Creative Education
2013. Vol.4, No.5, 315-321
Published Online May 2013 in SciRes (
Copyright © 2013 SciRe s . 315
Secondary Students’ Views about Creativity in the Work
of Engineers and Artists: A Latent Class Analysis
Danielle B. Harlo w, Karen Nylund- Gi bs on , Ashl ey Ivel and, Lauren Taylor
Department of Edu c at io n, Gevirtz Graduate Schoo l of Education, University of California Santa Barbara,
Santa Barbara, USA
Received March 6th, 2013; revised April 10th, 2 0 1 3 ; ac c e p t e d A pril 22nd, 2013
Copyright © 2013 Danielle B. Harlow et al. This is an open access article distributed under the Creative Com-
mons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro-
vided the original work is p roperly cited.
To close the current “innovation gap”, schools must help students develop creative thinking skills and
understanding of technical careers as requiring creativity. A first step is understanding students’ current
perceptions of the work of such careers. The goal of this study was to identify patterns in how secondary
school students viewed the work of a technical career, engineering, in comparison to art, a career more
traditionally associated with creativity. The sample consisted of 104 students (ages 13 - 15) entering their
first year of a program that intentionally integrates engineering, science, and art instruction. The students
in our study completed two assessments during their first week of school. On one side of a paper they
drew an engineer and answered questions about what engineers do. On the other side of the paper, they
drew an artist and answered questions about what artists do. These drawings and associated writing were
coded. We then used the statistical method of latent class analysis to model the responses. We identified
three patterns in the ways that students perceived the work of engineers and two patterns in the ways that
students perceived the work of artists.
Keywords: Engineering; Art; Latent Class Analysis; Secondary Education
In the United States, engineering has been largely absent
from pre-college education. This means that only those students
who choose to pursue engineering at the undergraduate level
have the opportunity to develop sophisticated understandings of
engineering. Fortunately, this is changing. The recently drafted,
A Framework for K-12 Science Education: Practices, Cross-
cutting Concepts and Core Ideas (NRC, 2012), considers sci-
ence and engineering practices as integral to science instruction.
This means that soon science teachers will be teaching engi-
neering practices and habits of mind to even our youngest stu-
Integrating engineering into K-12 programs has many poten-
tial benefits: “improved learning and achievement in science
and mathematics; increased awareness of engineering and the
work of engineers; understanding of and the ability to engage in
engineering design; interest in pursuing engineering as a career;
and increased technological literacy” (Katehi, Pearson, & Feder,
2009: pp. 49-50). Learning engineering will not only benefit
those who eventually pursue engineering or technology careers,
it will benefit all students. Indeed, engineering “habits of mind”
include systems thinking, creativity, optimism, collaboration,
communication, and attention to ethical considerations, the very
qualities considered imperative for working in the 21st century
(e.g., Trilli ng & Fadel, 2009).
Recently, business leaders have identified an “innovation
gap” between the qualities desired for employees and the actual
qualifications of applicants (e.g., Casner-Lotto & Barrington,
2004). The workforce is not prepared to do innovative work
and meet the demands of the new business and technology-
driven world. One contributing factor to this innovation gap
may be that instruction emphasizing creativity is lacking in
most engineering programs, both at the K-12 and post-secon-
dary levels.
Innovation is related to creativity through the process of en-
gineering design, the process of developing a new product or
solution to a problem that meets the goals of a client while
working within constraints. Researchers of engineering agree
on the stages of the design process, with slight variations in the
naming of the stages and acknowledgement that, in practice,
engineers may not pass through the stages in a linear fashion.
These stages include establishing a need, analyzing the task,
conceptual design, embodiment design, detail design, and im-
plementation (Howard, Culley, & Dekoninck, 2008). Carefully
moving through this process results in a solution that is appro-
priate for the task, but does not necessarily result in a solution
that is novel or innovative. To be innovative, one must engage
in the creative process of generating, evaluating and analyzing
ideas in one or more stages of the engineering design process
(Howard, Culley, & Dekoninck, 2008). In fact, the National
Academy of Engineering states that “creativity (invention, in-
novation, thinking outside the box, art) is an indispensable
quality for engineering, and given the growing scope of the
challenges ahead and the complexity and diversity of the tech-
nologies of the 21st century, creativity will grow in impor-
tance” (NAE, 2004: p. 55). This is in sharp contrast to a wide-
spread public belief that creativity is not part of engineering
(Kazerounian & Foley, 2007; Schmidt, 2011).
In efforts to help students develop creative skills in concur-
rence with more technical skills, some educators are calling to
integrate arts education into the more technical disciplines of
science, technology, engineering, and mathematics (STEM).
This movement has been termed STEAM education which
stands for science, technology, engineering, art, and mathemat-
ics. STEAM has been described as “science and technology,
interpreted through engineering and the arts, all based in a lan-
guage of mathematics” (Yakman, 2008). STEAM education,
which advocates for integrating STEM with an all-inclusive
definition of the arts, has been gathering momentum interna-
tionally. Two recent studies in Korea used a STEAM model to
integrate the sciences and arts at the elementary and early high
school levels (Yakman, 2010; Kwon, Nam, & Lee, 2011). Re-
lated movements initiated by art advocates include STEM-A
( and stem-arts (
Finally, the art of science learning
(, funded by the National
Science Foundation, explores how arts-based learning can im-
prove science learning at al l a ges.
The work of engineers involves designing a product that
meets a client’s needs and includes aesthetic considerations
along with financial, safety, and other constraints. In programs
that integrate STEM content with art, students engage in the
practices associated with engineering as well as art. Depending
on the curricular design, these may be simultaneous experi-
ences or separated in time. One would expect that participating
in the work of both engineering and art would help students
develop more sophisticated understandings about the work of
these professionals.
Since the early 1980’s, researchers have investigated stu-
dents’ ideas about the work of professionals by asking students
to draw. The initial Draw a Scientist Test (Chambers, 1983)
elicited several consistent characteristics that appeared in young
children’s drawings of scientists. Later iterations of this as-
sessment tool looked at the impact of media, gender, and older
students’ perceptions (for a review of this literature see Finson,
2002). The Draw a Scientist Test (DAST) has since been
adopted for a variety of professions. Knight and Cunningham
(2004) adopted the DAST for use with Engineering. In their
version, students responded to the following five questions, “In
your own words what is engineering?” “What does an engineer
do?” “Draw a picture of an engineer at work” “Do you know
any engineers?” “If yes, then who are they?” (p. 2529). They
administered the Draw an Engineer Test (DAET) to 384 stu-
dents in grades 3 - 12 and found that the most common activi-
ties associated with engineering were building (30%) and fixing
(28%) followed by creating (17%) and designing (12%). Later,
Capobianco, et al. (2011) administered the Draw an Engineer
Test (DAET) to 400 elementary school children. The children
in this study depicted engineers as mechanics, laborers, techni-
cians, and designers.
We used a modified version of this assessment to understand
students’ ideas about the work of both engineers and artists. We
hypothesized that students would consider artists to be creative
and engineers to be logical. We tested th is hypothesi s by having
freshmen in a high school engineering academy draw engineers
and artists and answer questions about the work of engineers
and artists. We then coded these data and modeled the data
using the statistical procedure of latent class analysis (LCA) to
better understand the patterns in the d at a.
Our study was guided by three research questions: 1) Are
there patterns in how secondary students entering an engineer-
ing academy think about the work of engineers? 2) Are there
patterns in how secondary students entering an engineering
academy think about the work of artists? 3) How are these pat-
terns related t o each other?
Study Design
This study included a qualitative stage and a quantitative
stage. We collected qualitative data consisting of student draw-
ings in response to specific prompts (described below). We then
coded these data through an emergent coding process. The
codes were then analyzed using latent class analysis. Below we
describe the research context, the methods of data collection
and the qualitative and quantitative data analysis process.
Research Context
This study took place at an Engineering Academy within a
public high school in California, USA. In the year prior to this
study, the school served approximately 2300 students of which
48% were Caucasian, 37% Hispanic, 9% Asian, and 2% were
African American. Nearly a quarter (23%) of the students were
disadvantaged socioeconomically. The school housed a nation-
ally recognized competitive engineering academy that admitted
approximately 100 freshmen each year. The Engineering
Academy enrollment in 2012-2013 was 104 students roughly
evenly divided by gender (53% female). The population was
nearly three quarters (74%) White, 19.6% Asian, 1.8% African
American, 1.8% Filipino and 1.8% American Indian. Sixteen
percent of the students reported that they were of Hispanic or
Latino ethnicity. These 104 freshmen made up our study popu-
In the Engineering Academy, physics, computer program-
ming, engineering, and art are intentionally integrated. Students
in this program spend one hour a day in the engineering acad-
emy doing course work dedicated to engineering, physics, art,
and computer science. The rest of their course work is com-
pleted in the regular high school. Curriculum is developed
around projects that require students to integrate knowledge
from all of these areas. For example, one project that students
complete is a light sculpture. The students design the sculpture
using computer aided design (CAD), use lathes and other
equipment to machine parts of the sculpture, and program the
lights to go on and off in an aesthetically pleasing way. This
requires knowledge of machining, circuits, art, and computer
programming. These smaller projects lead into a capstone pro-
ject in which the seniors form a robotics team and compete in a
national robotics competition.
Data Collection: Draw an Engineer and Draw an
During their first week of school, entering freshmen com-
pleted a modified Draw an Engineer Test. On one side of a
piece of paper, each student answered the questions, “In your
own words, what is engineering?” and “What does an engineer
do?” They also drew a picture of an engineer at work, and
wrote a brief description of what was going on in the drawing.
On the other side of the paper, the students answered similar
questions related to art, “In your own words, what is art?”
“What does an artist do?” They drew a picture of an artist at
Copyright © 2013 SciRe s .
work, and wrote a brief description of what was going on in
their drawing. Two responses to the engineering task and two
responses to the artist task were too light to be legible and could
not be used in the analysis. Thus our resulting data set consisted
of 102 responses to both the “Draw an Engineer Test” and to
the “Draw an Artist Test.”
Qualitative Analysis: Coding Drawings
The engineering drawings were coded with binary codes
(lack of presence/presence). The initial coding scheme was
based on that developed by Capobianco, and others (2011). It
was modified during initial coding as additional codes emerged
and codes were combined and collapsed. Since our population
was older than the population in Capobianco et al. (2011) and,
further, were entering a selective engineering academy we ex-
pected that our population would have more sophisticated un-
derstandings of the work of engineering. We had no initial
model for coding the artist task and thus, those codes were
Our aim was to construct a coding scheme that could be used
for both the engineering and artist portion of the activity. Our
resulting coding scheme consisted of the following types of
codes: Actions (what was done by engineers/artists), the output
(object), qualities of the output, purposes of the work, require-
ments (inputs) for the work, and details about who did the work.
Our initial scheme also coded particular types of clothing (e.g.,
lab coats on the engineers, berets on the artist) and tools (e.g.,
paintbrushes and screwdrivers) but these two categories of
codes were not included in the statistical analysis. Finally, we
coded the gender of the drawn engineer and or artist. This was
determined primarily by pronouns used in the students’ de-
scriptions of the picture (e.g., “The engineer is inspecting her
work,” “He is working on a robot”). Although many drawings
looked like they were intended to depict males or females (e.g.,
short hair, jeans for men, hair bows and skirts for females),
unless a gender-specific pronoun was used, they were assumed
to be non-specific to gender. As a result, we likely under-coded
the number of “male” drawings for both artists and engineers.
Our coding process included coding a subset of the data by
two independent coders. Coding was then compared and dis-
cussed. This led to a refinement of the coding scheme including
collapsing codes into larger categories. This process of coding,
comparing, and refining was continued until an interrater reli-
ability over 80% was attained. All data were then recoded with
the new coding scheme. Differences in codes during the final
round were discussed until consensus.
Quantitati ve An al ysi s: Latent Class Analysis
The second stage of the analysis involved using the statistical
method of LCA (Dayton, 1998; Magidson & Vermunt, 2004) as
a way of exploring whether there were patterns in students’
drawings. LCA is an exploratory multivariate analysis, similar
to cluster analysis, that can use the codes from students’ draw-
ings to identify latent classes of students. These classes are
based on binary codes meaning that the code was either present
or not (see Harlow, Swanson, Nylund-Gibson, & Truxler, 2011
for another example of using LCA to interpret students’ draw-
ings of science ideas). Different than the more commonly used
factor analysis, LCA groups individuals whereas factor analysis
groups items (or codes). In LCA, there are no assumptions
made about which codes would group together or how many
groups (or classes) students would be sorted into. LCA uses the
term “class” to refer to the grouping of individuals, but for this
paper we will use the term “group” to avoid confusion with the
term classes to refer to classrooms or age level groupings of
students. LCA models are fit in a series of steps. First, a one-
class model is fit and then the numbers of groups (classes) is
increased. Each new model (with an increase in the number of
groups) is compared to the previous model. The model with the
greater number of groups is selected only if increasing the
number of groups results in conceptually meaningful groupings
and provides a good statistical fit. This was done for engineer-
ing and art independently. In addition, fit indices were used to
help determine the ideal number of conceptually meaningful
groups (Nylund, Asparouhov, & Muthén 2007). The models
were run with Mplus version 7.0 (Muthèn & Muthèn, 1998-
We used a subset of the codes described above in the LCA
models. Specifically, we selected those codes that showed dif-
ferences among groups identified by LCA within each profes-
sion. This means that if a particular code was identified in
nearly all or very few responses, it was not included because
that code would not differentiate the groups. For example, the
code “communicate” was only found on one engineering re-
sponse so it was unlikely to show how the groups differed and
was not included for the engineering response. However, it was
found on 60 of the art responses and therefore was likely to
distinguish the groups and was included for the LCA analysis
of the artist responses.
The following codes were used for the latent class analysis.
Some codes were used for both the engineering and art LCA
(indicated with E/A). Others were used only for the engineering
(E) or the art (A) LCA.
1) Draw Male (E/A). Drawing was of a male and student
used male pronouns (he, him, his) when writing about what the
engineer or artist was doing.
2) Draw Female (E/A). Drawing was of a female and student
used female pronouns (she, her) when writing about what the
engineer or artist was doing.
What an artist or engineer does:
3) Create (E/A). The engineer or artist was thinking crea-
tively or being creative.
4) Imagine (E/A). The engineer or artist imagined or visual-
ized an idea.
5) Design (E/A). The engineer or artist designed something.
6) Make (E/A). The engineer or artist made something.
7) Skills (E). Verbs indicating a particular skill that could be
learned. Examples from engineering include computer pro-
gramming, drawing, and drafting. Examples from art include
draw, paint, and sing.
8) Do/Work (E). Vague statements such as, “He does work”
or “She does art”.
9) Logic (E). Going through a particular set of steps or solv-
ing a problem that has an assumed right answer. Words include
discover, solve, test, think.
10) Tinkering (E). The artist or engineer was manipulating an
existing object to improve it in some way
11) Maintain (E). The artist or engineer was manipulating an
existing object to repair or maintain.
12) Communicate (A). Art or engineering included commu-
nicating or expressing ideas.
Output/Product of work:
Copyright © 2013 SciRe s . 317
13) For others (E/A).The end product of the work of the en-
gineer or artist was for others. Included statements like, “for
others’ enjoyment” and “to help others”.
Requirements for work/Inputs:
14) Thought (E). Students indicated that thought was re-
quired or that the engineer or artist used his or her own
15) Ideas (E). Students indicated that ideas were required.
16) Knowledge (E). Students indicated that knowledge was
required. While this is similar to both thought and ideas, we
considered it separate because knowledge implies an existing
body of knowledge.
17) Emotion (A). Students indicated that emotions were re-
Qualitative Results
The number of responses coded with each code is listed in
Table 1. The first three codes (create, design, and imagine) are
codes we associated with traditional definitions of creativity. A
little over half of the students (53 students) explicitly associated
creativity (or the action of creating something) with engineering.
40 students stated that designing was part of engineers’ work
and 20 students claimed that imagining was part of engineers’
work. This is compared to 62 students stating that artists create
things or use creativity, 32 stating that artists imagine and only
7 reporting that artists design. Both codes of create and imagine
were more similar across the two disciplines than we antici-
Codes that were found much more often among the engi-
neering responses as compared to the artist responses include
work, construct, tinker, maintain, logic, knowledge, for others.
In contrast, codes that were more common among responses
Table 1.
Number of respo nses including each code.
Code Engineering Art
Create 53 62
Design 40 7
Imagine 20 32
Make 26 17
Do/work 17 6
Construct 71 4
Tinker 12 2
Maintain 22 0
Logic 33 5
Skills 40 89
Knowledge 37 1
Thought 15 25
Ideas 16 25
Communicate 0 60
Emotion 0 41
For others 41 19
Drew Male 41 27
Drew Female 10 16
about artists include skills, thought, ideas, communication and
emotion. Further, more of the responses about artists included
females while the responses about engineering included more
The qualitative analysis alone provides interesting insight to
how freshmen students in an engineering academy perceive the
work of engineers and artists. Over half the students see each
career as involving creativity. The students do indicate major
differences between the two careers. Engineering is perceived
to be more logical and requiring knowledge. In contrast, artists
are depicted as more emotional and requiring skills such as the
ability to paint, draw, or play instruments. A few of the students
noted connections between the work of artists and engineers.
For example, one student (student 64) wrote on the draw an
engineer test, “Really: an engineer is an artist” and conversely,
on the draw an artist test, wrote, “Really: an artist is an engi-
neer.” Other students made explicit connections between the
disciplines by calling engineering “an art.” What the qualitative
data does not allow us to immediately perceive is the patterns
within the ways of thinking about each career. For that we turn
to the statistical method of modeling the data using Latent
Class Analysis.
Deciding the Number of Groups
Table 2 presents the model fit information for the LCA mod-
els estimated for both art and engineering. The indices used for
the analysis are the commonly accepted ones used in LCA ap-
plications. With both the Bayesian information criteria (BIC)
and the Adjusted BIC (ABIC), the model with the lowest value
indicates the better fitting model. Also, two other fit statistics
used were the Lo-Mendell-Rubin likelihood ratio test (LMR
LRT, referred to as LMR) and the bootstrap likelihood ratio test
(BLRT). These provided a p-value that was used to compare
neighboring group models (i.e., 2-group vs. a 3-group). For fur-
ther information on LCA and the methods used to determine the
appropriate number of group (classes), see Nylund et al. (2007).
Table 2 presents fit information for both the engineering and
art LCA models considered. For the engineering drawings, we
fit 1 through 4 groups, and for the art we fit 1 through 3 groups.
When examining the results of the engineering drawings, both
the AIC (1645.65) and ABIC (1620.11) show significant de-
creases in the decreasing pattern when comparing 3 groups to 4,
suggesting that the addition of the 4th group does not provide a
significant increase in fit to the 3 group. This is further sup-
ported by the BLRT, which shows a non-significant p-value for
the 4-group solution, indicating that the 3-group model is pre-
ferred. The BIC shows an irregular pattern from those seen in
most applications, where it increases in values after the adding
more groups. The entropy for the engineering LCA is high
at .88, indicating good classification into the group. In sum the
majority of fit statistic for the engineering LCA point to a
3-group solution.
The model fit for the LCA of the art drawings consistently
pointed toward a 2-group model based on the AIC (980.76),
BIC (1030.82), and ABIC (970.80). This solution is further
justified by the LRM and BLRT that both show a non-signifi-
cant p-value for the 3-group solution, indicating that the 2-
groupsolution is the best fit for the art drawings. The entropy
for the art LCA is also reasonably high at .80. In sum, these
results indicate that a 2-class solution fits best for the art draw-
Copyright © 2013 SciRe s .
Copyright © 2013 SciRe s . 319
Table 2.
Fit indices for the engineering and art drawings model solutions.
Class Log Likelihood AIC BIC ABIC LMR/p-value BLRT/p-value Entropy
1 1883.01 1699.35 1738.58 1691.20 n/a n/a n/a
2 1611.21 1669.40 1750.47 1652.56 61.13/.14 -834.68/.01 0.82
3 1699.30 1645.65 1768.56 1620.11 55.01/.11 -803.70/.01 0.88
4 1727.98 1644.62 1809.37 1610.39 32.59/.35 -775.83/.18 0.88
1 280.35 1006.68 1030.39 1001.97 n/a n/a n/a
2 290.59 980.76 1030.82 970.80 44.95/.01 -494.34/.01 0.80
3 223.98 988.91 1065.31 973.71 11.61/.14 -471.38/1.00 0.87
LCA Model Results for Engineering
We use three groups to model the students’ responses to the
draw an engineer test. A clear difference can be seen between
Engineering Group 1 (57.6% of students) and Engineering
Group 3 (11.8% of students). Students sorted into Engineering
Group 1 (represented by black line in Figure 1) depict engi-
neers as professionals who have gained knowledge and skills
and use logical thinking to do their work. Of the three groups,
students sorted into Engineering Group 1 also have the lowest
probability (32%) of indicating that an engineer is creative.
In contrast, students sorted into Engineering Group 3 (dashed
line on Figure 1) depict engineers as creative (100%) and
imaginative (100%) individuals who use their own thoughts
and ideas to do engineering work. This group was also the most
likely to draw a female engineer. Engineering Group 2 (30.6%
of students, grey line on Figure 1) was more likely than stu-
dents sorted into the other two groups to indicate that engineers
maintained things and were the most likely of all the groups to
portray a male engineer.
Figure 1.
Three-class LCA model for engineering.
is creating something/solving problems in an unique way. In
other words, doing something using creative thinking and
thinking outside the box.” This student also wrote, “An engi-
neer does somet hing that can help/inspire others and does so me-
thing unique and special.” The drawing depicts plans for a chil-
dren’s hospital on the desk of an engineer. We can conclude
that it is a female engineer’s desk, although no individual is pic-
tured, through the description of her image, “she is drawing”.
Once an LCA is fit, we can explore which of the latent
groups each student belongs to. The model provides a probabil-
ity of being in that group, called a posterior class probability,
which allows us to feel confident in the assignment. In our ap-
plication, each model had high entropy, implying that students
are assigned with high precision, giving us confidence in using
these posterior probabilities to assign students to groups. Based
on their group assignment, we selected one sample student for
each of the three groups to show their drawings (see Figure 2).
LCA Model for Art
Students were sorted into two groups according to their re-
sponses to the draw an artist test. A little over half (53.2%) of
students were sorted into Artist Group 1 (black line on Figure
3). This line peaks at the code communicate (100% probability)
which also included phrases such as “express themselves.”
Students sorted into Artist Group 1 were also highly likely to
indicate that emotion (57%) was important. In contrast, students
sorted into Artist Group 2 (46.8% of students) had a low prob-
ability of indicating the importance of communication and
emotion and instead were most likely to mention creating
something (74%).
Student 77 was selected to represent Group 1. This student
had a high probability, 91%, of belonging to Engineering
Group 1, and lower probability, 9%, of being sorted into Engi-
neering Group 2. This student wrote that an engineer is “a per-
son who builds designs and maintains buildings, transportation
and motors.” Also, “Engineers help keep thing running and
working. They also build all sorts of stuff.” The drawing de-
picted a male engineer (as indicated by the description of the
picture) “welding something” and “designing something.”
Student 21 was sorted into Engineering Group 2 with 100%
probability and represents the group well. This student wrote
that engineering is “building and designing tools and stuff” and
drew an engineer “programming and figuring out how to build
an infrared radar.” As indicated by the description in the draw-
ing area, the engineer is male.
Combining Engineering and Art
Since we had each student’s response for both the artist and
the engineering exercise, we were able to cross their group
membership for art and engineering, allowing us to see which
patterns tended to go together. Looking at Table 3, we see that
the 59 students who were in Engineering Group 1 (the group-
focused on logic and knowledge) were roughly evenly divided
between the two classes of rt responses. A little over
Finally, for Engineering Group 3, we show the response of
student 43 who had a 99% probability of being sorted into En-
gineering Group 3. This student wrote, “I think engineering a
Figure 2.
Representative students for each of the three identified groups. Student 77 (left), Student 21 (middle), Student 43 (right) represent Groups 1, 2, and
3 respectively.
half of the students (52.5%) were sorted into Art Group 1, the
group focused on imagination and communication, while
slightly less than half (47.5%) were sorted into Art Group 2,
which focused on the product of art. A similar trend can be seen
for Engineer- ing Group 3 (the group most likely to indicate the
importance of creativity and imagination in engineering. The
trend for En- gineering Group 2, however, is different. Nearly
two thirds of this group (65.5%) were sorted into Art Group 1.
Students appeared to have three distinct views of the work of
engineers. The largest group of students (Engineering Group 1)
viewed engineering as largely analytical, highlighting that en-
Figure 3.
Two-class LCA model fo r art.
Table 3.
Cross tabulation of engi n e ering and art groups.
Group Art-1 Group Art-2
Group Engi neering-1 31 28
Group Engi neering-2 19 10
Group Engineering-3 7 5
gineering required knowledge and logical thinking. In contrast,
the smallest group of students (Engineering Group 3) viewed
engineering as a creative discipline. These students highlighted
creativity, imagination, thoughts, and ideas while downplaying
knowledge and logic. The other students (Engineering Group 2)
were the most likely of the three groups to consider engineers
as people who maintain things. These three groups of students
represent the heterogeneity identified within the entire popula-
tion of students. Note that that all three groups that emerged in
our study population appear to hold relatively sophisticated,
though incomplete, conceptions of the engineering profession.
Unlike earlier studies, we did not find groups of students that
held common misconceptions such as engineers as people who
drive trains or engineers as only technicians. This is not sur-
prising because our sample consisted of students entering a
competitive engineering academy. Likely, students who held
significant misconceptions about the engineering profession
would not have applied to this program.
The students were sorted into two different views of artists.
One group (Group 1) appeared to focus on the internal, less
tangible aspects of art—emotion and expressing oneself. In
contrast the other group (Group 2) which focused on the physi-
cal product.
In contrasting students’ understanding of engineering work
with that of the work of artists, we were able to investigate
whether they saw engineering as significantly less creative. In
fact, we did not see a large difference. While students did use
the terms create and creativity slightly more often when de-
scribing artists, they used terms like design significantly more
often when describing the work of engineering, indicating that
there was not the large difference in this area that we expected.
Instead, we found that the different groups of students were
more or less likely to depict engineering as a creative discipline,
with Engineering Group 3, representing only 11% of the stu-
dents, highly likely to understand engineering as creative and
Group 1, representing 57% of the students likely to see engi-
neering as logical. Ideally, students would see engineering as
requiring both creativity and logical and analytical thinking.
In the final step of comparing how each group of engineering
Copyright © 2013 SciRe s .
students were sorted into groups based on their art response, we
found that for two of the engineering groups (Groups 1 and 3),
the students were roughly evenly sorted into the two art groups.
Engineering Group 2, however, followed a different trend.
Nearly two thirds of these students (65.5%) were sorted into
Art Group 1. This is a surprising result because this group of
students was the most likely to depict engineering as maintain-
ing something. They were not particularly likely to describe
engineering as a creative or imaginative activity. This seems to
be in opposition to their view of artists. It may be that students
sorted into Engineering Group 2 were most likely to depict
stereotypical images of both careers.
Our study was limited in a number of ways. We had a small
sample size. Only 104 students were included in our study. A
larger sample size could result in a larger number of groups and,
thus, greater understanding of the differences among groups of
students about their perceptions of the work of engineers and
artists. Additionally, we did not include a comparison group.
Our study consisted only of students from one secondary school
and, further, only those who had self-selected, applied, and
been admitted to an engineering acad emy.
Further research should include comparison groups of stu-
dents who are entering a more typical secondary school pro-
gram and a larger sample of students. In our own further re-
search, we intend to conduct a longitudinal study of this popu-
lation of students. As students progress in their education, we
will conduct the draw an engineer test and the draw an artist
test at repeated points in their education. This will allow us to
understand how the patterns in their thinking change as they
participate in a specialized engineering academy that intention-
ally integrates engineering and art.
It is vital that as we strive to help the next generation of stu-
dents develop skills to be competitive in innovative careers like
engineering, that they develop sophisticated and comprehensive
understandings of the complexity of these careers. This is not
only an issue of training those students who choose to pursue
engineering careers, but a consideration that may impact re-
cruitment of new students to pursue engineering careers.
Barman, C. R. (1999). Students’ views about scientists and school sci-
ence: Engaging K-8 teachers in a national study. Journal of Science
Teacher Education, 1 0, 43-54. doi:10.1023/A:1009424713416
Casner-Lotto, J., & Barrington, L. (2006). Are they really ready to work?
Employers’ perspectives on the basic knowledge and applied skills of
new entrants to the 21st century workforce. New York: The Confer-
ence Board.
Chambers, D. W. (1983). Stereotypic images of the scientist: The draw-
a-scientist Test. Science Education, 67, 255-265.
Capobianco, B., Diefes-Dux, H. A., Mena, I., & Weller, J. (2011).
What is an engineer? Implications of elementary school student con-
ceptions for engineering education. Journal of Engineering Educa-
tion, 100, 304-328. doi:10.1002/j.2168-9830.2011.tb00015.x
Dayton, C. M. (1998). Latent class scaling analysis. Thousand Oakes,
CA: Sage Publications
Ehrlén, K. (2009). Drawings as representations of children’s concep-
tions. International Journal of Science Education, 31, 41-57.
Finson, K. (2002). Drawing a scientist: What we do and do not know
after fifty years of drawings. School Science and Mathematics, 102,
Harlow, D., Swanson, L., Nylund-Gibson, K., & Truxler, A., (2011).
Using LCA to understand students’ responses to the question, “What
is a Day?” Science Education, 95, 477-496.
Howard, T. J., Culley, S. J., & Dekoninck, E. (2008). Describing the
creative design process by the integration of engineering design and
congnitive psychology literature. Design Studies, 29, 160-180.
Kwon, S., Nam, D., & Lee, T. (2011). The effects of convergence edu-
cation based STEAM on elementary school students’ creative per-
sonality. In T. Hirashima et al. (Eds.), Proceedings of the 19th Inter-
national Conference on Computers in Education, Chiang Mai: Asia-
Pacific Society for Computers in Education.
Knight, M., & Cunningham, C. (2004). Draw an engineer test (DAET):
Development of a tool to investigate students’ ideas about engineers
and engineering. ASEE Annual Conference and Exposition, Salt Lake
City, 20-23 June 2004
Katehi, L., Pearson, G., & Feder, M. (2009). Engineering in K-12 edu-
cation: Understanding the status and improving the prospects. Wash-
ington DC: National Academy Press.
Kazerounian, K., & Foley, S. (2007). Barriers to creativity in engineer-
ing education: A study of instructors and students perceptions. Jour-
nal of Mechanical Design , 129, 761. doi:10.1115/1.2739569
Magidson, J., & Vermunt, J. K. (2004). Latent class models. In D.
Kaplan (Ed.), Handbook of quantitative methodology for the social
sciences. Newbury Park, CA: Sage Publication s .
Muthen, L. K., & Muthen, B. O. (19 98-2012). Mplus user's guide. Los
Angeles, CA: Muthen & Muthen.
National Academy of Engineering (2004). The engineer of 2020. Wash-
ington DC: National Academies Press.
National Research Council (2012). A framework for K-12 science edu-
cation: Practices, crosscutting concepts, and core ideas. Washington
DC: National Academy Press.
Nylund, K. L., Asparouhov, T., & Muthén, B. O. (2007). Deciding on
the number of classes in latent class analysis and growth mixture
modeling: A Monte Carlo simulation study. Structural Equation Mo-
deling, 14, 535-569. doi:10.1080/10705510701575396
Schmidt, A. (2011). Creativity in science: Tensions between perception
and practice. Creative Education, 2, 435-445.
Trilling, B., & Fadel, C. (2009). 21st century skills: Learning for life in
our times. San Francisco, CA: Jossey-Bass.
Yakman, G. (2010). What is STE@M?
Copyright © 2013 SciRe s . 321