Diversity in Design Teams: An Investigation of Learning Styles and their Impact on Team Performance and Innovation

Abstract

In this paper, we examine the role of diversity in design team performance, and discuss how diversity factors affect the dynamics and success of a design team. In particular, we focus on diversity in learning styles, as defined by Kolb’s Experiential Learning Theory. We also consider other demographic factors, such as discipline and gender. We present data gathered over two semesters of a multidisciplinary, project-based graduate level design course offered at the University of California at Berkeley. The data were captured through a series of surveys administered during the semester, first to collect diversity information on learning styles and standard demographics, and then to assess team performance as students reflected on their team interactions. We examine and compare the overall learning style breakdown of students in the class, along with an analysis of the teams. The results of our analyses offer insights into how students with different learning styles appear to contribute to design team performance. We provide recommendations that will help inform design educators on how to enhance overall team performance and innovation, with an understanding of learning style differences.

Full text

Diversity in Design Teams: An Investigation of Learning
Styles and their Impact on Team Performance and
Innovation*
KIMBERLY LAU
Department of Mechanical Engineering, University of California at Berkeley, Berkeley, CA 94709–1742, USA.
E-mail: lauk@berkeley.edu
SARA L. BECKMAN
Haas School of Business, University of California at Berkeley, Berkeley, CA 94709–1742, USA. E-mail: beckman@haas.berkeley.edu
ALICE M. AGOGINO
Department of Mechanical Engineering, University of California at Berkeley, Berkeley, CA 94709–1742, USA.
E-mail: agogino@berkeley.edu
In this paper, we examine the role of diversity in design team performance, and discuss how diversity factors affect the
dynamics and success of a design team. In particular, we focus on diversity in learning styles, as defined by Kolb’s
Experiential Learning Theory. We also consider other demographic factors, such as discipline and gender. We present data
gathered over two semesters of a multidisciplinary, project-based graduate level design course offered at the University of
California at Berkeley. The data were captured through a series of surveys administered during the semester, first to collect
diversity information on learning styles and standard demographics, and then to assess team performance as students
reflected on their team interactions. We examine and compare the overall learning style breakdown of students in the class,
along with an analysis of the teams. The results of our analyses offer insights into how students with different learning styles
appear to contribute to design team performance. We provide recommendations that will help inform design educators on
how to enhance overall team performance and innovation, with an understanding of learning style differences.
Keywords: learning styles; design teams; team performance; Kolb’s experiential learning
1. Introduction and background
With ever-changing technologies and rising market
competition, it is increasingly important to design
innovative products. Teamwork leads to innova-
tion more frequently than individual efforts [1], and
organizations that focus on new product develop-
ment invest in developing their teams to achieve a
high level of creativity and innovation. This begs
the question of how to best form and manage
teams that will successfully build quality products.
For example, should teams consist of experts from
the same field and with similar reinforcing experi-
ences, or should the teams be composed of experts
from diverse backgrounds and personality types?
Many companies rely on cross-functional teams to
benefit from diverse perspectives, experiences,
and design-for-X expertise, including members
from engineering, business, industrial design, and
more [2].
A variety of diversity factors may affect new
product development team performance outcomes.
Individual differences—be they cultural, gender, or
cognitive—cause people to approach a single situa-
tion in various ways. In the academic setting, such
differences may influence how a person learns,
solves problems, and interacts with peers and team
members.
In recent years, design education researchers have
begun exploring the relationship between learning
styles and learning in design. From this research, a
variety of learning characterizations have been
identified. Newland categorizes learners as com-
mon sense, dynamic, contemplative, and zealous
[3]. Leary classifies a person’s behavior along two
axes: dominant versus submissive and friendly
versus critical [4]. Felder examines learning under
sensory versus intuitive, visual versus auditory,
inductive versus deductive, and active versus reflec-
tive dimensions [5].
In his Experiential Learning Theory (ELT), Kolb
posits that a person acquires knowledge by grasping
and transforming experience [6, 7]. He defines these
experiences along two dialectically related continua:
Concrete Experience (CE) or Abstract Conceptua-
lization (AC), which measure how an individual
perceives information, and Reflective Observation
(RO) or Active Experimentation (AE), which mea-
sure how an individual processes information.
These two continua intersect to create four quad-
rants, each representing a different learning style
(Fig. 1). Each individual’s learning style is deter-
* Accepted 20 August 2011. 293
International Journal of Engineering Education Vol. 28, No. 2, pp. 293–301, 2012 0949-149X/91 $3.00+0.00
Printed in Great Britain # 2012 TEMPUS Publications.

mined by which combination of learning modes he
or she prefers for perceiving and processing infor-
mation.
The five learning styles are:
1. Assimilating (Abstract Conceptualization and
Reflective Observation): best at synthesizing a
wide range of information into a useful, logical
form
2. Converging (Abstract Conceptualization and
Active Experimentation): logical and orga-
nized, good at finding practical applications
for ideas and theories
3. Accommodating (Concrete Experience and
Active Experimentation): hands-on learning,
practical experience, sensing and intuitive
risk-takers
4. Diverging (Concrete Experience and Reflective
Observation): best at viewing concrete situa-
tions from many different points of view, facil-
itate idea generation
5. Balanced (Abstract Conceptualization and
Concrete Experience or Reflective Observation
and Active Experimentation): has no strong
preference for either extreme of the Processing
or Perception continuums combined, well-
balanced
Learning styles are particularly relevant to design
for its connection to innovation as a learning pro-
cess [8]. Although there have been extensive studies
relating to learning styles, research surrounding
Kolb learning styles in design teams has not yet
been fully explored. In this paper, we will examine
the effect of learning styles on design team perfor-
mance in the educational setting.
2. Subjects and methods
For this study, we gathered data from students
enrolled in ‘ME290P: Managing the New Product
Development Process: Design Theory and Meth-
ods’, a graduate-level, multidisciplinary design
course offered at University of California at Berke-
ley (UCB). This is a project-based learning class,
whereby engineering, business and science students
from UCB, along with industrial design students
from the California College of Arts (CCA), engage
in small design teams to solve a real-world, open-
ended design challenge. Over the semester, students
learn the tools and techniques of new product
development and apply them in their semester-
long class projects, while also developing skills
important for design and innovation outside the
academic environment [9].
This study was performed over two semesters of
ME290P, in Fall 2009 (N = 70, 16 teams) and in Fall
2010 (N = 75, 17 teams). Table 1 shows the break-
down of students, by discipline and gender.
We conducted this study with three surveys
during the semester. The first survey was adminis-
tered at the beginning of the semester and was
comprised of two parts: a demographic question-
naire and the Kolb Learning Style Inventory (LSI).
This survey served to help students understand their
K. Lau, S. L. Beckman and A. Agogino294
Fig. 1. Kolb Learning Styles. [6]

personal styles in observations, framing, and think-
ing, as well as the preferences of their teammates; the
results were intended to drive productive team
dynamics and processes from the start of the pro-
ject.
Midway through the semester, we administered a
Peer Review and Team Assessment survey to the
class. The purpose of this survey was for students to
provide feedback on the current state of their
project and team. The questions were divided into
seven sections: Goals, Roles, Processes and Proce-
dures, Relationships, Team Effectiveness, Team
Performance, and Time Management. The students
were also asked to evaluate each teammate on his or
her contributions to the team, in such ways as
dividing up 100 points among all team members,
including oneself. These results were presented to
the teams and served as a discussion point for
making improvements in the remainder of the
semester.
The third survey was administered at the end of
the semester and was similar to the mid-semester
survey with the goal of tracking improvements. The
results for Fall 2009 and Fall 2010 were analyzed
separately when appropriate because the surveys
were worded slightly different in each year.
3. Results and discussion
3.1 Learning styles of study population versus
general population
The distribution of learning styles in our entire
study population is shown in Fig. 2. Overall, the
class has a relatively similar learning style break-
down between the Fall 2009 and 2010 groups. The
students with a converging learning style are most
dominant across both semesters. The only differ-
ence is the marked paucity of divergers in Fall 2010.
Students with balanced learning styles are those
who have stronger preferences along a single axis,
either the Perception (AC+CE) or Processing
(AE+RO) Continuum. In our class, twenty-three
students demonstrated preferences in the Proces-
sing Continuum (AE+RO), for watching and doing,
versus four students for the feeling and thinking
Perception Continuum (AC+CE).
Table 2 shows the scores from each learning style
mode (Concrete Experience, Reflective Observa-
tion, Abstract Conceptualization, Active Experi-
mentation) for the two classes. The mean values
for each mode are relatively close and within 2
points of one another, but the range of individual
scores is wide (nearly 30 point differential for every
mode). This distribution is similar to that reported
for research universities in the Kolb manual on LSI
[7, p. 13].
Learning styles are also connected to our educa-
tional and professional experiences as shown by a
number of studies examining learning styles and
educational or career interests [10–13]. Kolb posits
that some learning styles will be typical in certain
vocations, because of the experiences one under-
takes in studying a specific profession [6, 11]. For
Diversity in Design Teams 295
Table 1. Class breakdown by discipline and gender—Fall 2009
and 2010, combined
Male Female Total
Engineering 41 13 54
MBA 33 10 43
Science 11 6 17
Industrial Design 11 8 19
Other 7 5 12
Total 103 42 145
Fig. 2. Learning Styles of Design Students.

instance, Kolb found that individuals in human-
related professions (educators, social workers, nur-
sing) tended towards concrete learning and were
more likely to be accommodators [7]. Engineers and
decision-makers were high in converging learning
styles, whereas professionals in the arts and huma-
nities were high in diverging styles. Mathematicians
and scientists mostly preferred the assimilating
learning style.
In our study group, the converging learning style
is most dominant (Table 3). This is not surprising
given the number of engineers and business students
in the class. However, there is a significant paucity of
divergers, except among the Industrial Design stu-
dents.
When comparing learning styles by gender,
women and men typically demonstrate different
learning style preferences [6]. In particular, men
score higher on the Abstract Conceptualization
spectrum and fit well with the Assimilating or
Converging styles. On the other hand, women
prefer practical, hands-on environments [14,15]
with either Diverging or Accommodating learning
styles.
In our study population, a higher percentage of
women exhibit the Assimilating learning style over
the Diverging learning style, and also have a higher
percentage of Assimilators than men (Table 4). On
the other hand, Kolb also found that learning styles
either changed over one’s academic career, or else
universities and graduate schools favor students
with higher Abstract Conceptualization (assimilat-
ing or converging). So it is not surprising that AC
was high for both men and women in our graduate
course, with the male percentage higher.
3.2 Learning style profiles of teams
To analyze learning styles on the project team level,
we identified each team’s overall learning profile by
averaging the team members’ individual scores on
the four stages of learning (CE, RO, AC, AE). In
Fig. 3, we illustrate the learning style profiles of two
distinct teams and of the class average. Team 1
represents the team with the most diverging learning
style in the class and Team 2 represents the team
with the most converging learning style in the class.
Each polygon represents one team’s learning style
profile. The points at which the polygons intersect
with each axis represent the team’s average score in
that respective continuum. The longer lines demon-
strate greater strengths in their respective quad-
rants. We can observe that Team 1 has a much
longer line in the ‘Diverging’ quadrant, representing
its more dominant learning style, while Team 2’s line
is longest between AC and AE in the ‘Converging’
region. The Class Average falls between these two
profiles and shows a stronger preference for conver-
ging.
3.3 Learning styles and team assessment results
With these aggregated learning style profiles, we
then examine how design teams rated themselves on
the mid-semester surveys to understand team coher-
K. Lau, S. L. Beckman and A. Agogino296
Table 2. Learning Style Scores
CE RO AC AE
Fall 2010 25.5 26.2 34.1 34.2 Mean
6.3 6.9 6.6 6.3 Std Dev.
15–39 13–41 20–46 21–47 Range
Fall 2009 26.1 28.2 32.4 33.2 Mean
6.6 7.0 7.1 7.5 Std Dev.
15–44 15–41 14–46 17–47 Range
Table 3. Learning Styles by discipline—Fall 2009 and 2010, combined
Engineering Business Industrial Design Other Total
Accommodating 5 (9%) 9 (21%) 4 (21%) 4 (14%) 22
Assimilating 9 (17%) 3 (7%) 4 (21%) 5 (17%) 21
Balanced 10 (19%) 9 (21%) 3 (16%) 5 (17%) 27
Converging 27 (50%) 19 (44%) 5 (26%) 12 (41%) 63
Diverging 3 (6%) 3 (7%) 3 (16%) 3 (10%) 12
Total 54 43 19 29 145
‘Other’ represents the Science and Humanities fields, such as Genetics and Plant Biology, Art History, and Information Science.
Table 4. Learning styles by gender—Fall 2009 and 2010, com-
bined
Female Male
Accommodating 8 (19%) 14 (14%)
Assimilating 7 (17%) 14 (13%)
Balanced 9 (21%) 18 (17%)
Converging 14 (33%) 49 (47%)
Diverging 4 (10%) 8 (8%)
Total 42 103

ence and performance. We compare design teams
with respect to the number of convergers within the
team because of the converging dominance in the
class. Tables 5 and 6 show the results of the mid-
semester survey, evaluated against the number of
convergers on a team for Fall 2009 and Fall 2010
respectively.
The bolded numbers represent the results that are
statistically significant (p < 0.05). Each column
represents a different group of teams, which are
clustered by the number of convergers in the team.
The symbols (*, {, and {) identify the pair of groups
in each row between which a statistically significant
difference was found. For example, in response to
Question 1: ‘As a team, we are clear about our
purpose’, the teams with one converger scored
significantly higher (4.25) in contrast with teams
with three convergers (3.7). The results were not
statistically significant between the other popula-
tions. In Question 8: ‘The team enjoys working
together’, the score attained by teams with one
converger (4.56) was significantly larger than both
the score of the team with two convergers (4.13) and
the score of the team with three convergers (3.91).
The results from Fall 2009 were normalized to a 5-
point scale.
The most striking observation here is that the
ratings significantly decrease as the number of
convergers on the team increases, specifically from
one to four convergers. This seems to imply that the
converging learners do affect design teams, with
fewer convergers providing greater benefit.
Indeed, converging learners are valuable to design
teams—they can find practical uses for ideas and
enjoy experimenting with new ideas. However, they
also prefer to internalize their theories before acting.
Perhaps an entire team of persistent thinkers trans-
lates to little or no reflective dialogue within the
team, and limited or slower success in teamwork.
Many of the questions showing statistically sig-
nificant results pertain to working as a team. Of
these, the most direct statement about team inter-
actions: ‘The team enjoys working together’, shows
teams with one converger rating highest of all. One
might have expected a more diverse team, particu-
larly one comprised of different learning styles, to
clash with one another; however, here the more
homogeneous teams, with respect to converging
learning styles, report more tension. This may also
be indicative of how teams spend their time
together. In questions relating to productivity (Q7,
Q10, Q16), teams with one converger report making
the best use of time. This could be because teams
with multiple convergers were so alike that team
members were complacent with one another, result-
ing in a lack of design momentum; or they may have
experienced greater conflict because of strong, simi-
lar personalities, and squandered time arguing over
simple ideas and tasks. More broadly, the teams
with one converging learner believe themselves to be
Diversity in Design Teams 297
Fig. 3. Learning Style Profiles.

the highest-performing teams (Q12) and with the
highest quality outputs (Q11), rating nearly one
point above teams with four converging learners.
Interestingly, when the teams were asked about
innovation: ‘Our team is innovative’, no group
showed statistically significant different results. So
although teams with one converging learner believe
they are most high-performing and productive of all
teams, they do not necessarily believe they are any
more innovative.
The leading question is thus how the learning
style profiles compare between the different teams,
with respect to the number of convergers, and
whether these perceptions are actually mirrored in
the team deliverables.
3.4 Learning styles and team performance results
Table 7 presents the average learning style profiles
of the entire team, clustered by the number of
convergers on each team. Recall that the converging
learning style is defined by the Abstract Conceptua-
lization + Active Experimentation combination and
was the predominant learning style in our sample.
As expected, we see that the scores for AC and AE
rise and the scores for CE and RO fall for the team as
the number of convergers increases.
We observe that the T1 and T2 teams have
remarkably similar team profiles (within 1 point),
yet T2 teams rate themselves lower than T1 teams in
all but one question of the mid-semester team and
peer assessments. This implies that it is not just the
learning profile of the converging learner that
matters to a team, but the number of convergers
on the team. Ultimately, the team benefits from a
very strong converging team member, but may need
equally strong non-converging teammates to bal-
ance the entire team out.
Table 8 shows the team’s actual project score by
K. Lau, S. L. Beckman and A. Agogino298
Table 5. Mid-semester Assessment results, by # convergers on team (Fall 2009)
1 converger 2 convergers 3 convergers 4 convergers
1 As a team, we are clear about our purpose. 4.25* 4.20 3.70* 3.61
2 The team is successfully achieving project goals to date. 4.37* 3.72* 4.22 3.83
3 The team is committed to learning about the tools, techniques and
process taught in this class.
4.13 3.93* 4.43* 3.89
4 The members of my team have a shared understanding of the roles
and responsibilities played by individuals on the team.
3.65* 3.63 3.70 3.19*
5 All members of the team have shared equitably in the tasks
performed to date.
3.73* 3.83{ 3.70
2.64*{
6 We have two-way communication with our speakers/design coaches. 4.05* 3.97{ 3.65
2.78*{
7 We spend sufficient time making sure the team is working on what we
are supposed to be doing.
4.05* 3.60 3.83 3.19*
8 The team enjoys working together.
4.56*{ 4.13* 3.91{ 4.03
9 As a team, we are accomplishing what we have set out to accomplish. 4.21* 3.93 3.70* 3.61
10 The time we spend together as a team is productive.
4.52*{ 3.93* 4.06{ 3.75
11 What we produce as a team are high-quality outputs.
4.40*{{ 3.97* 3.80{ 3.47{
12 Overall, we are a high-performing team.
4.29*{ 3.80* 3.59{ 3.47
Table 6. Mid-semester Assessment results, by # convergers on team (Fall 2010)
1 converger 2 convergers 4 convergers
13 We have discussed our individual learning goals for the class and the project
with each other.
4.41*{ 3.90* 3.75{
14 We have agendas for our team meetings.
4.45*{ 3.67* 3.44{
15 We have the skills and experience on the team that we need to be successful. 4.45* 4.07* 4.00
16 Our team meetings are productive. 4.45* 4.27 3.81*
17 I am learning valuable lessons about my own leadership by being on this team. 4.41* 4.13 3.81*
Table 7. Average Learning Style Profiles of Entire Team, by # Convergers
Learning Styles of Entire Team
Abstract
Conceptualization Active Experience Concrete Experience
Reflective
Observation
Teams with 1 Converger (T1) 32.1 34.0 26.0 27.8
Teams with 2 Convergers (T2) 33.1 33.9 26.4 26.5
Teams with 3 Convergers (T3) 33.1 35.5 25.7 25.7
Teams with 4 Convergers (T4) 36.5 35.2 25.0 23.3

external reviewers and faculty at the end of the
semester. These external reviewers included design
industry judges, who ranked projects according to
the quality of their mission statement, customer/
user needs, concept generation, concept selection,
prototype, and business analysis. This ranking is
taken as a proxy for greater innovation and overall
success. The table also includes the number of
convergers in the team, the team composition in
regards to learning styles, and gender.
We note in Table 8 that of the eleven teams with
only one converger, six appear in the top ten of the
list of highest performing teams. We also note that
the highest performing teams demonstrate gender
diversity. Conversely, the lowest performing teams
lacked gender diversity; three of the bottom four
teams were either all male or all female. Although
the lowest performing team had one woman and
three men, the team was dominated by the male
students; the female was a shy CCA undergraduate
student and a non-native English speaker. There is
no pattern that appears among teams with 2, 3, and
4 convergers; rather, they are sprinkled through the
grade distribution.
We also compared the grades with the midterm
evaluation scores to uncover any specific correla-
tions between how a team perceived itself and how
they actually performed at the end of the semester.
The results show little correlation between the mid-
semester team self-assessments and their actual
project performance when measured with the
entire class, with the highest r-value at 0.30. The
instructors speculate that their interventions may
have been effective overall—extreme problems were
addressed and corrective action taken. Student
feedback at the end of the semester praised the
value of the teamwork skills developed in the class.
An analysis of end-of-semester evaluation scores
and project grades did yield significant correlation
coefficients and many of the values were much
higher, indicative that the final team self-assessment
was correlated with final grades. For example, when
we compare the teams by the number of convergers
with their final team grades, we reveal some inter-
esting relationships. In Fig. 4, we see a high,
statistically significant correlation between how
productive teams with one converger believe their
meetings to be and their final project grade.
3.5 Learning styles and combined mid- and end-of-
semester analysis
Table 9 presents the comparison of results from the
midterm and end-of-semester surveys, for Fall 2009.
Overall, the post-semester scores are higher with the
Diversity in Design Teams 299
Table 8. Overall Project Score of Teams by # Convergers
Overall Score # Convergers Learning Style Breakdown Male Female Team
4.26 1 1 Accom, 2 Assim, 1 Con 1 3 2010–3
4.10 1 2 Accom, 2 Bal, 1 Con 2 3 2010–13
4.08 1 1 Con, 1 Assim, 1 Bal 2 1 2009–9
4.07 2 1 Accom, 1 Bal, 2 Con 2 2 2010–11
4.02 4 1 Assim, 4 Con 4 1 2010–1
4.01 1 1 Accom, 1 Con, 1 Bal 2 1 2009–5
3.95 1 1 Accom, 2 Div, 1 Bal, 1 Con 2 3 2009–3
3.94 1 2 Assim, 1 Bal, 1 Con 3 1 2010–15
3.92 4 1 Assim, 1 Bal, 4 Con 6 0 2010–14
3.92 2 1 Bal, 2 Con 3 0 2010–2
3.90 3 3 Con, 1 Assim, 1 Div 2 3 2009–8
3.90 2 2 Assim, 1 Bal, 1 Accom, 2 Con 0 6 2009–15
3.90 2 2 Con, 2 Div, 1 Accom, 1 Bal 1 5 2009–13
3.86 1 1 Bal, 1 Accom, 1 Con 0 3 2009–10
3.86 4 1 Assim, 4 Con 5 0 2010–12
3.85 2 2 Con, 1 Div 2 1 2009–12
3.84 1 1 Con, 1 Accom, 1 Bal 1 2 2009–16
3.83 0 2 Bal 0 2 2009–7
3.81 2 3 Assim, 2 Con 2 3 2010–17
3.75 2 3 Accom, 2 Con 4 1 2010–9
3.74 2 2 Bal, 2 Con, 1 Div, 1 Accom 1 5 2009–6
3.73 3 1 Div, 1 Accom, 3 Con, 1 Bal 3 3 2009–1
3.63 2 2 Con, 1 Assim, 1 Div 1 3 2009–14
3.53 1 1 Accom, 1 Assim, 2 Bal, 1 Con 4 1 2010–5
3.53 2 1 Accom, 1 Assim, 2 Con 1 3 2010–4
3.53 3 3 Con, 1 Bal, 1 Assim 0 5 2009–11
3.50 4 4 Con, 1 Div, 1 Accom 1 4 2009–2
3.50 3 1 Accom, 3 Con 3 1 2010–16
3.42 2 3 Bal, 2 Con 4 1 2010–10
3.40 1 2 Assim, 1 Bal, 1 Con 0 5 2009–4
3.29 2 1 Accom, 1 Bal, 2 Con 4 0 2010–6
3.07 0 1 Assim, 1 Bal, 1 Div 0 3 2010–7
3.01 1 2 Accom, 1 Con, 1 Div 3 1 2010–8

ones in bold being statistically significant. Here, we
see the Converging and Balanced students showing
the most significant perception of team improve-
ments. This is a favorable result, as it indicates the
students are likely becoming more comfortable with
themselves, their team, and project over time, or
that the teaching staff interventions were successful
in dissipating team conflict, or both.
4. Conclusions and recommendations
In this paper, we explored the Kolb learning styles of
students in a graduate-level design course over two
semesters. We found that the students in this course
were most dominant in the converging learning
style, and most lacking in the diverging learning
style. We also found that design teams with just one
converger generally performed better in their self-
perception of team performance than teams with
multiple convergers, at least before substantial
instructor intervention. There was some indication
that teams with a single converger dominated the
highest performing teams judged at the end of the
semester by external reviewers. As all of the teams
had diversity in learning styles, except those over-
dominated by convergers, we cannot draw any
other conclusions on the benefits of diversity in
learning styles. We do plan future research on
small projects composed of teams with homoge-
neous learning styles to investigate learning style
diversity impacts on teamwork further.
We note that the lowest performing teams lacked
gender diversity, as opposed to the teams at the top
of the rating list with stronger gender diversity. This
result could be a consequence of gender differences
in learning styles or personality types. The results
were only suggestive, but are strong enough to
motivate further research into this intersection of
cognitive styles and gender on design teams.
We also found that a mid-term evaluation of
perceived team performance with effective instruc-
tor intervention increased the team perception of
their final performance at the end of the semester.
This was further validated from positive teacher
evaluations on teamwork instruction and interven-
tions.
K. Lau, S. L. Beckman and A. Agogino300
Fig. 4. Mid-semester Evaluations versus Overall Project Score.
Table 9. Midterm and End-of-Semester Team Evaluations
Accommodating Balanced Converging
Pre Post Pre Post Pre Post
1 As a team, we are clear about our purpose. 4.38 4.58 4.47 4.85 3.87* 4.30*
2 As a team, we are clear about our shared values. 4.17 4.38 4.09* 4.55* 3.70 3.90
3 The team is committed to learning about the tools,
techniques and process taught in this class.
4.27 4.48 3.86* 4.62* 4.20 4.33
4 The members of my team have a shared understanding
of the roles and responsibilities played by individuals
on the team.
3.33 3.96 3.94* 4.47* 3.47{ 4.07{
5 As a team, we are accomplishing what we have set out
to accomplish.
3.96* 4.69* 4.24 4.62 3.87{ 4.47{
6 What we produce as a team are high-quality outputs. 4.27 4.58 4.09* 4.62* 4.00 4.30
7 We are taking advantage of the specific areas of
expertise of the individual members of the team.
3.44 4.27 4.24 4.50 3.88* 4.37*

These results provide support for recommending
diverse representation among design teams. Teams
that do not have such diversity may benefit from
interventions that encourage teams to think outside
their comfort zones and to assume different roles
amongst themselves to help spur more meaningful
progress and productive teamwork. Ultimately,
understanding and utilizing the different learning
styles will benefit design teams and enable members
to perform at their best levels.
References
1. A. Hargadon, How Breakthroughs Happen: The Surprising
Truth About How Companies Innovate, Harvard Business
School Press, Boston, 2003.
2. D. R. Denison, S. L. Hart and J. A. Kahn, From chimneys to
cross-functional teams: developing and validating a diagnos-
tic model, Academy of Management Journal, 39, 1996,
pp. 1005–1023.
3. P. Newland, J. A. Powell and C. Creed, Understanding
architectural designers’ selective information handling,
Design Studies, 8(1), 1987, pp. 2–16.
4. T. F. Leary, Interpersonal diagnosis of personality, Ronald
Press Co., NY, 1957.
5. R. M. Felder and L. K. Silverman, Learning and Teaching
Styles in Engineering Education, Engineering Education,
78(7), 1988, pp. 674–681.
6. D. A. Kolb, Experiential Learning: Experience as the Source
of Learning and Development, Prentice Hall, Englewood
Cliffs, N. J., 1984.
7. A. Y. Kolb and D. A. Kolb, The Kolb Learning Style
Inventory—Version 3.1, 2005 Technical Specifications,
HayGroup, 2005.
8. S. L. Beckman and M. Barry, Innovation as a Learning
Process: Embedding Design Thinking, California Manage-
ment Review, 50(1), 2007, pp. 25–56.
9. C. L. Cobb, A. M. Agogino, S. L. Beckman and L. Speer,
Enabling and Characterizing Twenty-First Century Skills in
New Product Development Teams, International Journal of
Engineering Education, 24(2), 2008, pp. 420–433.
10. Jonathan Fox and Suzanne Bartholomae, Student learning
style and educational outcomes: evidence from a family
financial management course, Financial Services Review, 8,
1999, pp. 235–251.
11. R. Loo, A Meta-Analytic Examination of Kolb’s Learning
Style Preferences Among Business Majors, Journal of Educa-
tion for Business, 77(5), 2002, pp. 252–256.
12. P. T. Engels and C. de Gara, Learning Styles of medical
students, general surgery residents, and general surgeons:
implications for surgical education, BMC Medical Educa-
tion, 10, 2010.
13. S. L. Stradley, B. D. Buckley, T. W. Kaminski, M. B
Horodyski, D. Fleming and C. M. Janelle, A Nationwide
Learning-Style Assessment of Undergraduate Athletic
Training Students in CAAHEP-Accredited Athletic Train-
ing Programs, Journal of Athletic Training, 37(4), 2002,
pp. 141–146
14. Marge Philbin, Elizabeth Meier, Sherri Huffman and Patri-
cia Boveri, ‘A survey of gender and learning styles’, Sex
Roles, 32, 1995, pp. 485–494.
15. S. E. Severiens and G. T. M. Ten Dam, Gender differences in
learning styles: a narrative review and quantitative meta-
analysis, Higher Education, 27, 1994, pp. 487–501.
Kimberly Lau is a doctoral candidate in the Department of Mechanical Engineering at the University of California at
Berkeley. She received her M.S. and B.S. degrees in Mechanical Engineering from the University of California at Berkeley,
and is a member the Berkeley Institute of Design (BiD).
Sara L. Beckman is a Senior Lecturer at the University of California at Berkeley, Haas School of Business. She has taught
for Stanford University’s Department of Industrial Engineering and Engineering Management, was a visiting faculty at
MIT’s Leaders for Manufacturing Program, ran the Change Management Team at Hewlett-Packard and consulted in
Operations Management at Booz, Allen & Hamilton. She has B.S., MS and Ph.D. degrees from Stanford University.
Alice Agogino is the Roscoe and Elizabeth Hughes Professor of Mechanical Engineering and affiliated faculty in UC
Berkeley’s Haas School of Business, Berkeley Institute of Design (BiD), Energy Resources Group (ERG) and SESAME
(Studies in Engineering, Science, and Mathematics Education). She also directs the Berkeley Expert Systems Technologies/
Berkeley Energy and Sustainable Technologies (BEST) Lab. Her Ph.D. is from Stanford University.
Diversity in Design Teams 301