O'Dowd, 1961), ment (n = 131, 61.8%) or the computer
male college students were asked about science department (n = 81, 38.2%).
the attractiveness of various scientific dis-
ciplines, while female college students Materials
were asked about their preferences for hav- The study consisted of completing two
ing men of science for husbands. forms. The first form was a standard size
The purpose of this study is twofold. sheet of paper with the following instruc-
First, it is to examine DAST drawings pro- tions printed one-half inch from the top
duced by college students for the presence margin:
of traditionally interpreted stereotypes. You have up to 15 minutes to com-
That is, to attempt to replicate the basic plete this portion of your
findings of the DAST as gender stereo- participation. It would be reasonable
types are presumptively dynamic, and as to utilize the entire 15 minutes.
such, subject to change over time. Sec- Please do not talk to the other peo-
ondly, our goal is to explore the similarities ple who are participating in this
and differences between a college-aged study, as they are busy working as
sample with the previous results obtained well. In the space below, please
from school children. DRAW A SCIENTIST.
The second form was a brief question-
Method naire used to ascertain age, year in school,
major, participant's sex, sex of drawing,
Participants whether the participant drew a famous sci-
The participants were 212 undergrad- entist (and his/her identity), identity of the
uate students enrolled in psychology or participant's favorite scientist (and reason
computer science service courses at Mis- for being such), type of scientist drawn,
sissippi State University. The sample number of instructors for the current
consisted of 47 freshmen (22.2%), 48 semester in the department tested, current
sophomores (22.6%), 62 juniors (29.2%), number of male and female instructors in
and 55 seniors (26%), who voluntarily par- the department tested, and sex of the last
College Student DAST Images...!/143

person with whom the participant had a table, a double helix, or other figures and
substanative conversation. tables). Symbols of technology were any
products of science (e.g., computers, tele-
Procedures phones). Captions were scored differently
All surveys were administered by one than the traditional method, as they were
female experimenter or one male experi- scored just for presence or absence, and
menter. During class-time administrations not for content.
(n = 3), the instructors were not present in The operational criteria for technical
two classes, and a female instructor was aspects of the drawings were defined a pri-
present in one. The instructors' sexes were ori by consensus between the researchers.
1 male and 2 female. Participants were told Drawings generated by 17 graduate stu-
the study consisted of 2 tasks and their dents (not used in this study) were jointly
instructions were on the forms. Partici- scored by the raters in order for training.
pants were also told they would get no All drawings were independently rated by
clarification about the instructions and no 2 clinical psychology graduate students.
participant asked for clarification. After
the 15 minutes for drawing elapsed, par- Results
ticipants were asked to cease drawing and Sex of Scientists
turn their drawings face down. Participants The reported sexes of the drawings were
then received the questionnaire. After com- 82% male (n = 173), 12.3% female (n =
pleting the questionnaire, participants were 26), and 5.7% nonspecified (n = 12). The
debriefed. One participant withdrew his drawings indicated as nonspecified includ-
drawing after debriefing, but allowed his ed stick people, a drawing with members
questionnaire to be used. of both sexes, and a drawing of a cartoon
Chambers' (1983) original scoring sys- dog. Female participants generated 2 non-
tem was used for all but one of the specified (1.8%), 92 male (82.1%), and 18
stereotypical science indicators. Lab coats female (16.1%) drawings, while male par-
were scored for presence (without regard ticipants drew 10 nonspecified (10.1%),
to color), eyeglasses were defined as two 80 male (81.8%), and 8 female (8.1%)
lenses (with or without arms), and facial drawings. An ANOVA found a significant
hair was any facial hair growth (e.g., mous- difference between sex of drawing and par-
taches, beards, jaw length or longer ticipants' sex (F = 8.123, p < .005).
sideburns, 5 o'clock shadow). Symbols of
research were scientific instruments and Types of Scientists
laboratory equipment (e.g., microscopes, Participants reported the types of sci-
pencils in pockets, test tubes, Bunsen burn- entist drawn as chemists (55.2%, n = 117),
ers, clip boards). Symbols of knowledge biologists (13.7%, n = 29), psychologists
were items that represent having knowl- (4.7%, n = 10), physicists (4.2%, n = 9),
edge (e.g., books, filing cabinets, computer scientists (2.8%, n = 6), non-
blackboards, posters with the periodic specific (8.5%, n = 18), an astronomer (n
-------------------

144 / College Student Journal

=1, .5%), a natural scientist (n = 1, .5%), class year and lab coats (F = 3.31, p <
and 11 participants failed to report the type .021). All other indicators were not sig-
of scientist drawn (5.2%). The 10 remain- nificantly different between class years.
ing cases (4.7%) are believed to be While 41 drawings (19.3%) had smoke
sarcastic, or noncompliant, responses and coming from apparatuses, in 2 drawings
include 3 mad/deranged scientists (1.4%), the scientists (1%) were smoking tobacco
3 nerd/geek scientists (1.4%), and 4 other (or possibly marijuana), and 1 (0.5%) had
comic responses (1.9%) with one each, smoke coming from the scientist's ears.
"Elvis," "pothead" "dumb," and an "elec- Computers were present in 10 drawings
tric" scientist (that is, one who was being (4.7%). A name tag was present on 29
electrocuted). (13.7%) scientists, with 8 being illegible
and the others representing either well-
Analysis of Drawings known fictional or nonfictional scientists,
Interobserver agreement was calculat- including 5 local computer science and
ed on 17 items for 211 drawings. The psychology faculty members, Kevorkian
raters' separate data were compared to and Einstein were among others named.
detect rater differences. Initial agreement Most DAST scientists had a tradition-
was 3,381 out of 3,587 chances to agree al hairstyle (n = 117, 55.2%). However, 44
(94.3%). The 206 disagreements were had wild hair (20.8%), 38 were bald
independently rated a second time by the (17.9%), 4 (1.9%) had nontraditional hair
same raters and agreement was 3,533 (e.g., males with long hair, or bald females
(98.5%). The remaining 54 disagreements - excluding stick figures), and 8 (4.2%)
were resolved by conference and the major- were undeterminable due to being covered
ity of those involved lab coats, symbols of by hats/scarfs, or being headless. Earrings
knowledge, facial hair, and captions. were visible on 3 males scientists (1.4%)
Lab coats occurred in 54.5% (n = 115) and no female scientists. The scientist was
of the drawings. Scientists wore eyeglass- represented by a simplistic stick person in
es in 145 drawings (68.4%) and had facial 12 drawings (5.7%), or an elaborate and
hair in 44 drawings (20.8%). Evidence of more "artistic" stick person in 9 drawings
research was present in 148 drawings (4.2%).
(69.8%). Symbols of knowledge were pre- FavoriteScientist
sent in 27 drawings (12.7%), and symbols The favorite scientist reported was
of technology were present in 21 drawings Albert Einstein (n = 105,49.5%). Given the
(9.9%). Captions were present in 71 principle result, it was curious that Marie
(33.5%) drawings. Curie was the second favorite (n = 7, 3.3%).
The mean number of 'stereotypical' The third favorites were instructors with
indicators present is 2.92 per drawing, whom the participant had classes (n = 6,
which compares to Chambers (1983) 2.8%), which included 2 current psychol-
reported 3.05 for 4th grade children and ogy department faculty members, 2
3.26 for 5th grade children. An ANOVA computer science department faculty mem-
detected a significant difference between bers (n = 3), and a former middle school
College Student DAST Images.../145

science teacher. Next were Isaac Newton or a genius (n = 20, 9.4%); the scientist's
and George Washington Carver (n = 5, hair or appearance (n = 11, 5.2%); the sci-
2.4% each). StephenHawkings, Benjamin entist was the participant's instructor (n =
Franklin, and Sigmund Freud were listed 2, .9%); and 8 participants (3.8%) cited
(n = 4, 1.9% each), and Thomas Edison other reasons that do not readily fit a clas-
and Dr. Frankenstein/mad scientist were sification scheme.
rated sixth most popular (n = 3, 1.4% for
each). Emmet Brown (from the Back to Multiple Regression
the Future Movies), James Watson and Using Mahalanobis distance to detect
Francis Crick (reported together), Bill Nye outliers (Mertler & Vannatta, 2002), three
(from the television show Bill.Nye the Sci- cases were identified as extreme. These
ence Guy), Alexander Graham Bell, cases were inspected and determined to be
Galileo Galilei, Richard Feynman, and older students, or students with a heavy
Johannes Kepler were each listed twice class load whose instructors were all the
(0.9% each). same sex. After due deliberation all of
Twenty-three different scientists were these cases were included in the analysis.
mentioned once (0.5% each) and includ-. An evaluation of linearity revealed no con-
ed, among others, Robert Oppenheimer, cerns. However, drawings reported as
Charles Darwin, Erik Erikson, Michael having a nonspecific sex were excluded
Faraday, Carl Sagan, Ivan Pavlov, Karen from analysis, because most people are
Horney, Dmitri Mendeleev, and Jean male or female and the nonspecific cases
Piaget. Nonexistent scientists, such as The may be instances of reactivity.
Professor (from Gilligan'sIsland), Dex- A standard multiple regression analy-
ter (from the cartoon, Dexter's sis was conducted to evaluate the prediction
Laboratory), and Brain (from the cartoon of DAST image sex from 10 variables (age,
Pinky and the Brain), along with the par- year, participant sex, favorite scientist,
ticipant (self as favorite), a participant's female instructors, male instructors, con-
aunt, a participant's uncle, and a partici- versation, experimenter sex, science, and
pant's father were also mentioned once. Einstein). Regression results indicate that
Twenty-nine participants did not report the overall model significantly predicts
a favorite scientist and an additional 13 DAST image sex, R1 = .133, R2adj =.075,
failed to report a reason for their selection. F (10, 151) = 2.31, p < .015. The model
The 170 reported reasons for being a par- accounts for 13.3% of the variance in sex
ticipant's favorite scientist can be classified of DAST images. A summary of regres-
into 7 categories: the scientist's theories sion coefficients is presented in Table 1
or intellectual contributions (n = 66, and indicates that only three (participant
31.1%); being the only scientist the par- sex, favorite scientist, and experimenter
ticipant could think of (n = 41, 19.3%); the sex) of the ten variables significantly con-
scientist was famous or funny (n = 22, tributed to the model. The 95% confidence
10.4%); the scientist was smart, brilliant, 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 I i
AZe -.0173 -.115 .325
YeUr -.0113 -.320 3749 -.009 -.026
Sex .10so 2.051 .042 .202 .165
Favofite 2.1700 2.12S .035 .207 .171
F1 .0500 .109 1.319 .139 .099 .107
MI .02S0 1.106 .271 .072 .090
Converation .0490 .077
~.101* .995 .321 .059 .0s1
Epeimenter -.1250 -2.095 -.132
Science .0890 .144 1.579 .022 .127
P17At
EinsWen 2047 WY,-)
11t71? MWJA W."
-------------

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,
Conve•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 represent a perceived public stereotype and
administrator -.243 to -.007, and do not not a personal belief or perception. It
contain the value zero. appears that the drawings generated by this
sample support that idea.
Discussion Indeed, the DAST is a tool initially used
One could anticipate more complex rep- to reveal a child's attitudes and beliefs
resentations in drawings as an individual about science, and the DAST has tradi-
develops a more complex knowledge of tionally been interpreted the same way for
the scientist, which may account for the children in any grade. The drawings gen-
increase in stereotypic indicators from first erated by these undergraduate students are
through fifth grade school children report- similar to those of children that appear in
ed by Chambers (1983). However, this the DAST literature, but at some level this
sample of college undergraduate students' sample cannot have the same beliefs and
drawings exhibited similar stereotypic attitudes that are inferred by traditional
characteristics to those of Chambers' fourth DAST interpretation, because 62.9% are
and fifth grade children. Symington and majors committed to a college science
Spurling (1990) challenged the tradition- degree. Thus, it appears that some changes
al interpretation of the DAST and in attitudes and beliefs that occur between
speculated that children's drawings may kindergarten and college, may not be well
College Student DAST Images...!/147

captured by the DAST. 72% of the students enrolled in doctoral


Specifically, we submit that many fac- programs were women. It seems reason-
tors 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 psy-
sion 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 psy-
istrator, 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.
jects their views. With college students, Although we will conclude by observ-
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 fea-
illustrate, that was not a problem with this sible explanation for these women's
sample. apparent dissonance, it should be noted
that, stick people were drawn by about
Conclusions 10% of the participants and it is doubtful
In theory, this sample with a mean age that even 10% of undergraduate students
of 21.02 years ranging from 18 to 31 years believe scientists are stick figures. In short,
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
remarkably similar to those of Chambers' References
Auguste, R., M., Wicherski, M., & Kohout, J. L.
study. From a perspective across time, (1999). APA Research Office.
exposure to years of science education has http://research.apa.org/mes96contents.html#c
seemingly had no effect on these partici- ontents
pants. Or, as a direct comparison between Beardslee, D. C., & O'Dowd, D. D. (1961). The
the two samples suggests, college students college student image of scientists. Science,
think no differently about science and gen- 133, 997-1001.
der than elementary school children. Brosnan, M. J. (1999). A new methodology, an old
The APA Research Office (Auguste, story? Gender differences in the "draw-a-com-
puter-user" test. European Journal of
Wicherski, & Kohout, 1999) reported that Psychology and Education, 14, 375-385.
72% of master's degree recipients in psy-
chology 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. Newton, L. D., & Newton, D. P. (1998). Primary
chilren's conceptions of science and the scien-
Fort, D. C., & Varney, H. L. (1989). How students tist: Is the impact on the National Curriculum
see the scientist: Mostly male, mostly white, breaking down the stereotype? International
and mostly benevolent. Sci&nce and Children, Journalof Science Education, 20, 1137-1149.
26, 8-13.
Parsons, E. C. (1997). Black high school females'
Kahle, J. B., Parker, L. H., Rennie, L. J., & Riley, images of the scientist: Expression of culture.
D. (1993). Gender differences in science edu- Journal of Research in Science Teaching, 7,
cation: Building a model. Educational 745-768.
Psychologist, 28, 379-404.
Raty, H., & Snellman, L. (1997). Children's
Maoldomhnaigh, M. C., & Hunt, A. (1988). Some images of an intelligent person. Journal of
factors affecting the image of the scientist Social Behavior and Personality,12, 773-784.
drawn by older primary school pupils.
Research in Science and TechnologicalEdu- She, H.-C. (1998). Gender and grade level differ-
cation, 6, 159-166. ences in Taiwan students' stereotypes of
science and scientists. Research in Science and
Maoldomhnaigh, M. C., & Mholain (1990). The TechnologicalEducation, 16, 125-135.
perceived expectation of administrator as a fac-
tor affecting the sex of scientists drawn by Song, J., & Kim, K.-S. (1999). How Korean stu-
early adolescent girls. Research in Science and dents see scientists: The images of the scientist.
Technological Education, 8, 69-75. InternationalJournal of Science Education,
21, 957-977.
Matthews, B. (1994). What does a chemist look
like? Education in Chemistry, September, Steinke, J., & Long, M. (1995). A lab of her own?
127-129 Portrayalsoffemale characterson children's
educationalscience programs.(Report No. ED
Mertler, C. A., &Vannatta, R. A. (2002). Advanced 384 937; CS 508 980). Paper resented at the
multivariate statistical methods: Practical Annual Meeting of the International Commu-
applicationand interpretation.(2nd ed.). Los
nication Association, 45th, Albuquerque, NM,
Angeles, CA: Pyrczak Publishing.
May 25-29, 1995. (ERIC Document Repro-
Newton, D. P., & Newton, L. D. (1992). Young duction Service, No. ED384937).
children's perception of science and the scien- Symington, D., & Spurling, H. (1990). The 'draw
tist. International Journal of Science
a scientist test': Interpreting the data. Research
Education, 14, 331-341. 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.

View publication stats

You might also like