International Research & Education in Design Conference 2019 — REDES2019
14 & 15 November 2019, Lisbon School of Architecture of the University of Lisbon, Lisbon, Portugal
Industrial designers problem-solving and designing: an EEG study
Sonia Vieira1, John S. Gero2, Jessica Delmoral3, Valentin Gattol4, Carlos Fernandes5, Marco
Parente6, António A. Fernandes6
1Politecnico di Milano, Italy
2University North Carolina at Charlotte, NC, United States
3Institute of Science and Innovation in Mechanical and Industrial Engineering-FEUP Portugal
4Austria Institute of Technology, Austria
5Saint John Hospital, Porto, Portugal
6Faculty of Engineering University of Porto, Portugal
This paper presents results from an experiment to determine brain activation differences between
problem-solving and designing of industrial designers. The study adopted and extended the tasks
described in a fMRI study of design cognition and measured brain activation using
electroencephalography (EEG). By taking advantage of EEG's high temporal resolution we focus on
time-related neural responses during problem-solving compared to design tasks. The experiment
consists of multiple tasks: problem-solving, basic design and open design using a tangible interface.
The tasks are preceded by a familiarizing pre-task and then extended to a fourth open design task
using free-hand sketching. The results indicate design cognition differences in the brain
measurements of task-related power and temporal analysis of transformed power between the
constrained problem-solving task and the open design tasks. Statistical analyses indicate increased
brain activation when designing compared to problem-solving. Results of time-related neural
responses connected to Brodmann’ areas cognitive functions, contribute to a better understanding of
industrial designers’ cognition in open and constrained design spaces and how the problem statement
can constrain or expand conceptual expansion.
Keywords: design, problem-solving, industrial designers, design neurocognition.
Introduction
The study of the cognitive behavior of industrial designers while designing, based on methods
such as protocol analysis (Ericsson and Simon, 1983, Kan and Gero, 2017), has produced
important results covering foundational aspects of design cognition. The notions of problem
space and solution space have been the ground of interpretations of the designing process
(e.g., Kruger and Cross, 2006) in the last fifty years of design research (Jones, 1963). The
problem-solving view of design claims that the designing process commences with an
exploration within the problem space (Goel and Pirolli, 1992). Alternative perspectives assert
that design thinking is primarily solution focused (Dorst, 2011; Darke, 1979). One of the
initial and core research questions is whether designing as a cognitive process is distinct from
problem-solving (Goel and Pirolli, 1992; Visser, 2009). Neurophysiological studies offer a
new integrative perspective into how brain behavior progresses during the designing process,
which makes them a robust tool for connecting to design cognition. Recent design studies
based on functional magnetic resonance imaging (fMRI) (Alexiou, et al., 2009; GoucherLambert, et al., 2017), electroencephalography (EEG) (Liu et al., 2018; Liu et al., 2016;
Liang, et al., 2017) and functional near-infrared spectroscopy (fNIRS) (Shealy, Hu and Gero,
2018) attempt to understand designing from a neurophysiological perspective. The present
paper describes a study from a larger research project whose goal is to correlate design
cognition with brain activation of designers across design domains. EEG's high temporal
resolution makes it a more suitable tool than fMRI (Hinterberger, et al. 2014; Dickter and
Kieffaber, 2014) to investigate designing as a temporal activity. The study reported in this
1
paper is based on the analysis of industrial designers’ brain activation using an EEG headset
in the context of performing problem-solving and design tasks in a laboratory setting. The
objective of the study is:
• investigate the use of the EEG technique to distinguish design from problem-solving
in industrial designers.
We adopt and extend the tasks described in a controlled experiment of an fMRI-based design
study (Alexiou, et al., 2009). That study suggested higher activation of the dorsolateral
prefrontal cortex is consistent for design tasks and ill-structured problems and recruits a more
extensive network of brain areas than problem-solving. We postulate the following
hypotheses:
Hypothesis 1. Design neurocognition of industrial designers when problem-solving and
designing are different.
Hypothesis 2. Neurocognitive temporal distributions of activations of industrial designers are
significantly different across design tasks.
Experiment Design
We have adopted and replicated two of the layout tasks described in the Alexiou, et al. (2009)
fMRI-based study. We extended their experiment to a third open layout design task with the
purpose of opening the solution space to produce a block experiment as depicted in Table 1
and Figure 1. The set of three tasks is preceded by a pre-task so that participants can become
acquainted with the physical interface and headset. The three tasks are followed by a fourth
open design free-hand sketching task. A tangible interface for individual task performance
was built based on magnetic material for easy handling. The pre-task was designed so that
participants can familiarize themselves with the use of the EEG headset, and necessary
corrections can be made before advancing to the block experiment, manoeuvring the magnetic
pieces that make up the physical interface and prevent participants from getting fixated in the
problem-solving Task 1. The block experiment consists of a sequence of 3 tasks: problemsolving, basic design and open layout design, as illustrated in Figure 1. We have matched
Tasks 1 and 2 with the problem-solving and design tasks from Alexiou, et al. (2009) in terms
of requests, number of constraints, stimuli and number of instructions. The open layout design
Task 3 provides an enlargement of the problem space and the solution space and the
opportunity of evaluating and reformulating the previous design solutions. In Task 4, the
participants are asked to propose and represent the outline design of a future personal
entertainment system, which is an ill-defined and fully unconstrained task unrelated to formal
problem-solving. The Mikado pick up sticks game was given to the participants to play in the
breaks between tasks to break their focus on the tasks.
Table 1: Description of the tasks.
Task 1 Problem-solving
In Task 1 the design of a set
of furniture is available and
three conditions are given as
requirements. The task
consists of placing the
magnetic pieces inside a
given area of a room with a
door, a window and a
balcony.
Task 2 Basic design
In Task 2 the same design
set of furniture is available,
and three requests are made.
The basic design task
consists of placing the
furniture inside a given room
area according to each
participant’ notions of
functional and comfortable
using at least three pieces.
2
Task 3 Open design
In Task 3 the same
design available is
complemented with a second
board of movable pieces that
comprise all the fixed
elements of the previous
tasks, namely, the walls, the
door, the window and the
balcony. The participant is
told to arrange a space.
Figure 1: Problem-solving Task 1, Basic Design Task 2 and Open Layout Design Task 3.
Differently from the original tasks (Alexiou, et al. 2009), the magnetic pieces were placed at
the top of the vertical magnetic board to prevent signal noise due to eye and head horizontal
movements. Two video cameras for capturing the participant’s face and activity and the audio
recorder were streamed in Panopto software (https://www.panopto.com/), Figure 2. One
researcher was present in each individual experiment to instruct and record the participant
performance. A period of 10 minutes for setting up and a few minutes for a short introduction
were necessary for informing the participant, reading and signing of the consent agreement
and discussing the experiment. The researcher sets the room temperature and draws each
participant’s attention to minimize the following actions as these affect the signal capture,
namely: blinking, muscle contractions, rotating the head, horizontal eye movements, neck
movements, pressing lips and teeth together in particular during the tasks. The researcher
follows a script to conduct the experiment so that each participant is given the same
information and stimuli. The researcher positioned the participants at the desk and checked
for metallic accessories that could produce electromagnetic interference. Before each task,
participants were asked to start by reading the text which took an average of 10s. Then the
subjects performed the sequence of five tasks previously described. In the breaks between the
tasks, participants played the Mikado game. The participants performed the tasks in a linear
sequence as the objective of the study is the measurement of brain activation of designers
through a sequence of tasks that gradually expand the design solution space from a problemsolving to basic and then open design tasks.
Figure 2: Audio, video and screen streaming in Panopto.
Electromagnetic interference of the room was checked for frequencies below 60Hz. The
experiments took place between March and July of 2017 and June and September 2018 in a
room with the necessary conditions for the experiment, such as natural lighting from above
sufficient for performing experiments between 9:00 and 15:00 and no electromagnetic
interference. The experiments took between 34 to 67 minutes. The EEG activity was recorded
using a portable 14-channel system Emotiv Epoc+. Electrodes are arranged according to the
10-10 I.S, Figure 3.
3
Figure 3: Emotiv Epoc+ Electrodes (10-10 I.S.) and experiment setup.
Participants
A total of 29 experiments were conducted with industrial designers. Due to EEG or video
recording issues five experiments were excluded. The analysis then proceeded based on the
EEG data recorded and processed for each of the 24 remaining experiments, and each of the
14 electrodes used for averaging, for each of the tasks. A z-transform was conducted to
determine outliers. The criteria for excluding participants were based on the evidence of 6 or
more threshold z-score values above 1.96 or below -1.96 and individual measurements above
2.81 or under -2.81. This resulted in a further two experiments being excluded leaving 22.
After the division of the Pow into time deciles (which provides the basis for the temporal
analysis) and based on the evidence of threshold values above two and a half average plus
standard deviation per channel, a further 4 experiments had to be excluded leaving 18.
The analysis is based on the experimental data of 18 industrial designers, aged 25-43 (M =
31.7, SD = 7.3), 10 men (age M = 35.1, SD = 7.2) and 8 women (age M = 27.5, SD = 5.1), all
right-handed. The study was approved by the local ethics committee of the University of
University of Porto. Each participant was reminded to use the bathroom and spit out any gum
before the start of the experiment. The researcher sat each participant at the desk, asking
him/her to untie hair and remove earrings and other metallic accessories, check if they are
using contact lenses as these may cause too much blinking and interfere with data collection.
Time was given to the participants, in particular in Tasks 3 and 4 so they could find a
satisfactory solution. Average time taken per task is as follows: Pretask, 101s, Task1, 90s,
Task2, 97s, Task3, 373s and Task 4, 725s.
Data Processing
For the present analysis, all the EEG segments of the recorded data were used for averaging
throughout the entire tasks, from beginning to end. In order the remove spurious effects such
those produced by eye blinks, jaw muscle contractions and speaking we adopt the blind
source separation (BSS) technique based on canonical correlation analysis for the removal of
muscle artifacts from EEG recordings (De Clercq, et al. 2006, Vergult, et al. 2007) adapted to
remove the short EMG bursts due to articulation of spoken language, attenuating the muscle
contamination on the EEG recordings (Vos, et al. 2010). The fourteen electrodes were
disposed according to 10-10 I.S, with a 256 Hz sampling rate, a low cutoff 0.1 Hz, and a high
cutoff 50 Hz. Data processing includes the removal of DC offset with the IIR procedure, and
BSS.
Data Analysis
We focus on the overall activation per channel, per task, per participant as the study aims to
determine how the results for problem-solving and designing can be distinguished. We
compare absolute values known as transformed power (Pow), and task-related power (TRP).
The Pow is the transformed power, more specifically the mean of the squared values of
microvolts per second (µV/s) for each electrode processed signal per task. This measure tells
us about the amplitude of the signal per channel and per participant magnified to absolute
4
values. We present Pow values on aggregates of participants’ individual results, per total task
and for each task deciles for the temporal analysis. The task-related power (TRP) is typically
calculated taking the resting state as the reference period per individual (Rominger, et al.
2018, Schwab, et al. 2014). We analyzed the EEG recordings of the resting periods prior to
the experiment of some of the participants and their results varied considerably, with some
participants showing signals that can be associated with the state of being nervous and
expectant and their cognitive effort and activity is unknown. As the focus of the study is to
determine how well designing can be distinguished from problem-solving, we take the
problem-solving Task 1 as the reference period for the TRP calculations. Thus, for each
electrode, the following formula was applied taking the mean of the corresponding electrode i,
in Task 1 as the reference period. By subtracting the log-transformed power of the reference
period (Powi, reference) from the activation period (Powi, activation) for each trial j (each
one of the five tasks per participant), according to the formula:
TRPi = log(Powi, activation)j - log(Powi, reference)j
(1)
By doing this, negative values indicate a decrease of task-related power from the reference
(problem-solving Task 1) for the activation period, while positive values express a power
increase (Pfurtscheller, Lopes da Silva, 1999). TRP scores were quantified for total power and
Pow temporal analysis was carried out by dividing each experiment session into deciles per
task (power and activation refer to brain wave amplitude). Data analysis included Pow and
TRP values on individual and aggregate levels using MatLab and open source software.
Analysis and Results
Preliminary results of total task-related power (TRP) across the 18 participants indicate that
the tasks can potentially be distinguished from each other using the TRP values. The open
design Tasks 3 and 4 show higher TRP from the constrained Task 1. The transformed power
(Pow), was calculated for each of the 5 tasks and electrodes. Results between the tasks for the
industrial designers are depicted in Figure 4. Higher activation in the open design Tasks 3 and
4, particularly in the channels of the right occipitotemporal cortex (F8 to O1), translates the
higher conceptual expansion in the problem and solution spaces.
Total Pow Industrial Designers
Total TRP Industrial designers (18)
Pretask
Task1
Task2
2
AF3
Task3
Task4
Pretask
Task1
100
90
80
70
60
50
40
30
20
10
0
AF3
AF4
1.5
1
F3
F3
F4
0.5
0
F7
F7
F8
-0.5
-1
-1.5
FC5
FC6
T7
Task4
AF4
F4
F8
FC6
T7
P8
O1
Task3
FC5
T8
P7
Task2
T8
P7
O2
P8
O1
O2
Figure 4: Task-Related Power (TRP) and Transformed Power (Pow).
To compare the TRP scores we performed an analysis by running a 4x2x7 repeatedmeasurement ANOVA, with the within-subject factors task, hemisphere and electrode. From
the analysis of the 18 participants we found a significant main effect of: task, p=.02, and
hemisphere, p=.02. There was no main effect for electrode, p=.60. A significant interaction
effect between the factors hemisphere and electrode was found: p<.01. In addition, we
conducted pairwise comparisons to check for differences among participants comparing
electrodes, hemisphere and task. The pairwise comparisons revealed that Task 4 differs
significantly from Pretask (p=.02) and Task 2 (p<.01). The transformed power (Pow), was
5
calculated for each of the 5 tasks, electrodes and deciles. To compare the Pow scores we
performed an analysis by running a 5x2x7 repeated-measurement ANOVA, with the withinsubject factors task, hemisphere and electrode. We found a significant main effect of: task,
p<.001, hemisphere, p<.001, and electrode, p<.001. The pairwise comparisons revealed that
Task 4 differs significantly from Task 1(p<.001) and Task 2 (p<.01), and Task 3 differs
significantly from Task 1(p<.01) and Task 2 (p=.01).
Temporal Analysis and Brodmann Areas
For a temporal analysis of the data, each experiment session is divided into ten equal
segments called deciles. The transformed power (Pow) for the constrained Task 1, and the
open design Tasks 3 and 4 across channels per decile is depicted in Figure 5. Problem-solving
Task 1 has increased general activation in deciles one and seven. Task 3 shows increased
general activation in deciles one, four, six, seven and ten. Task 4 shows higher variation of
temporal distributions of activations.
To compare the Pow scores for the deciles we performed an analysis by running a 5x2x7x10
repeated-measurement ANOVA, with the within-subject factors of task, hemisphere,
electrode and decile. From the analysis of the 18 industrial designers we found a significant
main effect of: task, p<.001, hemisphere, p<.001, and electrode, p=.001. A marginally
significant main effect was found for decile, p=.07.
Significant interaction effects were found between the factors: task and hemisphere, p=.01,
task and electrode, p<.001, task and decile, p<.001, and hemisphere and electrode, p<.01.
In addition, we conducted pairwise comparisons for hemisphere, electrode, decile and task.
The pairwise comparisons revealed that Task 4 differs significantly from Task 1 (p<.01) and
Task 2 (p<.01), Task 3 differs significantly from Task 1 (p<.01) and Task 2 (p=.02).
The pairwise comparisons also reveal significant differences between deciles: between the
first two deciles from which it can be inferred that participants are sorting out how to tackle
the tasks request; deciles four and six do not show differences with the others, from which it
can be inferred that a more reflective and incubation stage while maturing thinking about the
task request takes place; the third, fifth, seventh, eighth and ninth deciles differ from the last
one as the refinement of the solutions may differ from searching how to tackle the request.
Statistical analysis indicates significant increased activation of channels placed on the left and
right occipital and dorsolateral cortices in the open design tasks compared to the problemsolving task. These channels and their corresponding Brodmann areas (BA), are represented
across the deciles in Figure 5. In Figure 5, the circles indicate significant differences and the
numerals inside the circles are the Brodmann area number. Brodmann areas refer to unique
regions of the cortex and are associated with particular cognitive activities. Brodmann’s
studies on brain cells’ neuron structure and its cytoarchitectural organization in 52 areas
(1909) have been refined and correlated to various cortical functions and cognitive activities
by measuring blood flow in response to different mental tasks (Glasser, et al. 2016). Multiple
magnetic resonance imaging (MRI) measurements have resulted in an extended map with 97
new areas, besides the 83 areas previously reported (Glasser, et al. 2016) with each discrete
area containing cells with not only similar structure, but also function and connectivity.
Various cognitive functions and connectivity have been identified in studies using fMRI and
positron emission tomography (PET).
From the analysis of the open design Task 3, the time span for deciles is 36s. Single channel
significant activation takes place in three deciles. In the first decile, channel FC6 shows
increased activation of BA 44 whose cognitive functions are associated with inhibition
actions, monitoring actions, goals, expressing emotions, working memory, episodic memory
and object manipulation (Bernal and Altman, 2009). Such increased activation of FC6 takes
place in seven deciles for the open layout design Task 3.
6
The left temporal cortex and secondary visual cortex have differences in the second, third,
fifth, sixth and ninth deciles, with increased activation of BA37 associated with the functions
of monitoring shape, intentions, drawing, episodes, familiarity judgments and visual fixation
(Le, Pardo, Hu, 1998), and BA18, associated with the functions of spatial and emotional
visual processing, on the right hemisphere and visual word form and mental imagery on the
left hemisphere (Waberski, et al., 2008). Evidence for higher activation of the right
dorsolateral prefrontal cortex happens in
Pow In du strial Designers 1 d ecile
Pow In dustrial Design ers 6 decile
the fifth, sixth and tenth deciles in the
open layout task. Single activation of
channels in this region happen in the
third and seventh deciles. No channel
0808 45
shows decreased activation compared to
44 44
44 44
37
the constrained problem-solving Task 1.
3718
18 18
18
From the analysis of the open design
37
18
Task 4, the time span for deciles is 70s.
Pow Ind ustrial Design ers 2 decile
Pow In du strial Designers 7 decile
For each decile of 70s, statistically
significant differences between Task 4
and Task 1 take place in all the deciles.
Channel FC6 increased activation of
08
08
08
corresponding BA 44, whose cognitive
44 44
44
37
37
37 18
37
functions are associated with inhibition
18 18
37
18
actions, monitoring actions, goals,
18
expressing emotions, working memory,
Pow In du strial Designers 3 d ecile
Pow Industrial Design ers 8 d ecile
episodic memory and object
manipulation (Bernal and Altman, 2009)
takes place in all the deciles as well. The
right and left temporal and secondary
08
08
45
visual cortices have differentiating
4444
44
3718
37
21
37 18
21
contributions in the second to the sixth
18 18
18
37
18
37
and in the eighth to the ninth deciles,
18
with increased activation of BA37,
Pow In d u strial Design ers 4 decile
Pow In dustrial Design ers 9 decile
associated with the functions of
monitoring shape, intentions, drawing,
episodes, familiarity judgments and
visual fixation (Le, Pardo, Hu, 1998). As
08
08
47
Task 4 is an open design free-hand
4444
44 44
37
42 37
37
37 18
18
sketching task, drawing activates BA37
18
18 37
18 18
37
37
18
(Le, Pardo, Hu, 1998), and other areas of
18
the secondary visual cortex such as
Po w In dustrial Design ers 5 d ecile
Po w Indu strial Designers 10 d ecile
BA18 associated with the functions of
spatial and emotional visual processing,
on the right hemisphere and visual word
09
09
09
08
form and mental imagery on the left
08
45
45
45
47
45
44 44
hemisphere (Waberski, et al., 2008).
42 37
37
37 18
21
37 18
Evidence for higher activation of the
18
18 18 37
37
37
right dorsolateral prefrontal cortex just
18
takes place in the tenth decile. Spatial
Figure 5: Circles indicate channels that differ from memory, recall and planning among
Task 1 to Task 3 and Task 4 by deciles correlated
other functions attributed to BA09
with their Brodmann areas (numerals inside
(Slotnik, Moo, 2006) connected to
channel AF4, just show increase in
circles).
st
Ta sk 1
th
Ta sk 3
100
AF3
Ta sk 4
Ta sk 1
Ta sk 3
100
AF3
AF4
F3
F4
F4
60
60
40
F8
40
F7
20
20
0
0
FC5
FC6
T7
F8
FC5
FC6
T7
T8
P7
T8
P7
P8
O1
AF4
80
80
F3
F7
Ta sk 4
P8
O2
O1
O2
nd
Ta sk 1
th
Ta sk 3
100
AF3
Task 1
Ta sk 4
AF3
AF4
Ta sk 3
100
80
80
F3
F3
F4
F4
60
60
40
F7
F8
40
F7
20
20
0
0
FC5
FC6
T7
F8
FC5
T8
P7
FC6
T7
T8
P8
O1
P7
P8
O2
O1
O2
rd
Ta sk 1
th
Task 3
100
AF3
Task 4
Ta sk 1
AF3
AF4
Ta sk 3
100
80
AF4
F3
F4
F4
60
60
40
40
F7
F8
20
20
0
0
FC5
FC6
T7
F8
FC5
T8
P7
FC6
T7
T8
P8
O1
P7
P8
O2
O1
O2
th
Ta sk 1
th
Task 3
100
AF3
Task 4
Task 1
AF3
AF4
Ta sk 3
100
80
F4
AF4
F3
F4
60
60
40
F8
40
F7
20
F8
20
0
0
FC5
FC6
T7
T8
P7
FC5
FC6
T7
P8
O1
T8
P7
P8
O2
O1
O2
th
Task 1
AF3
th
Ta sk 3
100
Ta sk 4
Task 1
AF4
AF3
80
Ta sk 3
100
AF4
F4
F3
60
60
40
F8
40
F7
20
20
0
0
FC6
FC5
T8
T7
P8
P7
O1
Ta sk 4
80
F4
F3
F7
Ta sk 4
80
F3
F7
Ta sk 4
80
F3
F7
Ta sk 4
AF4
O2
F8
FC6
FC5
T8
T7
P8
P7
O1
O2
7
activation compared to Task 1 in the tenth decile. No channel shows decreased activation
compared to Task 1. The co-activation of channels of significant differences have two
moments of continuous and increasing engagement before and after the seventh decile.
Discussion and Conclusion
Results from this study demonstrate that EEG is both a practical and relevant technique to
study differences in industrial designers while problem-solving and designing. The results of
the analysis of the EEG data of the 18 participants show differences in the neurophysiological
activations of these industrial designers across tasks and provide initial support for Hypothesis
1: the design neurocognition of industrial designers when problem-solving and designing is
different, particularly in open design tasks, Task 3 and Task 4. Industrial designers show
higher transformed power (Pow) and distinct task-related power (TRP) differences from the
open design Task 3 and Task 4 to the constrained design Task 1. The neurocognitive temporal
distributions of activations are non-uniform, providing initial support for Hypothesis 2:
industrial designers show variation in the Pow between the problem-solving and design tasks,
across the deciles. On a qualitative level the current study shows evidence of a distinct
characteristic of increased Pow and TRP of Task 3 and Task 4. Increased activation is
associated with conceptual expansion (Abrahams, 2019) from which we infer that the design
space inherently expands as well in the designers’ search for the problem and the solution.
Evidence for higher activation of the right dorsolateral prefrontal cortex across in design tasks
(Alexiou, et al. 2009; Kounios and Beeman, 2009) is shown, particularly in the open layout
design Task 3. Evidence from fMRI studies (Alexiou, et al.2009) of a more extensive network
of brain areas in designing than problem-solving can be inferred from these EEG results.
Evidence for higher activation of the right occipitotemporal cortex is consistent for both open
design tasks. We can propose that for open design tasks the co-activation of channels of
significant differences, is consistent for the channels P7, O1, O2, P8 and FC6. In particular for
the open layout design, F4 and F8 also integrate the co-activation of channels of significant
differences, whose associated cognitive functions seem to be relevant for the design of spatial
solutions. Results from the time-related neural responses connected to Brodmann areas’
cognitive functions, contribute to a better understanding of industrial designers’ cognition in
open design tasks. These results can be correlated with previous cognitive studies that explore
similar hypotheses (Jiang, Gero and Yen, 2014).
Further detailed analyses are being carried out to provide a more in-depth and comprehensive
understanding of the neurophysiological differences between the tasks based on the temporal
analysis of frequency bands and their relation to cognitive functions.
Neuroimaging studies (i.e. fMRI, EEG, fNIRS) are more advanced in creative cognition
(Abrahams, 2019; Benedek, Jung, Vartanian, 2018; Gero, 2008; Gero, 2015; Kowatari et al.,
2009; Martindale and Hines, 1975; Vartanian and Goel, 2005; Xue et al., 2018), and visual
creativity, architecture and the arts (see review by Pidgeon, et al. 2016), than in design
research. However, no consensus has been found as results do not converge among studies
due to the different nature of the tasks and focus. Results from creative cognition studies with
focus on insight and divergent thinking problems, may not be particularly central to
understand creativity in the context of designing artifacts for the real world (Goel, 2014).
Consequently, the design neurocognition field emerges as promising to further a better
understanding of the acts of designing across domains and perhaps a more in-depth distinction
of creativity in the mental processes associated with design.
Cognitive studies of designers commenced some 50 years ago (Eastman, 1968) with the bulk
of the studies occurring in the last twenty years. Neuroimaging studies are a new approach to
studying design cognition that have the potential to provide an objective measurement of
brain behavior connected to cognition. The potential contributions of neuroimaging studies of
design cover a large number of areas including studying the effects of: design domains, tasks,
8
teams, tools and experience on design cognition. In particular, neuroimaging studies
contribute to a better understanding of design cognition and have implications for design
education, the development of design support and the management of design.
REFERENCES
Abrahams, A. (2019). The Neuroscience of Creativity, Cambridge, UK: Cambridge
University Press.
Alexiou, K., Zamenopoulos, T., Johnson, J. H., Gilbert S. J. (2009). Exploring the
neurological basis of design cognition using brain imaging: some preliminary results. Design
Studies, 30(6), 623-647.
Benedek,M., Jung, R., Vartanian, O. (2018). The neural bases of creativity and intelligence:
common ground and differences. Neuropsychologia, 118, 1-3.
Bernal, B., Altman, N. (2009). Neural networks of motor and cognitive inhibition are
dissociated between brain hemispheres: an fMRI study. Int J Neurosci. 119, 10, 1848-1880.
Darke, J. (1979). The primary generator and the design process. Design Studies, 1(1), 36-44.
De Clercq, W., Vergult, A., Vanrumste, B., Van Paesschen, W., Van Huffel, S. (2006).
Canonical correlation analysis applied to remove muscle artifacts from the
electroencephalogram. IEEE Transactions on Biomedical Engineering, 53, 2583-2587.
Dickter, C., Kieffaber, P. (2014). EEG Methods for the Psychological Sciences. Sage.
Dorst,K., (2011). The core of ‘design thinking’ and its application. Design Studies, 32(6),
521-532.
Eastman, C. M., (1968). Exploration in the cognitive processes of design, Defense
Documentation Report No. AD 671158, ARPA, Carnegie-Mellon University.
Ericsson, K. A., Simon, A. H. (1984). Protocol Analysis: Verbal reports as data. MIT Press.
Gero, J. S. (ed) (2008). Studying Design Creativity, Key Centre of Design Computing and
Cognition, University of Sydney, Australia.
Gero, J. S. (ed) (2015). Studying Visual and Spatial Reasoning for Design Creativity,
Dordrecht, Netherlands: Springer.
Glasser, M., Coalson, T., Robinson, E., Hacker, C., Harwell, J., Yacoub, E., Ugurbil,
K., Andersson, J., Beckmann, C., Jenkinson, M., Smith, S., Van Essen, D. (2016). A multimodal parcellation of human cerebral cortex. Nature. 536 (7615), 171-178.
Le, T., Pardo, P., Hu, X. “4 T-fMRI Study of Nonspatial Shifting of Selective Attention:
Cerebellar and Parietal Contributions. The American Physiological Society. Vol.79 No.3.
1998:1535-1548.
Goel, V. (2014). Creative brains: designing in the real world. Frontiers in Psychology, 8, 114.
Goel, V., Pirolli, P. (1992). The structure of design problem spaces. Cognitive Science, 16,
395-429.
Goucher-Lambert, K., Moss, J., Cagan ,J. (2017). Inside the mind: Using neuroimaging to
understand moral product preference judgments involving sustainability. ASME Journal of
Mechanical Design, 139(4), 041103-041103-11.
Hinterberger, T., Zlabinger, M., Blaser, K. (2014). Neurophysiological correlates of various
mental perspectives. Frontiers in Human Neuroscience (8): 1-16.
Jiang, H., Gero, J. S., Yen, C. C. (2014). Exploring designing styles using Problem-Solution
indexes,.In J. S. Gero (ed), Design Computing and Cognition'12, Springer, pp. 85-101.
Jones, J. C., Thornley, D. G. (eds) (1963). Conference on Design Methods. Oxford, UK:
Pergamon Press.
Kan, J. W.T., Gero, J. S. (2017). Quantitative Methods for Studying Design Protocols.
Springer.
Kounios, J., Beeman, M. (2009). The Aha! moment: The cognitive neuroscience of insight.
Current Directions in Psychological Science, 18(4), 210-216.
9
Kowatari Y., Lee S.H., Yamamura H., Nagamori Y., Levy P., Yamane S., Yamamoto M.
(2009). Neural Networks Involved in Artistic Creativity. Human Brain Mapping, 30, 16781690.
Kruger, C., Cross, N. (2006). Solution driven versus problem driven design: Strategies and
outcomes. Design Studies, 27(5), 527-548.
Liang, C., Lin, C., Yao, C., Chang, W., Liu, Y., Chen, S. (2017). Visual attention and
association: An electroencephalography study in expert designers. Design Studies, 48, 76-95.
Liu, L., Li, Y., Xiong, Y., Cao, J., Yuan, P. (2018). An EEG study of the relationship between
design problem statements and cognitive behaviors during conceptual design. Artificial
Intelligence for Engineering Design, Analysis and Manufacturing, 32, 351–362.
Liu, L., Nguyen, T., Zeng, Y., Ben Hamza, A. (2016). Identification of Relationships
Between Electroencephalography (EEG) Bands and Design Activities. ASME 2016
International Design Engineering Technical Conferences and Computers and Information in
Engineering Conference. Volume 7: 28th International Conference on Design Theory and
Methodology. Charlotte, North Carolina, USA, August 21–24, 2016.
Martindale, C., Hines, D. (1975). Creativity and cortical activation during creative,
intellectual and EEG feedback tasks. Biological Psychology, 3, 91-100.
Pfurtscheller, G., Lopes da Silva, F. (1999). Event-related EEG/MEG synchronization and
desynchronization: basic principles. Clin. Neurophysiol. 110, 1842-1857.
Pidgeon, L., Grealy, M., Duffy, A., Hay, L., McTeague, C., Vuletic, T., Coyle, D., Gilbert, S.
(2016). Functional neuroimaging of visual creativity: a systematic review and meta-analysis.
Brain and Behavior, 6(10), 1-26.
Purcell, T., Williams, P. & Gero, J. S., Colbron, B. (1993). Fixation effects: Do they exist in
design problem solving? Environment and Planning B, 20, 333-345.
Rominger, C., Papousek, I., Perchtold, C., Weber, B., Weiss, E., Fink, F. (2018). The creative
brain in the figural domain: Distinct patterns of EEG alpha power during idea generation and
idea elaboration. Neuropsychologia, 118, 13-19.
Shealy, T., Hu, M. and Gero, J. S. (2018) Neuro-cognitive differences between brainstorming,
morphologivcal analysis and TRIZ, ASME IDETC.
Schwab, D., Benedek, M., Papousek, I., Weiss, E., Fink, A. (2014). The time-course of EEG
alpha power changes in creative ideation. Frontiers in Human Neuroscience, 8, 1-8.
Slotnick, S., Moo, L. (2006). Prefrontal cortex hemispheric specialization for categorical and
coordinate visual spatial memory. Neuropsychologia. 44, 9, 1560-1568.
Vartanian, O. and Goel, V. (2005). Neural correlates of creative cognition, in C. Martindale,
P. Locher and V. Petrov (eds) Evolutionary and Neurocognitive Approaches to the Arts,
Baywood Publishing Company, 195-206.
Vergult, A., De Clercq, W., Palmini, A., Vanrumste, B., Dupont, P., Van Huffel, S., et al.
(2007). Improving the interpretation of ictal scalp eeg: BSS-cca algorithm for muscle artifact
removal. Epilepsia, 48, 950-958.
Vos, D., Riès, S., Vanderperren, K., Vanrumste, B., Alario, F., Huffel, V., Burle, B. (2010).
Removal of muscle artifacts from EEG recordings of spoken language production.
Neuroinform, 8, 135-150.
Visser, W. (2009). Design: one, but in different forms. Design Studies, 30(3), 187-223.
Waberski, T. Gobbele, R., Lamberty, K., Buchner, H., Marshall, J., Fink, G. (2008). Timing
of visuo-spatial information processing: Electrical source imaging related to line bisection
judgements. Neuropsychologia. 46, 1201–1210.
10