ArticlePDF Available

Work week productivity, visual complexity, and individual environmental sensitivity in three offices of different color interiors

Authors:

Abstract and Figures

This report is the fourth in a series from a large scale study that examines the effects of three office color interiors (white, predominately red, and predominately blue-green) on worker productivity. Matched on relevant variables, participants were assigned to one of three offices and performed simulated office tasks for four consecutive days. Productivity was measured through workers' task performance and task accuracy, taking into account individual differences in environmental sensitivity (i.e., stimulus screening). The findings suggested that the influences of interior colors on worker productivity were dependent upon individuals' stimulus screening ability and time of exposure to interior colors. Implications of office workers' long-term productivity are discussed in relation to issues concerning the visual complexity of interior environments. © 2007 Wiley Periodicals, Inc. Col Res Appl, 32, 130 – 143, 2007
Content may be subject to copyright.
Work Week Productivity, Visual
Complexity, and Individual Environmental
Sensitivity in Three Offices of Different
Color Interiors
Nancy Kwallek,
1
*Kokyung Soon,
1
Carol M. Lewis
2
1
The University of Texas at Austin, School of Architecture, Department of Interior Design, Austin, TX 78712
2
The University of Texas at Austin, School of Social Work, Center for Social Work Research, Austin, TX 78712
Received 26 August 2004; revised 22 May 2006; accepted 26 July 2006
Abstract: This report is the fourth in a series from a
large scale study that examines the effects of three office
color interiors (white, predominately red, and predomi-
nately blue-green) on worker productivity. Matched on
relevant variables, participants were assigned to one of
three offices and performed simulated office tasks for four
consecutive days. Productivity was measured through
workers’ task performance and task accuracy, taking into
account individual differences in environmental sensitivity
(i.e., stimulus screening). The findings suggested that the
influences of interior colors on worker productivity were
dependent upon individuals’ stimulus screening ability
and time of exposure to interior colors. Implications of
office workers’ long-term productivity are discussed
in relation to issues concerning the visual complexity of
interior environments. Ó2007 Wiley Periodicals, Inc. Col Res
Appl, 32, 130 143, 2007; Published online in Wiley InterScience
(www.interscience.wiley.com). DOI 10.1002/col.20298
Key words: interior design; office workers; office inte-
rior; productivity; environmental sensitivity; visual com-
plexity; task performance; task accuracy; work week
INTRODUCTION
The importance of environmental effects on employees’
productivity and morale has been suggested,
1–3
yet very lit-
tle experimental research on the long-term effects of
interior color on workers’ productivity in the office environ-
ment has been reported. In a review of over 200 color stud-
ies, Beach et al.
4
concluded that few scientifically sound
findings can be drawn about the impact of color on human
performance. Evidently, most studies have often lacked the
necessary experimental control and design measures that
could generate more conclusive results. Other problems,
such as poor resemblance of the experimental room settings
to the real office conditions, inadequate measurement of
productivity and various psychological outcomes, and the
lack of long-term assessment, have severely limited the gen-
eralization and application of the color research findings in
real world settings.
In earlier reports on a large scale research project con-
ducted by the first author and associates,
5–7
the effects of
three interior colors on worker mood, job satisfaction, and
cognitive performance were examined over a 4-day work
week relative to individual differences in environmental
sensitivity. As in most published research, standardized
measurements of productivity, including task performance
and task accuracy, have been relatively short in terms of
length and scope. Thus, one of the goals in the research
project was to determine if cumulative effects or trends of
worker productivity exist from one full work day to the
next. This is the fourth report of the project that focuses
on the effects of color schemes on week-long worker
productivity.
Color and Arousability
There is a prevailing perception that some colors are more
arousing than other colors. Specifically, warm colors, such
as red, are assumed to have more arousing effects on human
responses than cool colors, such as green and blue. On
the basis of his work with brain-impaired participants,
*Correspondence to: Nancy Kwallek (e-mail: n.kwallek@mail.utexas.edu).
Contract grant sponsors: Interface Flooring Systems, BASF Corpora-
tion, International Interior Design Association.
V
V
C2007 Wiley Periodicals, Inc.
130 COLOR research and application
Goldstein
8
proposed that red, a warm color, has an ‘expan-
sive’ property. The color red increases human receptiveness
to external stimuli and induces a state of excitation, which
would affect an individual’s emotional and motor res-
ponses. In contrast, he suggested that green, a cool color, has
a ‘contractive’ property, which provokes human with-
drawal from the external stimulation and reduces one’s
receptiveness to external influences. In addition, he sug-
gested that green has a soothing effect on emotion and
performance.
Subsequent experimental studies have generally sup-
ported this notion. Using white, red, and blue illuminations,
Gerard
9
reported that, among his participants, there were
higher rates of palmar conductance and cortical activation
in response to red rather than to blue. In terms of blood
pressure, respiratory, and eye blink, the study recorded an
increase in activity levels when participants were exposed
to red lights and a decrease when exposed to blue lights.
Similarly, Wilson
10
found that participants appeared to be
more aroused when exposed to red slides than to green
slides, as measured by Galvanic Skin Response (GSR).
Using similar methods, Jacob and Hustmyer
11
tested for the
relative arousability of red, green, blue, and yellow on par-
ticipants’ GSR, heart rate, and respiration. They found that
blue was significantly less arousing than both red and
green, but not yellow.
Effects of Color on Human Performance
The different arousal effects of colors may result in
changes in an individual’s cognitive and psychomotor per-
formances. In early studies on this topic, Pressey
12
con-
ducted a series of experiments to identify the impact of
color upon multiple aspects of physiological and mental
responses, including finger rhythmic movement, pressure
estimation, judgments of pleasantness of touch substances,
multiplication, free association, short-term memory of
nonsense syllables, and continuous choice reaction. He
concluded that the degree of color saturation controlled
by light, affects participants’ finger tapping, multiplica-
tion, and continuous reaction performances, whereas the
effects of hue per se could not be detected in all experi-
ments. In a later study by Hammes and Wiggins,
13
high
and low anxious participants were exposed to three differ-
ent color illuminations, including red, blue, and white,
while performing a perceptual motor steadiness task. No
effect of color illumination on motor task performance for
either low or high anxious participants was found.
Similarly, Dubrovner et al.,
14
contrasted the participants’
number of responses and reaction time on the black-and-
white cards and the color cards of the Rorschach Test. They
reported no significant effects of color on cognitive per-
formance or reaction time. In another study by Goodfellow
and Smith,
15
twenty-five female participants performed two
psychomotor tasks in one of five color conditions where a
tabletop in a booth was painted either red, blue, yellow,
green, or gray. No significant differences for either motor
task were found across the five color conditions. In short,
by controlling the color effects through color illumination
or color on the surface of restricted objects, these studies
generally failed to detect the effects of color on human psy-
chological and motor performances.
Later studies have focused on investigating color as an
‘environmental effect’ rather than as an isolated stimu-
lus. Nakshian
16
had forty-eight participants perform nine
tasks consisting of fine motor tasks, psychophysical judg-
ments, and gross motor tasks in three painted, partitioned
color surrounds. He found significant group differences
for two motor tasks, including hand tremor and motor in-
hibition. On both tasks, participants performed better in
the green condition than in the red. In a similar vein,
Ainsworth et al.,
17
conducted an experimental study on
the effects of red, blue-green, and white interior colors on
participants’ mood and performance. However, using the
words typed, number of errors in typing, and a ratio of
errors to words typed to measure work performance, they
found no significant performance difference between the
groups. They suggested that the relative short-term per-
formance measurement might have contributed to the non
significant results.
In Kwallek et al.’s study
18
comparing the effects of red
and blue offices on worker mood and productivity, partici-
pants made significantly more errors on a 20-min typing
task when they switched from one color room to another,
particularly from blue to red. It was concluded that be-
cause red may produce subjective arousal in participants
and that because blue may have the opposite effect, the
participants’ performance may have been negatively
affected due to the sudden shift from a nonarousing envi-
ronment (blue) to a highly arousing one (red).
In another Kwallek et al. study
19
on worker perform-
ance in a white, green, or red office interior, participants
in the white office made significantly more errors than
participants in the red office. Simultaneously, participants
found the white office to be less distracting and more pre-
ferred as an office environment than the red office. One
consideration for the findings is the lack of saturation of
the white office interior compared with the green and red
office interiors. Assuming that the red and green office
interiors were more distracting than the white office inte-
rior because of their higher saturation, participants may
have narrowed their attention to the tasks at hand in try-
ing to cope with the environmental stress. Consequently,
they could focus more on the tasks in the red and green
offices and made fewer errors than in the white office.
The authors further commented that the amount of effort
required to ignore distracting stimuli may have a detri-
mental impact on task performance in the long run.
Another explanation for the findings is that the red
office was more arousing for participants in this environ-
ment. Since red has been considered to be subjectively
and physiologically arousing, the authors contended that
the green and white office interiors did not boost the par-
ticipants’ overall arousal as high as the red office interior
that was required for more optimal performance on this
particular task.
Volume 32, Number 2, April 2007 131
In another large scale study
20
on the effects of nine dif-
ferent hues on short-term worker productivity, the findings
supported the notion that participants performed worse in
the white office interior than in any of the other eight in-
terior colors (red, green, orange, yellow, blue, beige, gray,
and purple). Further analysis of the data by grouping the
colors according to saturation and value revealed that par-
ticipants performed worse in lighter colored offices as
opposed to darker colored offices. This provides some
indications that saturation and value of color are impor-
tant in determining the effects of color on worker
productivity.
Recent studies on environmental effects of color have
taken into account other relevant factors that might mod-
erate the effects of color on performance. In an experi-
ment conducted by Stone,
21
she found that participants’
task performance within two environmental colors signifi-
cantly differed as the level of task difficulty varied. Spe-
cifically, participants in a blue environment performed
better when performing high demand tasks rather than
low demand tasks. In contrast, participants in a red envi-
ronment performed better when they were assigned low
demand tasks than high demand tasks. In addition, she
also found that environmental view, such as the presence
of a scenic picture, might influence the participants’ per-
formance in different interior colors by serving as a com-
parative point to the task demand. For instance, while
decreased performance of participants was expected in the
red (highly stimulating) interior with high demand tasks,
the soothing effects of the scenic picture might mitigate
the negative effects of the high arousing environment and
high demand tasks. Furthermore, she suggested that time
factor seems to play a role in the relationship of environ-
mental colors and performance as well. On the basis of
the results of her short-term tasks, the performance differ-
ences in different experimental conditions were likely to
increase over time.
In a similar vein, Etnier and Hardy
22
recruited thirty
undergraduate participants to perform psychologically and
physically challenging tasks in three monochromic color
rooms. They concluded that the surrounding color had no
direct effect on task performance, although under particu-
lar circumstances, such as individual color preferences,
might moderate the impact of environmental color on
individual performances. This conclusion is supported by
Cockerill and Miller’s study,
23
where the children in the
study performed the Lafayette Grooved-pegboard Test
faster and more accurately while they wore the goggles
with their most preferred color rather than their least pre-
ferred color.
One plausible conclusion derived from these studies of
color in relation to possible moderating effects is that
color has an indirect effect on human performance, and
the strength of the effect depends on the type of stimuli,
performance indicators, and individual preferences of
color. Unfortunately, among all factors investigated, the
degree of individual sensitivity to the exposure of envi-
ronmental stimuli, a fundamental individual difference in
relation to the response to interior colors, has largely been
ignored in the previous literature. Therefore, to better
understand the effects of environmental color on human
performance, this study takes into account both the com-
plexity of external stimuli as well as the individual differ-
ences in responding to environmental stimuli as important
variables in our investigation.
Visual Complexity and Arousability
The extent to which an individual is visually aroused is
dependent upon the type and rate of imagery information
embedded in the environment. Berlyne
24,25
discussed vis-
ually-related environmental stimuli in terms of visual
complexity and how it relates to human visual preferen-
ces. He assumed that preferences for visual complexity
correspond to the arousal state of the individuals. In addi-
tion, he also hypothesized that individuals strive to main-
tain their optimal level of arousal or to restore the optimal
level when the equilibrium between the degree of visual
(external) complexity and individual’s arousal state is
disrupted. That is, individuals who experience higher
arousal level are expected to prefer lower visual complex-
ity to restore their optimal arousal level. The reverse is
true for the individuals who experience lower arousal
level.
An alternative explanation for arousal level related to
visual complexity has been suggested by Walker.
26,27
According to Walker, there is a direct relationship between
the psychological complexity of individuals and their opti-
mal complexity level, i.e., individual’s most preferred level
of visual complexity. Psychological complexity is con-
sidered to be an individual difference that varies according
to a person’s physiological arousal level. Driven by their
psychological complexity, individuals actively search for
levels of visual complexity that correspond to the optimal
arousal level of the current moments. Consequently, the
individual’s preference for the surrounding environment is
determined by the match between one’s arousal level eli-
cited by the external environment and his or her current
psychological complexity, including subjective experience,
behavioral and neurological activities. A subsequent study
by Twiford et al.
28
supported Walker’s notion that individ-
ual differences in response to visual complexity determine
his or her preference to the environment.
By considering color as an environmental cue to human
responses, Ku
¨ller
29–31
examined the effects of interior
color in terms of visual complexity. He created two ex-
perimental rooms, where one room was full of colors and
patterns to represent an ambient surrounding with high
visual complexity, whereas another room was colored in
monochromatic grey to represent an environment with
low visual complexity. He reported that the recorded elec-
troencephalogram (EEG) alpha related to cortical activity
was significantly lower on the participants in the colorful
room than in the sterile grey room, implying higher level
of arousability on the participants in the colored room in
response to visually more complex environments. In addi-
132 COLOR research and application
tion, Ku
¨ller also observed a physical response to counter-
act over stimulation by lowered heart rates among the
participants in the colorful room during the 3-h long
experiment.
In short, Ku
¨ller’s study basically supported Ber-
lyne’s
24,25
assumption of human’s tendency to maintain
equilibrium between visual information rate and physical
arousability. It is possible that visual complexity attributed
to color could be investigated along several dimensions on
the basis of Ku
¨ller’s conclusion. Such dimensions might
be the number of hues involved, the relative amount of
space each hue occupies, the relative spatial position to
each other, and differences in saturation and value. How-
ever, Ku
¨ller was short of addressing the issue of individual
difference in arousal level responding to the complexity of
external stimuli, as proposed by Walker.
26,27
Fortunately,
Mehrabian’s
32,33
studies of trait arousability can effec-
tively fill in the theoretical gap between the impact of inte-
rior color, visual complexity, and individual difference in
processing visual stimuli.
Arousability, Stimulus Screening, and Performance
Mehrabian
32,33
considered arousability as a trait associ-
ated with individual differences in processing environ-
mental stimuli. Closely related to arousability, stimulus
screening is considered a neurological process that oper-
ates in the opposite way of arousability. While trait arous-
ability can be defined ‘by strength of arousal response to
sudden increases in complexity, variation, novelty, and/or
unexpectedness of stimuli’’ (p.3),
33
stimulus screening is
conceptualized as individual innate ability to routinely
block or filter out irrelevant stimuli within one’s sur-
rounding. These habitual levels of stimulus filtering can
be measured in terms of perceptual selectivity in multiple
sensory inputs. Screeners are individuals who are more
capable of screening irrelevant stimuli and thus are more
adept at simplifying information from sensory input. In
contrast, nonscreeners are those less proficient in eliminat-
ing nonessential stimuli and are more likely to be over-
whelmed by the complexity of the perceptual information.
Consequently, under situations of high visual complexity,
screeners should experience faster habituation to a similar
stimulus, whereas nonscreeners are more likely to be
aroused and take a longer time to return to a baseline arousal
level. In measuring stimulus screening, Mehrabian
32
devel-
oped a questionnaire that categorizes individuals into low,
moderate, and high screeners based on a range of scores for
each category on this instrument (refer to the method sec-
tion). Individuals who are more likely to be distracted by
external stimuli are called low screeners and the reverse is
true for high screeners.
In their study of the effects of environmental stimuli on
an individual’s state of arousal and task performances,
Yermolayeva-Tomina
34
found that individuals who were
easily distracted by irrelevant stimuli showed decrements
in performance in the presence of extraneous stimuli,
whereas others experienced an improvement in perform-
ance when extraneous stimuli were introduced. Additional
evidence came from Mehrabian and Russell’s
35
study that
concluded that a decrease in performance by low screen-
ers was reported only when a distracting stimulus was
presented during the performance of a moderate to highly
complex task, suggesting the important role of individual
stimulus screening ability in the study of environmental
factors and human performance.
In the first published report
5
from this large scale study
examining the effects of three office interior colors
(monochromatic white, predominately red, and predomi-
nately blue-green), productivity on short term office tasks
was examined and reported. The findings indicated that
high screeners performed better on a standardized short-
term proofreading task in the red office and poorer in the
blue-green office than low screeners. The reverse was true
for low screeners; they performed better on the short-term
office task in the blue-green office and poorer in the red
office. The results also suggested that interior colors alone
may not have discernible impact on productivity. Interior
colors appear to affect productivity only when individual
stimulus screening ability is taken into account. In light
of the findings on color, arousal, and psychomotor tasks,
the findings related to worker performance were inter-
preted as an extension of the Yerkes–Dodson arousal-per-
formance curve.
36
Briefly, the Yerkes–Dodson principle
proposes that there is a curvilinear relationship between
arousal and performance. As an individual’s level of arousal
increases, so does performance, up to a certain point. After
reaching an optimal level of arousal, any increase in arousal
will lead to decreased performance.
If the red interior is inherently arousing as previous lit-
erature has suggested, then high screeners are less likely
to feel overwhelmed in terms of arousal. As a result,
those individuals may have performed better than low
screeners in the red interior because they are closer to an
optimal level of arousal in terms of productivity than low
screeners. Low screeners in the red interior may overshoot
their point of optimal arousal because of their inability to
screen out irrelevant stimuli, resulting in them becoming
overly excited physiologically and psychologically in the
presence of the red interior. In contrast, if the effect of a
blue-green interior is inherently relaxing, high screeners
may not experience enough stimulation in the presence of
such a color office environment in order reach an optimal
level of arousal for peak performance. However, in a sim-
ilar environment, low screeners may receive just enough
stimulation to function optimally. These findings reaffirm
the notion that individuals may respond differentially to a
particular interior color depending on their individual
characteristics such as stimulus screening ability.
Hypotheses
In terms of task performance, it was predicted that the
interior color, derived from Beach et al.’s
4
hypotheses
concerning optimal combinations of color value and satu-
ration, would only affect performance when individual
Volume 32, Number 2, April 2007 133
stimulus screening ability was taken into account. On the
basis of earlier results found for short term tasks, low
screeners were predicted to have higher performance than
high screeners in the office with the blue-green interior.
In contrast, high screeners were expected to perform bet-
ter than low screeners in the red interior. We also
hypothesized that the pattern of difference in the white in-
terior office to be similar to the red interior office, since
the white color resembles the red color in terms of their
starkness and level of arousal based on previous studies.
In addition, based on the recommendation of previous
studies
17,21
on the possible augmentation of the effects of
interior color over time, we predicted that the perform-
ance differences between the three groups of workers
(high/moderate/low screeners) increases over the 4-day
period.
For task accuracy, we predicted that low screeners
would make fewer errors than high screeners in the blue-
green interior office. The pattern would be reversed in the
red interior and white interior offices. Similar to the task
performances, we hypothesized that the interior color
would exert a long-term impact on the workers’ task ac-
curacy. Specifically, we expected that the predicted pat-
terns of differences between low screeners and high
screeners would increase over the 4-day work week.
METHOD
Subject Screening and Selection
Participants in this study were recruited through a
state-sponsored human resource center, by placement of
advertisements in city newspapers, and through other job
recruitment centers in the city. All prospective partici-
pants were screened on a Friday prior to being selected
and assigned to a particular experimental condition. They
were given a timed typing task, a self-report questionnaire
related to their general psychological and physical condi-
tions, and four questionnaires to complete, including the:
Questionnaire Measure of Stimulus Screening and Arous-
ability (QMSSA)
32
; Ishihara Color Blindness Test
(ICBT)
37
; Eysenck Personality Inventory (EPI)
38
; and,
Jenkins Achievement Striving Activity Scale (JASAS).
39
After screening, participants were eliminated if they:
typed fewer than 40 words per minute with five or more
errors; were diagnosed as color blind by the ICBT; scored
five or more points on the Social Desirability scale of the
EPI (see below for rationale); were reported to be dys-
lexic and experiencing psychological and physical impair-
ments that would interfere with office task performances;
or, indicated prior knowledge of the purpose of the
experiment.
Over 400 participants were scheduled for screening and
only 200 actually appeared. Of those, 120 participants
passed the screening tests and 90 participants (67 women,
74%; 23 men, 26%) actually completed the 4-day work
week experiment with 30 individuals dropping out before
and during the experiment due to cancellations, no shows,
and other circumstances. The mean age of the workers
was 33.2 years (Standard Deviation ¼11.5). Participants
were paid $200 upon completion of the experiment.
Instruments
The Mehrabian’s Questionnaire Measure of Stimulus
Screening and Arousability (QMSSA)
32
contains 40 items
and measures individual differences in automatic screen-
ing of and habituation to irrelevant stimuli in the sur-
rounding. Responses on each item range from 4¼very
strongly disagree to þ4¼very strongly agree with a
score of 0 ¼neither agree or disagree. The participants’
total scores can range from 160 to þ160, with higher
scores denoting higher stimulus screening ability. Exam-
ples of the items include ‘I am not influenced as much as
most people by the weather’’ (positive wording) and ‘I
am strongly moved when many things are happening at
once’ (negative wording). Kuder-Richardson reliability
coefficient of this questionnaire was 0.92 for this sample.
The Ishihara Color Blindness Test (ICBT)
37
was used
to identify color vision deficiency of congenital origin
among potential participants. The test consists of fourteen
plates, each with a circular image consisting of colored
dots as in a pointillist painting. Numerals within the
circles of dots are distinguishable if the individual has
normal color vision. Only the first eleven plates were used
to detect general color deficiency. If <10 plates were read
correctly, then a subject was identified as color blind and
thus was eliminated from participating in the experiment.
The Eysenck Personality Inventory (EPI)
38
is a 57-item,
dichotomously rated (yes or no) questionnaire. The EPI
consists of three scales: Extroversion-Introversion, Neurot-
icism-Stability, and Social Desirability. The Extroversion-
Introversion scale (24 items) measures an individual’s gen-
eral engagement or inhibition with regard to the social sur-
rounding. A sample item of this scale includes ‘Do you
like going out a lot?’ The Neuroticism-Stability scale
(24 items) assesses the general emotional responsiveness
and liability of the subject in relation to internal and exter-
nal stressors. A sample item of this scale includes ‘‘Do
you worry about awful things that might happen?’’ The
Social Desirability scale (9 items) identifies whether par-
ticipants exhibit a ‘desirability response set’’ that might
influence the validity of their responses. A sample item of
this scale includes ‘Do you find it very hard to take no for
an answer?’ The total scores for the first two scales can
range from 0 to 24 by summing the number of yes
responses and were used for checking purposes in the sub-
ject matching procedure. The Social Desirability scale,
however, has a possible range of scores from 0 to 9. A
participant who scored five and higher on this scale sug-
gested that his or her scores might not be valid indicators
of any psychological dimension measured on the basis of
Eysenck and Eysenck’s
38
recommendation. Thus, the par-
ticipant was eliminated from further participation in the
study. The test-retest reliabilities of this inventory were
reported to range from 0.82 to 0.97.
134 COLOR research and application
The Jenkins Achievement Striving Activity Scale
(JASAS)
39
contains seven items and assesses the presence
of achievement-related behaviors and attitudes. The total
possible score for an individual ranges from 7 to 35, with
a 5-point rating scale from 1 to 5 that has different label
responses depending on the questions. Examples of items
include ‘How seriously do you take your work?’’ and
‘How often do you set deadlines or quotas for yourself in
courses or other activities?’’ The scale was originally
derived from an exploratory factor analysis of the Jenkins
Activity Survey,
40
and subsequently replicated for a sec-
ond sample in a confirmatory factor analysis. The sample
reliability coefficient for this scale was 0.79 in this study.
Office Design
Each of the three offices, which was 8 ft 8 in. (2.63 m)
wide, 11 ft 6–1/2 in. (3.52 m) long, and 8 ft (2.44 m)
high, was of a different color interior, but identically fur-
nished (Fig. 1 for the floor plan). Each had an office desk
and return, a posture chair, an occasional chair, a memory
typewriter, a wall clock, and three framed black-printed
generic certificates on the wall parallel to the desk.
Desk accessories included a wooden paper tray and
wooden card file box, a metal tape dispenser, stapler, and
bookends. An identical book (with beige jacket) was
placed in each desk’s bookends. Identical beige tele-
phones, a green leafy plant, a phone message pad, a clear
glass cup holder for pens and pencils, and a clear glass
paper clip holder were also placed on each desk (See pho-
tos in the first report
5
).
Each individual office, including all four walls and the in-
terior of the office door, was painted. The window in each
office was closed off with drywall and painted like the rest of
the wall so that no natural light entered the office to elimi-
nate any fluctuations of natural daylight. The temperature
was controlled between 73 and 758F (22–248C). The acous-
tics were buffered with ‘white noise’’ machines. Small
built-in circulation fans kept a constant air flow and tempera-
ture. Artificial light was measured using an Illumination
Quality Meter. At the center desk area in each office, the
light was identically set at 600–620 lux (60–62 foot-can-
dles). Glass fiber luminance ceiling panels were rearranged
with opaque acoustic ceiling panels in a drop ceiling frame-
work so the light could be measured as identical foot-candles
at the center of the desk work area of each office. Daylight
fluorescent lamps were used in the offices having a color
temperature of 5000 K and a color rendering index of 90.
All paint colors in the experiment are identified using
the Munsell Color Notation (MCN) system. An attempt
was made to use pure pigments for all latex paint colors
without blending in any other adjacent color on the color
wheel. A Munsell Color Notation of five for the color
provides the purest possible color.
To further interpret color notations, the value of a color
precedes the slash mark while the saturation is designated
FIG. 1. Floor plan of the Environmental Design Research Center.
Volume 32, Number 2, April 2007 135
by the number following the slash mark. The value range
of a color is nine steps between pure white and pure black.
In designing the color schemes for value, an attempt was
made to create colors having an equal distance of separa-
tion between the lightest, medium, and darkest values cho-
sen for the three colors for each scheme. Selecting a
darker value than five was not feasible because the pig-
ment of the color neutralized too much in the artificial
light used in the offices. Thus, for the colors coordinated
on the walls in the red and blue-green offices, the values
selected were 9/ for the lightest value, 7/ for the median
value, and 5/ for the darkest value. As the number
increases, the color becomes more saturated. Conversely,
as the number decreases, the color becomes less saturated.
However, due to a color’s wavelength, some colors at the
same value level appear more saturated than other colors.
For example, because red has a longer wavelength, it
appears more saturated than its opposite color (blue-green)
at the same degree of saturation. On the basis of Munsell
color theory, red reaches its maximum saturation at /14 of
the saturation scale, whereas, due to its shorter wave
length, blue-green reaches its maximum saturation at /10.
To generate the most saturated red color in latex paints, /
12 was used. In contrast, the most saturated for the blue-
green was /8. Thus, to create a three-step spread of satura-
tion for red and blue-green from the most to the least satu-
rated, the saturation steps selected were /12, /7, and /2 for
red and /8, /5, and /2 for blue-green.
It is relatively rare to have monochromatic color offices
in the real world setting, except white. Thus, to ensure the
experimental office rooms resemble the variations of hues,
values, and saturations in the real world, complementary
color schemes were developed. Complementary schemes
would also provide the testing and comparison of a pre-
dominately warm with a predominately cool color scheme.
In order for the enclosed environment to be perceived as
spacious and thus pleasant or vice versa, the three office
interiors were painted different color schemes based on
Beach et al.’s
4
recommendation. Beach et al.
4
suggested
that, to create color combinations that would create the
most spacious and pleasant interior over a long period of
time, the largest area of the interior should be a high value
color and low in saturation, the second largest area should
be a medium value color and medium in saturation, and
finally the trim areas (accent) in the interior should be high
in saturation and either high or low in value.
Accordingly, the first office was painted white (2GY
9/.5) on all four walls and door including the desk return
and appointments. The color white chosen was the gov-
ernment standard (No. 595b-27875) and it is the ubiqui-
tous color utilized in most office buildings. It served as
a controlled condition for the three office comparison
(Table I).
The second office was painted with a predicted un-
pleasant color scheme that was expected to hinder low
screeners’ but not high screeners’ performance. The red
color selected for the top two-thirds of the wall area in
this office (including the inside of the office door) was 5R
5/12. The complement of red, blue-green, was selected for
the lower third of the walls, desk, and return (5BG 7/5).
For accent and trim molding, the selected color was
5R 9/2.
The third office was painted with a predicted pleasant
color scheme that was expected to facilitate low screen-
ers’ but not necessary high screeners’ performance. The
complement of the red color, a blue-green (5BG 9/2), was
selected as the predominant color in this office. It was
selected for the top two-thirds of the wall area, including
the inside of the office door. Medium red (5R 7/7) was
selected for the lower third of the wall, desk and return.
For accent on wood and metal accessories, and trim mold-
ing, a strong bluish-green (5BG 5/8) was selected (See
photos of the office interiors in the first report
5
).
EXPERIMENTAL PROCEDURES
Participants who passed the screening procedure and
showed up the week they were assigned were matched
across office conditions based on gender and stimulus
screening ability. Other worker characteristics (i.e., age,
typing speed, personality, and motivation) that might
influence the workers’ task performance were evaluated
to ensure no prior group differences (Table II).
Participants performed a variety of office tasks throughout
a 4-day (Monday through Thursday), 8-h (9:00 a.m. to 5:00
p.m.) work week. During each working day, participants
were permitted two 15-min breaks and a 1-h lunch break.
On any given day, all three workers in each office performed
the same tasks in the same order and time frame. However,
office tasks were organized into two weekly schedules, A
and B, which alternated from one week to the next. Though
the daily schedules within each weekly schedule were identi-
cal, the weekly schedules differed in terms of which daily
schedule applied from 1 day to the next. Two brief office
tasks (filing index cards and catalogue pricing), each totaling
15 min in duration, were performed by the workers at the
beginning of the work day and after breaks and lunch to ac-
climate the workers to the office environment. Additional
TABLE I. Munsell color notations for three office interiors.
Interior
Monochromatic
white office
Predominantly
red office
Predominantly
blue-green office
Upper wall (2/3) 2GY 9/.5 5R 5/12 5BG 9/2
Lower wall (1/3) 2GY 9/.5 5BG 7/5 5R 7/7
Trim/accent 2GY 9/.5 5R 9/2 5BG 5/8
136 COLOR research and application
measures were administered at the end of each day and at
the end of the work week (i.e., Friday). The findings of these
measures, however, have been discussed in the previous
reports
5–7
and are not part of the results for this study.
Measurement of Task Performance and Accuracy
For each work day, office workers were administered a
typing task, a zip code proofreading task, and a text proof-
reading task. The typing task was administered in the morn-
ing and in the afternoon for 75-min periods. It consisted of
typing a manuscript following an original English text. The
original text provided to the office workers was longer than
any of them could finish. They were instructed to perform as
best as they could. The zip code proofreading task consisted
of checking prepared manuscripts of zip codes against the
original manuscripts for errors for 75-min periods each work
day. Similar to the typing task, the original zip code manu-
script was longer than any office worker was capable of
completing. The office workers were told to accurately proof-
read as many as they could complete. Finally, the text proof-
reading task required the office workers to proofread error-
containing manuscripts against the original manuscripts and
to correct the identified errors. However, a significant number
of participants managed to finish this task within the allotted
time, especially toward the end of the work week.
The task performance and task accuracy of these workers
were assessed. The typing performance score of an office
worker was measured as the total number of words typed
during the allotted time. The typing task accuracy was calcu-
lated by summing the total errors, including spelling, tempo-
ral, spatial, and miscellaneous errors. For this report, only the
afternoon typing performance scores and errors were ana-
lyzed. It is believed that the longer exposure of office work-
ers to the office environment in the afternoon session, as
opposed to the morning session, would make afternoon typ-
ing performance a better indicator of the influence of the
office color on worker performance. For the zip code proof-
reading task, performance was measured by the number of
zip codes checked during the allotted time. The task accu-
racy of the office workers was measured by the number of
errors made during zip code proofreading task. Finally, the
performance on the text proofreading task was assessed by
adding the number of words read, and the task accuracy was
determined by the missed correction of errors in the error-
containing manuscripts.
Strategies for Data Analysis
To determine the moderating effect of stimulus screen-
ing ability on the relationship between interior colors and
workers’ productivity and based on previous studies, par-
ticipants were grouped based on one of three levels of
stimulus screening ability. Using the scores derived from
the QMSSA,
32
individuals with scores below 36 were
designated low screeners, individuals with scores between
36 and 0 were designated moderate screeners, and indi-
viduals with scores above 0 were designated high screen-
ers. These categories were used in the matching procedure
to ensure that an equal number (n¼30) of low, moderate,
and high screeners were assigned to each of the three stim-
ulus screening groups. Together with three office color
interiors, the study created nine experimental conditions,
with each condition consisting of about 10 participants.
A preliminary test using multivariate analysis of variance
(MANOVA) was performed prior to the actual data analysis
to ensure that there were no group differences based on age,
baseline performance (typing speed), personality, and
achievement motivation. Next, repeated-measure analyses of
variance (RM-ANOVAs) were performed with office color
interiors (white, predominantly red, and predominantly blue-
TABLE II. Demographic and personality characteristics of matched workers for three offices.
White office Red office Blue-Green office
nRange MSDnRange MSDnRange MSD
Sex
Male 8 7 8
Female 22 23 22
Age 30 20–56 32.03 10.79 30 20–68 33.13 11.73 30 18–59 34.53 12.09
Typed words
per minute 30 41–76 55.20 10.70 30 41–86 57.26 11.00 30 40–86 58.20 11.92
Stimulus screening
ability scores
a
30 87 to þ65 14.40 34.76 30 98 to þ119 16.43 45.47 30 83 to þ91 15.63 38.57
Achievement striving
Activity scores
b
30 7–32 24.96 4.64 30 20–35 25.33 3.81 30 14–31 25.16 3.96
EPI Extroversion-
introversion scores
c
30 1–16 10.66 3.33 30 2–17 11.10 3.95 30 2–19 11.20 3.40
EPI Neuroticism-
stability scores
d
30 1–24 8.20 5.12 30 0–20 7.46 5.50 30 1–17 8.10 4.65
EPI Social
desirability scores
e
30 0–4 2.30 1.26 30 0–4 2.43 1.13 30 1–4 2.40 .85
a
Test norms are considered a mean ¼24 and standard deviation ¼39.
32
b
For a sample of 1035 college students, the mean was 23.96 for males and 24.44 for females.
39
c
Test norms are those considered between a raw score of 10 and 14.
38
d
Test norms are those considered between a raw score of 7 and 12.
38
e
Individuals scoring >4 were not selected for the study.
38
Volume 32, Number 2, April 2007 137
green) and stimulus screening ability (low, moderate, and
high) as the independent variables. Six groups of dependent
variables were analyzed separately in the model to test for
the overall effects, including: typing performance; zip code
proofreading performance; text proofreading performance;
typing errors; zip code proofreading errors; and, text proof-
reading errors over four consecutive work days. According
to the hypotheses, we expected that there were significant
three-way interactions (interior color stimulus screening
ability day) in all six RM-ANOVAs. The significant alpha
level on the RM-ANOVA was set at P0.05.
For post-hoc analysis, preplanned independent-sample T-
tests were used to describe the patterns of interaction
between color schemes and stimulus screening ability on the
performance and error measures at the beginning (day 1) and
the end (day 4) of the experiment. However, the statistical
power of the T-tests to reject the null hypothesis (i.e., there
is no difference between two comparison groups) was attenu-
ated considering the relatively small number of participants
in each experimental condition (n¼10). Instead, to com-
pensate for the power attenuation resulting from the small
number of participants, one-tailed T-tests were used in the
analysis. Since RM-ANOVAs for the overall test set the sig-
nificant level of probability at P0.05, the risk of commit-
ting Type I error (i.e., saying that there is a group difference
when in reality there is none) in the one-tailed T-tests would
be reduced. T-tests that showed probability between 0.05 and
0.08 were reported as well since they could imply important
significant differences if the statistical power increased.
RESULTS
Overall, multivariate analysis of variance test suggested
that there were no pre-experimental group differences in
terms of age, baseline performance, and individual per-
sonality and achievement motivation that may influence
workers’ performance across three interior colors, F(2,87)
¼0.19, non significant (hereafter ns; Table II). Thus, the
matching procedure for subject randomization was consid-
ered successful.
Task Performance
In terms of typing performance, a significant interaction
effect of office color interior and stimulus screening over
the 4 week days was found, F(12,209) ¼2.11, P<0.05.
On day 1, the post-hoc T-test suggested that high screen-
ers performed similarly compared with low screeners,
t(1,15) ¼1.11, ns, in the white office. Likewise, in the
red office, high screeners performed similarly compared
with low screeners, t(1,18) ¼0.233, ns, and moderate
screeners, t(1,17) ¼1.37, ns. In contrast, low screeners
seemed to perform better than high screeners in the blue-
green office, t(1,21) ¼1.45, P¼0.08 (Fig. 2).
On day 4, however, the T-test suggested that high
screeners performed better than low screeners, t(1,15) ¼
1.63, P¼0.06, in the white office. In the red office, the
result suggested that high screeners performed better than
moderate screeners, t(1,17) ¼1.48, P¼0.08. In contrast,
low screeners did not perform better than high screeners
in the blue-green office, t(1,21) ¼0.28, ns.
In terms of zip code proofreading performance, there was
no significant interaction effect of office color interiors and
stimulus screening over the 4 week days, F(12,209) ¼1.56,
ns. However, a significant interaction effect was found by
comparingonlyday1andday4,F(4,81) ¼2.80, P<0.05.
On day 1, in the white office, high screeners performed simi-
larlytolowscreeners,t(1,15) ¼0.27, ns, and moderate
screeners, t(1,20) ¼0.90, ns. In the red office, high screeners
also did not perform better than low screeners, t(1,18) ¼
1.18, ns. However, in the blue-green office, moderate screen-
ers performed better than both low screeners, t(1,16) ¼2.00,
P<0.05, and high screeners, t(1,17) ¼1.81, P<0.05.
Low screeners also performed better than high screeners,
t(1,21) ¼1.79, P<0.05, in the blue-green office (Fig. 3).
On day 4, in the white office, the result suggested that
high screeners performed better than moderate screeners,
t(1,20) ¼1.46, P¼0.08. In the red office, the t-test sug-
gested that high screeners did better than low screeners,
t(1,18) ¼1.44, P¼0.08. In addition, in the blue-green
office, moderate screeners were shown to perform better
than both low screeners, t(1,16) ¼2.03, P<0.05, and
high screeners, t(1,17) ¼1.81, P<0.05. However, low
FIG. 2. Typing performances of workers in three offices by stimulus screening ability over four work days. [Color figure
can be viewed in the online issue, which is available at www.interscience.wiley.com.]
138 COLOR research and application
screeners did not perform better than high screeners,
t(1,21) ¼0.48, ns, in the blue-green office.
In terms of text proofreading, repeated-measure analy-
sis of variance suggested that there was a significant inter-
action effect of office color interiors and stimulus screen-
ing over the 4 week days, F(4,81) ¼2.20 (Roy’s largest
root), P¼0.08. On day 1, high screeners did not perform
better than low screeners in both the white office, t(1,15)
¼0.00, ns, nor in the red office, t(1,18) ¼0.46, ns. How-
ever, in the blue-green office, moderate screeners had bet-
ter performance than both low screeners, t(1,16) ¼1.51,
P¼0.08, and high screeners, t(1,17) ¼1.45, P<0.05
(Fig. 4).
On day 4, high screeners again did not perform better
than low screeners in both the white office, t(1,15) ¼
0.87, ns, nor in the red office, t(1,18) ¼0.46, ns. In con-
trast, in the blue-green office, moderate screeners still
performed better than high screeners, t(1,16) ¼1.45, P¼
0.08, but not low screeners, t(1,16) ¼0.41, ns.
To summarize, partly consistent with our hypothesis,
while different screeners had similar levels of perform-
ance in typing and zip code proofreading tasks at the start
of the work week (day 1), high screeners had consistently
outperformed either low or moderate screeners toward the
end of work week (day 4) in both the white and predomi-
nantly red office interiors. Performances of moderate and
low screeners were indistinguishable in most cases. In
contrast, although high screeners had the worst perform-
ances in typing and zip code proofreading in the predomi-
nantly blue-green office interior at the beginning of the
work week, their performance was comparable to the low
screeners by the end of work week. Contrary to our hypo-
thesis, moderate screeners, surprisingly, stood out to be
the best performers in the predominantly blue-green office
interior on typing, zip code proofreading, and text proof-
reading tasks at the end of the work week, statistically
and/or numerically. Moreover, we did not find significant
differences on the text proofreading task as hypothesized,
possibly due to the ceiling effect (i.e., a significant por-
tion of the participants were able to finish the task on
time), and only the overall pattern of differences between
different screeners in the predominantly blue-green inte-
rior was consistent with the hypothesis.
Task Accuracy
In terms of typing accuracy, there was no overall signifi-
cant interaction between office color interior and stimulus
screening ability over the 4-day period, F(12,209) ¼0.74,
ns. On day 1, however, the t-test suggested that low screen-
FIG. 4. Text proofreading performances of workers in three offices by stimulus screening ability over four work days.
[Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
FIG. 3. Zipcode proofreading performances of workers in three offices by stimulus screening ability over four work days.
[Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
Volume 32, Number 2, April 2007 139
ers made significantly more errors than high screeners in the
blue-green office, t(1,21) ¼1.76, P<0.05. No significant
interaction effect of office color interior and stimulus screen-
ing ability was found in either the white or the red office.
On day 4, there was no significant group difference found in
any of the three interior colors (Fig. 5).
In terms of zip code proofreading accuracy, there was
also no overall significant color-screening ability inter-
action over the 4 work days, F(12,209) ¼0.815, ns.
However, there was a 4-day average significant interac-
tion between stimulus screening ability and office color
interior, F(4,81) ¼2.90, P<0.05. On day 1, high
screeners made fewer errors than both low screeners,
t(1,15) ¼1.74, P¼0.05, and moderate screeners,
t(1,20) ¼2.00, P<0.05, in the white office. This pat-
tern persisted on day 4, as high screeners again made
fewer errors than low screeners, t(1,15) ¼3.43, P<
0.01, and moderate screeners, t(1,20) ¼2.39, P<0.05,
in the white office (Fig. 6). However, no significant
group difference was found in the red and blue-green
offices.
In terms of text proofreading accuracy, no overall signifi-
cant interaction effect of stimulus screening ability and color
interior was found, F(4,81) ¼1.23, ns.Therewasnosigni-
cant difference found in any of the offices (Fig. 7).
In brief, except for zip code proofreading task, no over-
all pattern of difference (color-stimulus screening-day
interaction) was detected in task accuracy. This can be
due to the fact that, on average, the number of errors
made by the participants was small and it prevented us
from identifying significant group difference in task accu-
racy. The color-stimulus screening interaction was only
reported in the white office interior in terms of task accu-
racy. In general, low and moderate screeners committed
more errors in both typing and zip code proofreading
tasks toward the end of the work week, compared with
high screeners. However, contrary to the hypothesis, this
pattern was not found in either the predominantly red or
blue-green office interior.
DISCUSSION
In comparison to the results found for short-term produc-
tivity tasks in the first report of this study,
5
the results
found for the longer-term tasks were both similar to and
different from the previous short-term results. Congruent
with earlier results, the impact of interior color on worker
performance was dependent upon individuals’ stimulus
screening ability. However, based on the typing and zip
code proofreading performances, at least part of the pat-
FIG. 6. Zipcode proofreading errors of workers (converse of accuracy) in three offices by stimulus screening ability over
four work days. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
FIG. 5. Typing errors of workers (converse of accuracy) in three offices by stimulus screening ability over four work days.
[Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
140 COLOR research and application
tern of the color-stimulus screening interaction differed
from our hypotheses. In contrasting low and high screen-
ers (and sometimes moderate) in the predominantly blue-
green and predominantly red office interiors, we had pre-
dicted that low screeners would perform poorer than high
screeners in the predominantly red office interior, while
the reverse would be true in the predominantly blue-green
office interior at the end of the work week. However, the
hypothesis about the performance of low and high screen-
ers in the predominantly red office interior was only par-
tially supported. Overall, by the end of the experiment,
high screeners did better in typing and zip code proof-
reading tasks than low (and moderate) screeners in the
predominantly red office interior. In addition, the perform-
ances of low and high screeners in the white office inte-
rior were expected to be similar to the predominantly red
office interior. The results suggested that the hypothesis
was supported, as the pattern of differences in the white
office interior closely resembled the predominantly red
office interior.
The result also indicated that the final performances of
three levels of screeners were significantly different from
the initial performances. In the white and predominantly
red office interiors, there was no initial difference
between the screener groups in terms of typing and zip
code proofreading performances. However, the high
screeners eventually outperformed low and sometimes
moderate screeners by the end of the experiment. This
result seems to support the notion that interior color, as
an environmental stimulus, has an enduring and differen-
tial effect on workers’ performance based on individual
environmental sensitivity. In other words, interior color
can ultimately affect a worker’s overall productivity in
the long run depending on the duration of exposure. It
also addresses the limitations posed by earlier color and
worker performance studies, such as Ainsworth et al.’s,
17
that long-term assessment is needed to detect the impact
of environmental color on human performances.
However, the performance difference between low and
high screeners (and moderate) in the predominantly blue-
green office interior did not hold up to expectation. Spe-
cifically, there was no difference found between low and
high screeners in both typing and zip code proofreading
performances by the end of the work week. Interestingly,
moderate screeners appeared to have the best performance
in all tasks among the three stimulus screening groups.
This may be understood through the possible confounding
effect of visual complexity in the interior space of the ex-
perimental room. Although both the predominantly red
and predominantly pale blue-green office interiors had
one predominate color scheme, the spaces introduced a
complimentary (secondary) color in the interior (see the
explanations of the office design in the Method section),
which added a dimension of visual complexity because of
the additional hues, values, and saturations.
On the basis of Ku
¨ller’s
30,31
study as well as the
Yerkes-Dodson
36
principle, the predominantly blue-green
office interior might appear to have more contrasting
aspects that provide different levels of optimal arousal to
the workers than for which was initially designed and
anticipated. With the blue-green wall color, a more no-
ticeable contrast was created in the predominantly pale
blue-green interior with the more saturated blue-green
trim and medium saturated red color of the lower third of
the wall space. Contrary to the predominantly blue-green
office interior, in the predominantly red office interior, the
pale pink trim and the bottom medium blue-green color
were overwhelmed by the more saturated red walls. This
design made the predominantly red interior appear to be
less intense than expected. Therefore, the predominantly
blue-green interior might be considered visually more
complex than expected because the contrasts are more no-
ticeable. This amount of visual complexity appeared to be
most conducive to the moderate screeners’ optimum per-
formance. On the other hand, it may appear to be too
much stimulation for low screeners but not enough stimu-
lation for high screeners.
The sporadic significant results found in task accuracy
were interesting in many ways. In the first report,
5
errors
in the performance of short-term standardized clerical
tasks were not associated with neither interior color nor
stimulus screening for both the first and fourth days on
FIG. 7. Text proofreading errors of workers (converse of accuracy) in three offices by stimulus screening ability over four
work days. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
Volume 32, Number 2, April 2007 141
which participants were assessed. For this study, individ-
ual stimulus screening ability matters only in the white
office interior, and the effect was small. The finding was
reasonable since the small number of errors made by the
workers does not have enough statistical power to detect
the significant group differences. This finding might also
suggest that task performance and task accuracy were two
different aspects of productivity and were affected differ-
ently by the interior colors. Specifically, the predomi-
nantly red and predominantly blue-green interiors only
affect workers’ task performances but not their task accu-
racy after stimulus screening is taken into consideration.
Future studies should further explore the effect of stimu-
lus screening on task accuracy under different color
interiors.
Also, it is puzzling that the findings of the text proof-
reading task are not consistent or even contradictory to the
other the tasks. There are two factors that may have con-
tributed to the inconsistency. First, a preliminary analysis
of the performance on the text proofreading task indicated
that a significant portion of the office workers managed to
finish proofreading the manuscripts within the allotted
time throughout the four work days. This phenomenon
might have created a ‘ceiling effect’’ that limits the range
of participants’ performance and subsequently the results
of the task. Second, the text proofreading performance and
accuracy relied heavily on individual English language
proficiency levels compared with the other two tasks.
Unfortunately, the authors might have overlooked the lan-
guage factor of the participants during the screening pro-
cess and were unable to control for the effects of individ-
ual language proficiency on the text proofreading task.
Finally, one intriguing issue related to the time effect is
that changes in the performance differences between
workers in different conditions during the work week
were significant. This suggests that both the lengths of
time for the task and for the study are important moderat-
ing factors in determining the effects of environmental
colors on relevant variables such as task performance.
Considering the Yerkes-Dodson’s arousal-performance
curve,
36
the impact of interior colors on low and high
screeners may change over time due to familiarity and
habituation effects. Specifically, while both low and high
screeners were initially aroused by more stimulating col-
ors (e.g., red) at a similar level, high screeners more
quickly adapted to the novel environmental stimuli and
experienced faster habituation of the heightened state of
arousal than low screeners. Assuming that the predomi-
nantly red interior color is inherently distracting to office
occupants, the habituation effect may explain why high
screeners eventually outperformed low (and sometimes
moderate) screeners at the end of the work week. This
finding supports the speculation from the previous studies
that time factor should be considered in the study of inte-
rior colors and human performances.
Beyond considerations concerning the period of time
over which tasks are performed, the complexity of the
environment, and individuals’ sensitivity toward the envi-
ronment, studies of the environmental color and its impact
on individuals in the workplace are severely limited when
relying on basic clerical performance measures (e.g., typ-
ing and proofreading) alone. This study is no exception.
Future studies should assess other skills, such as creativity
and problem solving, that are more relevant to the
demands of 21st century offices. This study also did not
investigate the effects of stress and color preference, two
other important factors of individual differences, that may
intervene worker task performance and task accuracy.
Last, but not least, computer technology has been a much
more integrated component of offices nowadays than it
was when this study was first designed and implemented.
Further studies should place computer technology as the
central issue in the office design as well as the measure-
ment of workers’ performance.
How visual complexity and individual environmental
sensitivity can theoretically be explained in terms of the
enduring effects of environmental color has yet to be
adequately addressed. One possible conclusion derived
from these series of studies is that the interior environ-
ment, particularly interior color, has an impact on the
cognitive and affective functioning of individuals, which
is mediated by individual environmental sensitivity and
possibly visual complexity. However, the outcome meas-
ures in the current study only contribute a small piece of
the puzzle. In addition to task performance measures, neu-
ropsychological and physiological measures might be
explored and utilized to evaluate the impact of interior
environment on individuals in a variety of settings.
ACKNOWLEDGMENTS
We thank Harold Woodson and Maria Morrissey for their
assistance in producing this manuscript.
1. Brill M, Margulis TS, Konar E. Using Office Design to Increase Pro-
ductivity, Vol.1: Grand Rapids. Michigan: Westinghouse Furniture
Systems; 1984.
2. Brill M, Margulis TS, Konar E. Using Office Design to Increase Pro-
ductivity, Vol. 2: Grand Rapids. Michigan: Westinghouse Furniture
Systems; 1984.
3. Oldham GR, Fried Y. Employee reactions to workspace characteris-
tics. J Appl Psychol 1987;72:75–80.
4. Beach LR, Wise BK, Wise JA. The human factors of color in envi-
ronmental design: Critical review. NASA technical report NCC 22-
404, NASA/Ames Research Center, Moffett Field, CA, 1987.
5. Kwallek N, Woodson H, Lewis CM, Sales C. Impact of three interior
color schemes on worker mood and performance relative to individual
environmental sensitivity. Color Res Appl 1997;22:121–132.
6. Kwallek N, Soon K. Perception and sensitivity to colour and office
space characteristics after a work week. In: Nieves JL, Herna
´ndez-
Andre
´s J, editors. The 10th Congress of the International Colour
Association. AIC; 2005. p 159–162. Proceedings of AIC, Granada,
Spain, May 8–13.
7. Kwallek N, Soon K, Woodson H, Alexander JL. Effect of color
schemes and environmental sensitivity on job satisfaction and per-
ceived performance. Percept Mot Skills 2005;101:473–486.
8. Goldstein K. Some experimental observations concerning the influen-
ces of colour on the function of the organism. Occup Ther 1942;21:
147–151.
142 COLOR research and application
9. Gerard RM. Differential effects of colored lights on psychophysio-
logical functions. Unpublished doctoral dissertation, University of
California, Los Angeles, 1958.
10. Wilson GD. Arousal properties of red versus green. Percept Mot
Skills 1966;23:947–949.
11. Jacobs KW, Hustmyer FE. Effects of four psychological primary col-
ors on GSR, heart rate, and respiration rate. Percept Mot Skills
1974;38:763–766.
12. Pressey SL. The influence of color upon mental and motor effi-
ciency. Am J Psychol 1921;32:326–356.
13. Hammes JA, Wiggin SL. Perceptual-motor steadiness, manifest anxi-
ety, and color illumination. Percept Mot Skills 1962;14:59–61.
14. Dubrovner RJ, Von Lackum WJ, Jost H. A study of the effect of
color on productivity and reaction time in the Rorschach test. J Clin
Psychol 1950;6:331–336.
15. Goodfellow RAH, Smith PC. Effects of environmental color on two
psychomotor tasks. Percept Mot Skills 1973;37:296–298.
16. Nakashian JS. The effect of red and green surroundings on behavior.
J Gen Psychol 1964;70:143–161.
17. Ainsworth RA, Simpson L, Cassell D. Effects of three colors in an
office interior on mood and performance. Percept Mot Skills 1993;
76:235–241.
18. Kwallek N, Lewis CM, Robbins AS. Effects of office interior color on
workers’ mood and productivity. Percept Mot Skills 1988;66:123–128.
19. Kwallek N, Lewis CM. Effects of environmental colour on males
and females: A red or white or green office. Appl Ergon 1990;21:
275–278.
20. Kwallek N, Lewis CM, Lin-Hsiao JWD, Woodson H. Effects of nine
monochromatic office interior colors on clerical tasks and worker
mood. Color Res Appl 1996;21:448–458.
21. Stone NJ. Environmental view and color for a simulated telemarket-
ing task. J Environ Psychol 2003;23:63–78.
22. Etnier JL, Hardy CL. The effects of environmental color. J Sport
Behav 1997;20:299–312.
23. Cockerill IM, Miller BP. Childrens’ color preferences and motor
skill performance with variation in environmental color. Percept Mot
Skills 1983;56:845–846.
24. Berlyne DE. Conflict, Arousal, and Curiosity. New York: McGraw-
Hill; 1960.
25. Berlyne DE. Complexity and incongruity variables as determinants
of exploratory choice and evaluative ratings. Can J Psychol 1963;17:
274–290.
26. Walker EL. Psychological complexity and preference: A hedgehog
theory of behavior. In: Berlyne DE, Madsen KB, editors. Pleasure,
Reward, Preference: Their Nature, Determinants, and Role in Behav-
ior. New York: Academic Press; 1973.
27. Walker EL. Psychological Complexity and Preference: A Hedgehog
Theory of Behavior. Monterey, CA: Brooks/Cole; 1980.
28. Twiford JR, Haude RH, Sterns HL. Effects of induced arousal on
preference for visual complexity. Percept Mot Skills 1978;46:1155–
1158.
29. Ku
¨ller P. The use of space: Some physiological and philosophical
aspects. In: Korosec-Serfaty P, editor. Appropriation of Space. CIACO;
1976. p 154–163. Proceedings of the 3rd IAPS, Strasbourg, France, June
21–25.
30. Ku
¨ller R. Physiological and psychological effects of illumination and
colour in the interior environment. J Lighting Vis Environ 1986;10:
33–37.
31. Ku
¨ller R, Mikellides B. Simulated studies of color, arousal, and
comfort. In: Marans RW, Stokols D, editors. Environmental Simula-
tion: Research and Policy Issues. New York: Plenum; 1993. p 163–
190.
32. Mehrabian A. Manual for the Questionnaire Measure of Stimulus
Screening and Arousability. Los Angeles, CA: UCLA; 1976.
33. Mehrabian A. Theory and evidence bearing on a scale of trait arous-
ability. Curr Psychol 1995;14:3–28.
34. Yermolayena-Tomina LB. Concentration of attention and strength of
the nervous system. In: Gray JA, editor. Pavlov’s Typology. Oxford:
Pergamon; 1964.
35. Mehrabian A, Russell JA. An Approach to Environmental Psychol-
ogy. Cambridge, Massachusetts: MIT Press; 1974.
36. Yerkes RM, Dodson JD. The relation of strength of stimulus to ra-
pidity of habit-formation. J Comp Neurol Psychol 1908;18:459–482.
37. Ishihara S. Ishihara’s tests for colour-blindness, Report No. 24, Plate
edition. Kanehara, Tokyo, Japan, 1993.
38. Eysenck HJ, Eysenck BG. Manual for the Eysenck Personality In-
ventory. San Diego, CA: Educational and Industrial Testing Service;
1968.
39. Helmreich RL, Spence JT, Pred RS. Making it without losing it:
Type A, achievement motivation, and scientific attainment revisited.
Pers Soc Psychol Bull 1988;14:495–504.
40. Jenkins CD, Zyzanski SJ, Rosenman RH. Progress toward validation
of a computer-scored test for the Type A coronary-prone behavior
pattern. Psychosom Med 1971;33:193–202.
Volume 32, Number 2, April 2007 143
... Colour contrast can contribute to complexity when highly noticeable (with high intensity) (Kwallek et al., 2007) and may provide coherence in some regulated levels and compositions (Kaplan & Kaplan, 1989). However, there are no studies regarding the measurement of colour contrast intensity in complex and coherent interiors, possibly due to the complexity of the visual perception process in the human brain when encountering spatial compositions that communicate too many impactful factors. ...
... According to the basic definitions, complexity is defined by variety and contrast, especially more noticeable contrasts, according to Kwallek et al. (2007). However, the outcomes of this study showed that the dark-light contrast (high-noticeable contrast) was the least effective in complex hotel lobbies showing almost the score 0. This outcome demonstrates that there is a very low level of reliance on dark-light contrast when designers tend to enhance complexity. ...
Article
Full-text available
To understand the relationships between cognitive variables like visual complexity, coherence, and colour contrast in interior spaces, direct numerical analysis is crucial. Conventional approaches are limited due to the brain's struggle to process visual information without cognitive manipulation. However, advancements in artificial intelligence enable direct examinations. This study used a convolutional neural network to assess the intensity of colour contrasts in images of 5-star hotel lobby interiors with high levels of coherence and complexity. The results indicated that visually complex lobby interiors have less warm-cool and dark-light contrast but more pronounced complementary contrast than coherent lobby interiors. Additionally, a negative correlation was identified between complementary and warm-cool contrasts across all perspectives. These results underscore the potential influence of specific colour contrast types on the cognitive experience of interiority.
Chapter
Colour plays a vital role in modern architectural design, particularly in healthcare and wellness environments. This section provides an insight into how colour can significantly improve the patient experience in healthcare facilities by influencing their mood, perception and overall wellbeing. Research has shown that specific colours can evoke specific emotions and antipathies, such as yellow’s ability to make people feel hopeful, so environmental colours in healthcare-related buildings are crucial in reducing patient stress, enhancing well-being and creating a comfortable environment. In addition, colour plays an important role in the usability of medical equipment and environmental aesthetics.
Article
Full-text available
This systematic literature review paper explores the use of extended reality (XR) technology for smart built environments (SBE) and particularly for smart lighting systems design. Smart lighting is a novel concept that has emerged over a decade now and is being used and tested in commercial and industrial built environments. We used PRISMA methodology to review 270 research papers published from 1968 to 2023. Following a discussion of historical advances and key modeling techniques, a description of lighting simulation in the context of extended reality and smart built environment is given, followed by a discussion of the current trends and challenges.
Thesis
Full-text available
The study was about motivational practices and teachers’ performance in private secondary schools in Mbarara District, Uganda. The study aimed at; establishing the relationship between administrative support and teachers‟ performance. The cross-sectional and descriptive survey designs were adopted on a sample of 214 comprising Board of governors, PTA Chair persons, head teachers and teachers from selected private secondary schools. Data analysis involved use of descriptive statistics and of Pearson correlation coefficient to determine the relationship between motivational practices and teachers‟ performance. Descriptive results revealed that administrative support had vital role on teachers’ performance. The findings also showed that there was a statistically positive relationship between administrative support and teachers performance. Thus, it was recommended that administrators should provide a conducive work environment to teachers leading to good performance.
Article
Full-text available
In Experiment I patterns differing in irregularity, amount of material, or incongruity were projected in turn for either 0.5, 1.3, or 4 sec. S decided which of each pair to look at again. Typically, the more irregular (MI) pattern was selected for a short initial exposure and the less irregular (LI) for the longer one. In Experiment II patterns were presented singly and S rated each on a scale of "pleasingness" or "interestingness." MI patterns were more interesting, but LI patterns more pleasing. Differences are related to perceptual curiosity and diversive exploratory behavior.
Chapter
Color psychology is applied all the time in our everyday environments, in the form of paint and graphics both outdoors and indoors, and also in the form of artificially lit spaces. Are the effects of color on the human organism so extreme that our physical and mental hygiene depends on them? Could lack of color in our internal and external spaces not only cause ugliness and boredom but also be responsible for physiological discomfort or stress reactions? The aim of the present chapter is to review the research on color in three areas that have caused considerable confusion during the last 50 years. Most of this research has been carried out in simulated situations, some of which seem to have been barren of any contact with reality.
Article
60 males and 60 females between 6 and 11 yr. of age performed a modified Grooved-pegboard Test while wearing goggles with eyepieces of varying hues. Initially, subjects identified their most-preferred and least-preferred colours. Analysis showed that subjects made significantly fewer errors and took significantly less time while working under their preferred colours. There were significant performance differences between age groups but not between sexes.
Article
Preferences for visual stimuli varying in level of complexity were obtained from 58 male undergraduate students. Subjects in an induced-arousal group showed preferences that were monotonically related to complexity level, while subjects in a non-aroused control group expressed highest preferences for intermediate levels of complexity. The data conform most closely with the psychological complexity notion of Walker (1964, 1973).