Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Contents lists available at ScienceDirect
Building and Environment
journal homepage: www.elsevier.com/locate/buildenv
Access to daylight and view in an office improves cognitive performance and
satisfaction and reduces eyestrain: A controlled crossover study
Anja Jamrozik
a,∗
, Nicholas Clements
a,b
, Syed Shabih Hasan
a,c
, Jie Zhao
a,b,c
, Rongpeng Zhang
a,b,c
,
Carolina Campanella
a,b,c
, Vivian Loftness
d
, Paige Porter
a,c
, Shaun Ly
a,c
, Selena Wang
a,e
,
Brent Bauer
a,b
a
Well Living Lab, Inc., Rochester, MN, USA
b
General Internal Medicine, Mayo Clinic, Rochester, MN, USA
c
Delos Labs, Delos Living LLC, New York, NY, USA
d
School of Architecture, Carnegie Mellon University, Pittsburgh, PA, USA
e
Psychology, The Ohio State University, Columbus, OH, USA
ARTICLE INFO
Keywords:
View quality
Natural light
Window access
Cognitive function
Performance
Office satisfaction
ABSTRACT
Windows provide access to daylight and view, both of which have been linked to positive outcomes for occu-
pants, including improved satisfaction, well-being, and performance. However, window access can also cause
discomfort and eyestrain from glare. This controlled crossover study tested the occupant impacts of two modern
shading systems designed to provide daylight and view while minimizing glare: windows with manually-con-
trolled motorized mesh shades (Mesh Shades) and windows with automatic tinting (Dynamic Tint). Ten parti-
cipants spent fourteen weeks working in a living lab in which three conditions were non-consecutively repeated
for two-week periods: Mesh Shades, Dynamic Tint, and a baseline condition lacking daylight and view (Blackout
Shades). Participants' cognitive function performance, satisfaction, and eyestrain in the baseline Blackout Shades
condition were compared to the same measures in the Mesh Shades and Dynamic Tint conditions. Two aspects of
cognitive function performance—Working Memory and Inhibition—improved in both the Mesh Shades and
Dynamic Tint conditions. Satisfaction with light as well as with the overall environment improved in both the
Mesh Shades and Dynamic Tint conditions. Eyestrain symptoms were reduced in both the Mesh Shades and
Dynamic Tint conditions. There were no statistical differences between settings with Dynamic Tint and mo-
torized Mesh Shades on measures of cognitive function performance, satisfaction, or eyestrain symptoms. This
research demonstrates that providing access to daylight and view in an office environment using modern shading
methods can improve occupants’cognitive performance and satisfaction while reducing eyestrain.
1. Introduction
People spend more than 90% of their time indoors [1–3]. For adults,
much of their waking life is spent at work. The average worker in the
OECD (Organisation for Economic Co-operation and Development)
countries now works 36.6 h per week on their main job [4]. The
workplace environment can improve or detract from occupants' sa-
tisfaction, performance, and well-being [5–7], so it is important to
design workplaces that support, rather than worsen, workers’experi-
ence.
Daylight and views provided by window access are two factors
important to office occupants [8], as demonstrated by multiple studies
conducted in the 1980's and 1990's. In a 1983 study of nearly 500 office
workers in New Zealand and England, 99% of respondents said that
offices should have windows [9]. In a 1989 study of 59 American
university participants' window preferences and factors that influence
those preferences, participants reported their window choices were
based on being able to access a view outside and sunlight, an improved
mood, and improved performance [10]. Window access can also predict
workers' satisfaction with lighting and other aspects of the environment
[6,11,12], and, in some cases, job satisfaction [11]. Office workers
believe that daylight is better than other light sources [9,13,14] and
that they do their best work in natural light conditions [13]. In the late
1990's and continuing into the 2000's, studies demonstrated that
availability of daylight can have a positive impact on occupants' health
outcomes [15–19], and exposure to daylight may be particularly
https://doi.org/10.1016/j.buildenv.2019.106379
Received 1 March 2019; Received in revised form 6 August 2019; Accepted 25 August 2019
∗
Corresponding author. 221 1st Ave SW, Rochester, MN, 55902, USA.
E-mail address: anja.jamrozik@delos.com (A. Jamrozik).
Building and Environment 165 (2019) 106379
Available online 26 August 2019
0360-1323/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
T
important early in the day [16,17].
Daylight impacts are also important to employers. In case studies of
differently-sized companies from across the United States, Romm and
Browning reported that companies moving their workers to buildings
designed to prioritize daylight reported lower absenteeism, increased
productivity, and fewer mistakes [20]. In a 1998 study of 100 diverse
employees in a company in Southern Europe, workers in spaces with
more daylight penetration were more satisfied with their jobs, reported
better well-being, and indicated they were less likely to quit their jobs
[21].
Beyond the benefits of daylighting, there is growing evidence of the
importance of providing office occupants views outside. In a study of
2500 office workers in the U.K., 89% reported that having a view out of
the office was very important [14]. In a 2000 study of 1800 Danish
office workers, study participants preferred workplaces near windows
and cited “the possibility to look out”as the most positive aspect of
having a window [22]. Despite dramatic changes to the office en-
vironment and in how work is completed since the mid-1900's, re-
searchers continue to identify links between views and positive impacts
on occupants, including their health, well-being, and attention span
[23–27].
The effects of employees having access to view are also important to
employers. In a 2003 study of 100 American call center customer ser-
vice representatives, workers with better views showed improved pro-
ductivity, as measured by faster call processing times [28]. In a 2002
study of 141 American office workers, those with windowed offices
reported spending more time working than workers in offices without
windows [29]. In the study of 100 Southern European workers refer-
enced earlier, a view of natural elements was found to reduce the effects
of stress on workers’intention to quit [21].
While daylight and views have significant benefits, window access
can also have unintended side effects. Being too close to a window can
create problems with glare and radiant heat gain or loss [30,31]. In a
1982 study of 235 Canadian office workers, workers in buildings with
large glazed areas (68% of office wall area) reported more eye strain
and had higher absenteeism compared to workers in buildings with
smaller glazed areas (11% of office wall area) [32]. The study of 1800
Danish office workers referenced earlier found that the most negative
aspects of having access to a window were glare problems and over-
heating [22]. While access to daylight and views can reduce the risk of
eyestrain and headache, modern studies demonstrate that day lit win-
dows can create a strong contrast with the walls surrounding them,
potentially increasing eyestrain and headaches [18].
Window shading controls have been developed to reap the benefits
of daylight and view availability while minimizing discomforts asso-
ciated with excessive sunlight. Window shades and blinds come with
several levels of control: fully manual (e.g., blinds operated by pulling a
cord), motorized (e.g., roller shades operated by pushing a button), and
automatically controlled (operated automatically based on time of day
and/or weather conditions). Window shades move only up and down,
while window blinds can move up and down and can be opened, tilted,
and closed.
When they are available, office occupants can use blinds and shades
to avoid excessive sunlight, glare, and overheating [33]. A 2013 review
of research on shade and blind use found that most occupants do not
change the shade position more than once a week or once a month, and
some do not change the shade position at all [34]. The main reason
occupants close shades and blinds is to prevent glare [34]. As observed
in the 1970's and 1980's, in northern hemisphere office buildings with
manually controlled blinds, people are more likely to lower blinds on
the southern facade (80%) than on the northern façade (50%) [35].
Once occupants bring blinds down to avoid direct sunlight, they are
unlikely to change their position again [35–37].
Modern offices provide motorized and automated options for shade
control. A pilot study of 8 Austrian office workers in 2006 found that
motorized blinds were used three times more frequently by occupants
than fully manual blinds [38]. However, if motorized blinds and shades
are fully automatic, they are often disabled [39], or overridden by oc-
cupants [40,41]. For example, in a 2014 study of 40 office buildings in
the Netherlands, most of the installed automatically-controlled blinds
were disabled by occupants [39]. Because automatic motorized shading
that fully satisfies occupants has yet to be developed [34], manually-
controlled motorized mesh shades were considered as one of the state-
of-the-art shading technologies to be tested in the current research.
A newer technology for view and daylight control is electrochromic
glass, in which thin film coatings of electrochromic materials are
electronically controlled to achieve various levels of window tinting.
These thin films typically consist of metal oxides whose material
properties, like light transmittance, can be changed when voltage is
applied. This allows window systems equipped with electrochromic tint
to dynamically manipulate the light transmittance and solar heat gain
coefficient to reduce glare and improve thermal control. Electrochromic
windows can tint following predefined operational schedules or they
can tint automatically depending on the time of day and outdoor
weather conditions. If occupant preferences differ from these schedules,
occupants can override automation when desired [41]. Less research is
available on people's use and adoption of electrochromic glass as
compared to shades, but initial findings are promising [42–45].
Therefore, automated electrochromic glass with occupant override was
also tested in this study.
The research was conducted in a living lab, a facility in which study
participants occupy a simulated real-world environment for an ex-
tended period [46,47]. This method provides a well-controlled and
naturalistic environment that supports typical occupant behavior, al-
lows for environmental monitoring, and permits occupants' unfolding
reactions to the environment to be studied. Using the living lab para-
digm, we tested whether manually-controlled motorized mesh shades
(Mesh Shades condition) and automatically-controlled but able to be
overridden electrochromic glass tint (Dynamic Tint condition) experi-
mental conditions would improve office occupants’performance, sa-
tisfaction, and minimize discomfort, as compared to a baseline condi-
tion with no access to daylight and view, achieved by covering windows
with blackout shades. The study was novel in combining positive fea-
tures of lab and field studies: participants were in a carefully-controlled
environment, and conditions were varied to allow for the discovery of
causal links to behavioral outcomes. During the study, participants
spent their days completing their regular work tasks in the lab, allowing
for their true reactions to changes in their work environment to be
measured using validated cognitive performance tests and ques-
tionnaires.
The experimental design, predictions, and data analysis plan were
all pre-registered on the Open Science Framework (https://osf.io/
hfv58/?view_only=3e8d2aae48c34b4ca7a00dd353bb4431).
2. Materials and methods
2.1. Overview of study design
This study was conducted at the Well Living Lab facility set up as an
open office [47]. Employees of Mayo Clinic were recruited to partici-
pate in the study, which was conducted over a fourteen-week period.
Lab participants spent their workdays in the office space where they
completed their regular work tasks. After a two-week acclimation
period, each experimental condition (Blackout Shades, Mesh Shades,
Dynamic Tint) lasted two weeks and was repeated, with condition order
randomized to reduce order effects, see Table 1. The other environ-
mental conditions (temperature, electrical lighting, ventilation con-
trols) were kept consistent throughout the study.
Motorized roller blackout shades (Mermet Blackout-White, visual
light transmittance (T
V
) of 0%, Lutron Electronics Co., Inc.) fully cov-
ered windows in the Baseline condition, eliminating natural light pe-
netration and view access, and were open and inoperable in Mesh
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
2
Shades and Dynamic Tint conditions.
A motorized roller mesh shade with an openness factor of
1.7 ± 0.75% and T
V
of 7.6 ± 1.6% (E Screen - THEIA™, White/Pearl,
Lutron Electronics Co., Inc.) was selected based on building location
and orientation (see Fig. 1 for detailed specifications). On the first day
of the first week of the Mesh Shades conditions, the mesh shades were
opened completely before participants arrived and participants could
adjust positions of the window shades to their preferred height using
wall-mounted controls installed next to each window. Mesh shades
were kept open and inoperable during Blackout Shade and Dynamic
Tint conditions.
Shade position data were collected every minute and whenever the
shade position was requested to be changed and are reported as percent
open (e.g. fully open is 100%, closed is 0%). Shade position time series
data were analyzed for average openness, per day counts of shade po-
sition changes, and time spent in each quintile of shade openness.
Each window was also equipped with an electrochromic tinting
system (View, Inc.) capable of being set to four tint levels, ranging from
clear glass with high T
V
(58%) in tint level 1 to nearly completely
darkened glass with low T
V
(1%) in tint level 4 (see Fig. 1 for specifi-
cations). In the Dynamic Tint condition, tinting transition times and tint
level for each window were determined with an algorithm (View, Inc.),
based partially on photosensor data from the lab's rooftop sensor and
façade orientation. During the Dynamic Tint condition, participants
could interrupt the automation and set the desired tint level of each
window using an application installed on tablets in the office space.
Application-based tint changes made by occupants were programmed
to persist for 3 h before returning to the level determined by the au-
tomation algorithm. Tint level 2 had a T
V
of 40% and was used in the
Mesh Shades condition to simulate a typical existing office with tinted
double or triple pane windows with low solar heat gain. In the Blackout
Shades condition, all windows were set to tint level 4 to reduce heat
flux through windows. Data downloaded from the electrochromic tint
control unit described how and when occupants changed the tint level.
In Fig. 2 are photographs of a window in the Mesh Shades and Dynamic
Tint conditions.
During the experiment, participants completed surveys at the end of
each workday to assess environmental and work-related satisfaction,
eyestrain, and other outcomes.
Daily objective measures of participants’cognitive function perfor-
mance were collected every afternoon. Cognitive function performance
encompasses a wide range of human abilities, such as spatial cognition,
insight problem solving, and deductive reasoning. The current research
measured executive functions, the general-purpose control mechanisms
that regulate cognition and make it possible for us to plan and execute
goal-directed actions [48,49], since these abilities are necessary for a
wide range of work and everyday activities [49]. There is general
agreement that at least three different executive functions are corre-
lated but separable from one another [48,50]: working memory up-
dating, inhibition, and task switching. Working memory updating refers
to our ability to hold, manipulate, and update information in memory.
Inhibition refers to our ability to deliberately inhibit automatic re-
sponses when it is necessary to do so. Task switching refers to our
ability to shift between different tasks or operations. All three were
measured using peer-reviewed, validated electronic tasks from cogni-
tive psychology and neuroscience.
Participants’survey responses and cognitive function performance
in the Blackout Shades condition were compared to their responses and
performance in the Dynamic Tint and Mesh Shades conditions. The
hypothesis was that access to daylight and view in the Mesh Shades and
Dynamic Tint conditions would improve performance and satisfaction
and reduce eyestrain symptoms as compared to no access (i.e., the
Blackout Shades/Baseline condition).
An additional group of participants from the same work unit as the
lab participants remained in their regular office and completed cogni-
tive testing at the same time as the lab participants each day. We
compared cognitive performance of participants in the lab and in the
regular office to understand whether lab participants’performance was
representative of the larger population from which they were drawn.
2.2. Participants
Twenty participants consented to participate in the study. Ten (3
male, 7 female, M
age
= 49.30, SD
age
= 8.45) moved to work in the
living lab for the duration of the study, and ten (1 male, 9 female,
M
age
= 52.70, SD
age
= 11.48) stayed in the work unit's regular office
space. The study was approved by the Mayo Clinic Institutional Review
Table 1
Daylight and view experimental conditions and study sequence.
Condition Dates Week # Blackout Shade Mesh Shade Window Tint
Acclimation April 6, 2017–6/17/2017 1, 2 –– –
Blackout Shades (Baseline) 6/18/2017–January 7, 2017
8/13/2017–8/26/2017
3, 4
11, 12
Closed Open Level 4
Dynamic Tint February 7, 2017–7/15/2017
7/30/2017–December 8, 2017
5, 6
9, 10
Open Open Automated level 1–4; Able to be overridden
by occupants
Mesh Shades 7/16/2017–7/29/2017
8/27/2017–September 9, 2017
7, 8
13, 14
Open Open on 1st day; Controllable by
occupants
Level 2
Fig. 1. Technical specifications of the roller shades and the electrochromic glass
used in the study.
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
3
Board.
The sample size of the current study was smaller than that of most
studies on the effects of daylight and view reviewed, and similar to the
size of exploratory studies such as [38]. As described in section 2.3, the
sample size of the lab cohort was dictated by lab space size and desk
arrangement to minimize differences in experience between partici-
pants.
Participants were recruited from a work unit whose management
approved recruitment of participants for a research study, and who
could conduct work duties from a remote location. Additional inclusion
criteria were: adults 18–65 years old, able to provide informed consent,
minimum of one year experience performing current work duties, able
to work 20–40 h per week. Exclusion criteria were: severe hearing loss
requiring the use of hearing aids, severe vision problems, sensitivity to
light triggering headaches or epileptic episodes, cognitive disabilities
severely interfering with typical office work, physical disability prohi-
biting a variety of typical office tasks, clinical depression or severe
mood disorders, drug (illegal or prescription narcotic), or alcohol de-
pendency, women who were pregnant or intended to become pregnant
during the study, and employees who were in a performance im-
provement plan or spent more than 50% of time working outside of the
office.
At the start of the study, lab participants' chronotype was assessed
using the Morning-Eveningness Questionnaire (MEQ) [51,52], a well-
validated self-report questionnaire developed to measure whether an
individual's circadian rhythm produces peak alertness in the morning,
evening, or in between. Five participants had an intermediate chron-
otype, four a moderate morning chronotype, and one a definite
morning chronotype.
2.3. Office configuration
Participants’normal work environment was an office with in-
dividual cubicles in a mid-rise office building. Cubicles typically had a
double desk surface and storage for paperwork and files. Access to
daylight and view varied, with some workers close to windows and
others toward the interior of the building. Window shading devices and
interior lighting varied throughout the space, but most windows had
fully manual window blinds. Ten participants who stayed in their office
experienced this environment throughout the study.
Ten participants were moved to a 124 m
2
(19.1 m length, 6.5 m
width, 2.6 m height) open office made up of three experimental mod-
ules at the Well Living Lab, see Fig. 2a. Windows along the north
(W01–W02) and east (W03–W08) facades had dimensions of
2.39 m × 1.75 m and 2.39 m × 4.27 m, respectively.
To minimize differences in experience between study participants
(e.g., access to daylight/view, likelihood of glare), all desks were ar-
ranged along the east facade of the office space, with the edge of each
desk 1.22 m away from the inside surface of the windows. This desk
arrangement dictated the sample size of the lab cohort (N= 10).
As much as possible, participants’individual work environment at
the lab (desks, cubicles, storage, work equipment) was modeled on their
normal work environment. Desks were separated by 1.35 m tall parti-
tions and 1.58 m tall partitions separated desks from the rest of the
office space to provide visual privacy. Desks were an adjustable sit-
stand model, and participants were each assigned a desk location.
Participants were provided computers, telephones, printers, fax ma-
chines, and internet to complete their work.
2.4. Lighting environment and lighting measures
Correlated color temperature (CCT) tunable LED lighting
(0.2 × 0.6 m, Rubik Tunable White 3-cell, Acuity Brands Lighting, Inc.),
grouped to simulate typical troffer lights used in office environments,
provided electrical lighting in the office, see Fig. 3c. LED light levels
and CCT were adjusted using an illuminance spectrophotometer (CL-
500A, Konica Minolta Sensing Americas, Inc.) and set so the desks re-
ceived, on average, 300 lux of 4000 K electric light in the horizontal
plane during the study, see Supplementary materials S1.
A spatial assessment of horizontal illuminance, CCT, and spectral
power distribution of natural light under the range of expected ex-
perimental conditions was conducted using the sampling points in
Fig. 3c, see Supplementary materials S1.
Illuminance (Lux1000, Wovyn LLC, N= 10, 1-min sampling in-
terval) sensors and CCT sensors (Color Lux1000, Wovyn LLC, N= 10,
1-min sampling interval), able to measure illuminance and estimate
CCT were installed on desk surfaces (see Fig. 3b) to measure temporal
variability in horizontal illuminance and CCT. Illuminance (N= 16, 10-
min sampling interval) and CCT (N= 8, 10-min sampling interval)
sensors were installed on window surfaces facing outside at the middle
of the windows (height of 1.6 m) to measure vertical illuminance and
CCT. See Supplementary materials S2 for details on sensor calibration.
Fig. 2. Photographs of a window seen from desk 5. The top panel shows the shading range within the Mesh Shades condition. In this condition, window tint was set to
level 2 to simulate conventional low solar heat gain office glass. Quintiles of shade openness are pictured along with the measured percent of the workday spent in
each quintile in this condition (averaged across windows). The lower panel shows the tint range within the Dynamic Tint condition. The tint levels are pictured along
with the measured percent of the workday spent in each tint level in this condition (averaged across windows). (For interpretation of the references to color in this
figure legend, the reader is referred to the Web version of this article.)
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
4
Fig. 3. (a) Layout of experimental modules, desks, and windows; (b) layout of illuminance, correlated color temperature (CCT), air temperature, and relative
humidity sensors in the office space; and (c) electric lighting design and spatial lighting assessment sampling points. (For interpretation of the references to color in
this figure legend, the reader is referred to the Web version of this article.)
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
5
2.5. Other environmental conditions and monitoring
The air temperature set point was initially 22.8 °C (73 °F), and
participants could request adjustments by contacting the participant
coordinator. One set point adjustment (up to 23.9 °C (75 °F)) was made
during the third week of the study (Thursday, 6/21/2017) and kept for
the rest of the study. The relative humidity (RH) set point was 40%. See
Supplementary materials for details on office ventilation (S3), addi-
tional environmental measures (S4), and ratings of view quality (S5).
2.6. Cognitive function performance measures and processing
Participants were each provided a dedicated tablet (iPad, Apple
Inc.) to use for cognitive tests. The tests were delivered through a web
application linked on the tablet, and participants were reminded by
email to take the tests between 1 and 3 p.m.
As described in Section 2.1, three aspects of executive function
performance—working memory updating, inhibition, and task switch-
ing—were measured. Working memory was measured using the Op-
eration span test [53,54], in which participants solve math problems
while remembering sets of letters. Participants are given limited time to
solve each math problem and are asked to solve the problems accu-
rately (at least 80% accurately in the current study). The dependent
measure is the Load Score–-the proportion of memory items (letters)
participants can recall correctly while maintaining good performance
on the math task. Working memory data from days on which a parti-
cipant failed to maintain adequate math performance (16.8% of cases)
were removed from analyses, following standard practice for this task.
The key question tested was whether the Load Score was higher in Mesh
Shades and Dynamic Tint conditions vs. Baseline.
Inhibition was measured using the Stroop test [55], in which par-
ticipants respond to what color words (which are names of colors)
appear in. Sometimes words appear in a congruent color (BLUE written
in blue), sometimes the colors are incongruent (BLUE written in green).
Participants are asked to identify the color as quickly as possible. The
dependent measure is the reaction time difference between correctly
answered Incongruent vs. Congruent trials. While specific values of the
colors (blue, red, green) were chosen to be visible to people who are
color blind, a test at the start of the study additionally identified any
participants who were color blind (none were).
Task switching was measured using the magnitude/parity test
[56,57]. In this test, participants monitor the color of digits (1–4, 6–9)
and, depending on the color of the number, either answer whether a
number is greater versus less than five or whether the number is even
versus odd. Sequential trials were categorized as either as “Stay”or
“Switch”trials. In Stay trials, the trial follows the same kind of trial
(e.g., even vs. odd trial followed by even vs. odd trial). In Switch trials,
the trial type varies from the one that came before it. Participants are
asked to respond as quickly as possible. The dependent measure is the
reaction time difference between correctly answered Switch vs. Stay
trials.
Inhibition and Task Switching reaction times shorter than 200 ms
and longer than 3000 ms were trimmed to remove outliers, and reaction
times were log-transformed to remove skew. For Inhibition, the key
question tested was whether the difference between Incongruent and
Congruent trials was smaller in Mesh Shades and Dynamic Tint con-
ditions vs. Baseline. For Task Switching, the key question tested was
whether the difference between Switch and Stay trials was smaller in
Mesh Shades and Dynamic Tint conditions vs. Baseline.
See Supplementary materials S6 for information on practice trials
given to participants during the experiment acclimation period.
In addition to the tests, the application included questions to assess
factors previously linked to cognitive function: caffeine intake, aerobic
exercise, mindfulness practice, positive and negative affect, and alert-
ness, see Supplementary materials S7 for details.
2.7. Survey design and measures
Lab participants completed surveys about their daily experience at
the end of each workday. Participants were given a set of questions
from an adapted “right now”version [58] of the Cost-effective Open-
Plan Environments (COPE) survey [59] to assess their satisfaction with
the environment, including the environment as a whole, lighting
(overall, light for computer work, light for paper-based tasks), view,
ability to alter physical conditions, and other factors (e.g., air quality,
temperature, noise). The survey also assessed satisfaction with work-
related variables: the department/agency, job satisfaction, and pro-
ductivity. An item based on the COPE wording was added to assess
productivity in isolation. See Supplementary materials S8 for an addi-
tional measure of productivity collected during the study. All ratings
were on a scale from 1 (very dissatisfied) to 7 (very satisfied). Partici-
pants also reported when they experienced glare, and what environ-
mental factors should be improved to support their effectiveness at
work.
Every weekday, participants also completed a Headache & Eyestrain
questionnaire [60] to assess their eye-related symptoms, including eye
fatigue, blurred vision, irritability, and difficulty focusing. All symptom
ratings were on a scale from 1 (none) to 4 (severe).
Once a week, on Fridays, lab participants were also asked about any
sick-building symptoms (SBS) they experienced during the week and
whether these symptoms improved when they were not in the office, as
measured through a SBS symptom checklist [61].
2.8. Behavioral data analysis
To characterize the relationship between experimental condition
and cognitive function performance, as well as survey ratings, linear
mixed-effects analyses were performed using the lme4 (Version 1.1–17)
package in R (Version 3.5.0). Mixed-effects analyses allowed for the
modeling of variation in how individual participants reacted to each
experimental condition. Experimental condition was included as a ca-
tegorical fixed effect with three levels (Blackout Shades, Mesh Shades,
and Dynamic Tint) in models of all outcomes. Models of the Inhibition
and Task Switching measures included the fixed effect of trial type (e.g.,
Switch vs. Stay) and an interaction between experimental condition and
trial type to identify differences in the reaction difference between trial
types under different environmental conditions (e.g., if the difference
between Switch and Stay trials was smaller in the Mesh Shades con-
dition than the Blackout Shades condition). To account for practice or
fatigue effects, changes in cognitive function performance were mod-
eled over the course of the experiment period by including a fixed effect
of week number. For the Task Switching and Inhibition tasks, changes
in cognitive function performance were modeled for each daily ex-
periment session by including a fixed effect of trial number (within the
daily task). Following current best practices to evaluate the significance
of fixed effects of models fit with lme4 [62], p-values were derived
using Satterthwaite approximations for degrees of freedom with the
lmerTest package (Version 3.0–1). Intercepts for the random effect of
participant and by-participant slopes for the effect of experimental
condition and trial type (when applicable) were included in each
model.
Estimates of the marginal and conditional coefficients of determi-
nation were calculated using the MuMIn package in R (Version 1.43.6).
The marginal R
2
estimate represents the variance explained by a
model's fixed effects, and the conditional R
2
estimate represents the
variance explained by a model's fixed and random effects.
In exploratory follow-up analyses, behavioral outcomes in the Mesh
Shades and Dynamic Tint conditions were compared using the lsmeans
package (Version 2.27–2) when testing for a main effect of condition, or
by running the same models using only data from the Mesh Shade and
Dynamic Tint conditions when testing for an interaction between ef-
fects (for the Inhibition and Task Switching measures).
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
6
The effect of weekday and covariates of interest (caffeine intake,
amount of aerobic exercise, amount of mindfulness practice, positive
affect, negative affect, alertness) were separately tested for inclusion in
each of the cognitive function models. In follow-up analyses, the cog-
nitive performance of lab participants was compared to that of control
participants (those who stayed in the work unit's regular workspace),
while accounting for practice and fatigue effects.
3. Results
3.1. Environmental measurement results
3.1.1. Shade and tint control
Per window, there were on average ( ± SD) 1.9 ± 3.2 shade posi-
tion changes/day during the Mesh Shade condition. Per window, tint
overrides were detected on average 0.3 ± 0.6 times per day during the
Tint condition. Fig. 4 demonstrates the by-week shade position changes
and tint overrides per day for each window. There were no consistent
patterns in frequency of shade position change or tint level adjustments,
see Supplementary material S9 for further summaries.
3.1.2. Lighting and other environmental measurements
Desktop illuminance data collected during the Blackout Shades
condition agreed with the initial spectrophotometer-assessed desktop
illuminance measurements, averaging 293 ± 11lx. Relative to base-
line, desk-level horizontal illuminance increased on average by
270 ± 153 lx and 239 ± 153 lx in the Mesh Shades and Dynamic Tint
conditions, respectively. There were no consistent differences in hor-
izontal illuminance between Mesh Shades and Dynamic Tint conditions.
See Supplemental materials S10-14 for additional environmental mea-
sures.
3.2. Behavioral measurement results
3.2.1. Cognitive function performance across daylight and view conditions
Cognitive function performance improved in daylight and view
conditions, though not all three cognitive function measures were im-
pacted to the same degree. Working Memory and Inhibition (our ability
Fig. 4. Summaries of shade position changes per day during work hours (6:00–18:00) on weeks of Mesh Shade conditions (a,b) and tint state overrides per day during
weeks of Dynamic Tint conditions (c,d): (a) Time series of shade position changes per day by week, (b) histogram of position changes per day by window, (c) time
series of tint state overrides per day by week, (d) histogram of tint state overrides per day by window. (For interpretation of the references to color in this figure
legend, the reader is referred to the Web version of this article.)
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
7
to deliberately inhibit automatic responses when it is necessary to do
so) improved reliably in both the Mesh Shades and Dynamic Tint
conditions as compared to Baseline (Blackout Shades), while Task
Switching was not reliably impacted by experimental condition.
For the Working Memory task, the dependent variable was the Load
Score, calculated daily, and the independent variables were experi-
mental condition and week number in experiment. Participants’Load
Score increased slightly over the course of the experiment period, as
seen in the positive effect of week number, see Table 2. Taking this
improvement into account, the Load Score improved reliably in the
Mesh Shades and Dynamic Tint conditions vs. Baseline.
For the Inhibition task, the dependent variable was reaction time for
each correctly-answered trial, and the independent variables were ex-
perimental condition, trial type (Congruent vs. Incongruent), week
number in the experiment, and sequence number (the trial number
within the task). Models included the interaction between trial type and
experimental condition. Across the study, the Inhibition task effect re-
mained reliable: people were slower to respond to Incongruent trials
than Congruent trials. People's overall reaction time decreased slightly
over the course of the experiment period (as seen in the negative effect
of week number) and increased slightly over the course of daily sessions
(as seen in the positive effect of sequence number). Taking these pat-
terns into account, the key measure, the difference between
Incongruent and Congruent trials, decreased reliably in Mesh Shades
and Dynamic Tint conditions, vs. Baseline (as seen in the negative effect
of the interaction between trial type and these conditions), see Table 3.
For the Task Switching task, the dependent variable was reaction
time for each correctly-answered trial, and the independent variables
were experimental condition, trial type (Switch vs. Stay), week number
in the experiment, and sequence number (the trial number within the
task). Models included the interaction between trial type and experi-
mental condition. Across the study, the Task Switching effect remained
reliable: people were slower to complete Switch trials than Stay trials.
People's overall reaction time decreased slightly over the course of the
experiment period and over daily sessions. There were no differences in
the key measure, the difference between Switch and Stay trials, across
the environmental conditions, see Supplementary materials S15. The
effects of the Dynamic Tint and Mesh Shades conditions on cognitive
Table 2
Working Memory Load Score results. The Load Score of lab participants improved re-
liably in Mesh Shades and Dynamic Tint conditions vs. Baseline. B = fixed effect estimate,
SE B = standard error of fixed effect estimate.
Descriptive statistics
Group in lab
Week Condition MSD
3 Blackout 0.646 0.268
4 Blackout 0.702 0.244
11 Blackout 0.702 0.289
12 Blackout 0.724 0.289
7 Shades 0.695 0.292
8 Shades 0.761 0.226
13 Shades 0.777 0.247
14 Shades 0.763 0.306
5 Tint 0.681 0.313
6 Tint 0.681 0.263
9 Tint 0.722 0.268
10 Tint 0.763 0.245
Group in office
Week Condition MSD
3Office 0.763 0.271
4Office 0.686 0.296
11 Office 0.696 0.331
12 Office 0.659 0.306
7Office 0.670 0.293
8Office 0.709 0.280
13 Office 0.681 0.297
14 Office 0.749 0.248
5Office 0.699 0.274
6Office 0.664 0.302
9Office 0.685 0.301
10 Office 0.774 0.264
Model summary - lab participants
B SEB ß SEß t p
Working Memory Load Score
Intercept (Blackout shades) 0.672 0.074 −0.310 0.275 9.124 < .001
Sheer shades 0.033 0.014 0.142 0.058 2.302 0.022
Tint 0.029 0.014 0.119 0.053 2.045 0.046
Week Number 0.008 0.002 0.031 0.007 4.463 < .001
R-square marginal R-square conditional
0.023 0.791
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
8
function performance were comparable and there were no statistical
differences between these conditions, see Supplementary materials S16.
See Supplementary materials S17 for covariate analyses. The pat-
tern of findings described above remained the same after accounting for
all significant covariates. See Supplementary materials S18 for a com-
parison between participant groups (lab group and office group). The
groups did not differ in their cognitive performance, suggesting the lab
group was representative of the larger work group.
3.2.2. Environmental and work-related satisfaction
For each of the satisfaction analyses, the dependent variable was a
daily survey rating, and the independent variable was experimental
condition. People were more satisfied in conditions that provided access
to daylight and view, see Table 4 for model details and S21 for de-
scriptive statistics and model R
2
values. Overall environmental sa-
tisfaction was higher in the motorized Mesh Shades condition and the
Dynamic Tint condition vs. the Baseline with covered windows. Like-
wise, measures of satisfaction with light conditions were higher in Mesh
Shades and Dynamic Tint conditions, including overall light quality,
light on the desk for computer work, and access to view.
People's satisfaction with the ability to alter physical conditions
improved in both the Mesh Shades and Dynamic Tint conditions vs.
Baseline. Satisfaction with the aesthetic appearance of people's work
area improved in the Dynamic Tint condition.
There were no significant differences between conditions for other
aspects of environmental satisfaction or broader work-related satisfac-
tion.
Participants were asked when they experienced glare.
Unsurprisingly given that the experimental space had windows facing
east, participants reported experiencing glare from daylight and on
their computer screens in the morning in the two conditions that pro-
vided access to daylight (Supplementary materials S19).
Table 3
Inhibition results. The difference between Incongruent and Congruent trials for lab participants decreased reliably in Mesh Shades and
Dynamic Tint conditions vs. Baseline. B = fixed effect estimate, SE B = standard error of fixed effect estimate.
Descriptive statistics
Group in lab
Week Condition Congruent trials M(ms) Congruent trials SD (ms) Incongruent trials M(ms) Incongruent trials SD (ms)
3 Blackout 908.903 279.569 1204.710 6381.727
4 Blackout 845.443 319.561 980.329 366.516
11 Blackout 818.493 289.383 911.694 353.281
12 Blackout 822.629 270.542 913.454 424.291
7 Shades 870.568 406.091 956.609 394.945
8 Shades 867.510 301.269 959.809 387.151
13 Shades 846.381 488.597 897.878 328.341
14 Shades 839.505 329.704 934.552 394.771
5 Tint 861.277 271.242 982.328 354.289
6 Tint 902.052 1875.787 973.652 801.791
9 Tint 844.181 338.386 949.537 453.210
10 Tint 869.116 463.503 954.518 468.268
Group in office
Week Condition Congruent trials M(ms) Congruent trials SD (ms) Incongruent trials M(ms) Incongruent trials SD (ms)
3Office 997.174 399.732 1196.842 780.438
4Office 969.673 539.289 1141.563 806.590
11 Office 951.919 2094.288 1043.570 2102.477
12 Office 1204.581 11827.204 1018.202 2580.778
7Office 890.902 402.896 1029.357 511.845
8Office 897.457 830.115 1203.186 5732.074
13 Office 875.233 769.174 1069.301 4871.739
14 Office 879.939 601.103 941.151 452.091
5Office 930.924 401.485 1115.672 1412.053
6Office 1008.644 1944.988 1110.917 706.008
9Office 921.219 1344.721 1138.684 3270.026
10 Office 871.435 697.536 996.316 608.814
Model summary - lab participants
B SEB ß SEß t p
Inhibition reaction time
Intercept (Blackout shades) 6.7700 0.0299 −0.0148 0.0995 226.452 < .001
Sheer shades 0.0128 0.0096 0.0426 0.0319 1.338 0.207
Tint −0.0086 0.0067 −0.0285 0.0223 −1.279 0.224
Incongruent trial (vs. Congruent) 0.1090 0.0107 0.3630 0.0357 10.173 < .001
Sequence Number 0.0001 0.0000 0.0004 0.0001 3.351 0.001
Week Number −0.0087 0.0004 −0.0289 0.0013 −22.688 < .001
Sheer shades * Incongruent trial −0.0207 0.0057 −0.0690 0.0190 −3.635 < .001
Tint * Incongruent trial −0.0137 0.0058 −0.0455 0.0194 −2.347 0.019
Marginal R-squared Conditional R-squared
0.0396 0.1753
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
9
When asked to rank environmental factors that should be improved
to support people's effectiveness at work, over 70% of responses ranked
“window access”first in the baseline Blackout Shades condition. In
contract, in the Mesh Shades and Dynamic Tint conditions, the number
one improvement was divided among noise, temperature, and privacy,
see Supplementary materials S20.
The effects of the Dynamic Tint and Mesh Shades conditions on
ratings of environmental satisfaction were comparable and there were
no statistical differences between these conditions, see Supplementary
materials S21.
Table 4
Environmental and work-related satisfaction ratings from the COPE survey by
environmental condition. B = fixed effect estimate, SE B = standard error of
fixed effect estimate.
B SEB ß SEß t p
Environmental satisfaction
Intercept (Blackout
shades)
4.723 0.422 −0.249 0.282 11.181 < .001
Sheer shades 0.598 0.201 0.399 0.134 2.978 0.016
Tint 0.655 0.242 0.437 0.161 2.711 0.025
Overall light quality satisfaction
Intercept (Blackout
shades)
4.158 0.503 −0.589 0.302 8.262 < .001
Sheer shades 1.544 0.538 0.926 0.323 2.871 0.018
Tint 1.478 0.545 0.887 0.327 2.712 0.024
Light for computer work satisfaction
Intercept (Blackout
shades)
4.422 0.534 −0.503 0.336 8.287 < .001
Sheer shades 1.232 0.521 0.775 0.328 2.364 0.042
Tint 1.111 0.465 0.699 0.293 2.389 0.040
Light on the desk for paper-based tasks satisfaction
Intercept (Blackout
shades)
4.469 0.518 −0.504 0.335 8.632 < .001
Sheer shades 1.225 0.541 0.793 0.350 2.264 0.050
Tint 1.096 0.486 0.710 0.314 2.257 0.050
Access view satisfaction
Intercept (Blackout
shades)
1.711 0.222 −1.070 0.087 7.693 < .001
Sheer shades 4.446 0.438 1.734 0.171 10.152 < .001
Tint 4.545 0.402 1.773 0.157 11.314 < .001
Alter physical conditions satisfaction
Intercept (Blackout
shades)
3.424 0.403 −0.592 0.220 8.487 < .001
Sheer shades 2.052 0.448 1.118 0.244 4.584 0.001
Tint 1.762 0.392 0.960 0.213 4.501 0.001
Aesthetic appearance of your work area satisfaction
Intercept (Blackout
shades)
5.409 0.323 −0.124 0.299 16.756 < .001
Sheer shades 0.191 0.092 0.177 0.085 2.086 0.067
Tint 0.151 0.069 0.140 0.064 2.181 0.048
Cleanliness of your work area satisfaction
Intercept (Blackout
shades)
5.597 0.331 −0.038 0.311 16.929 < .001
Sheer shades 0.072 0.089 0.068 0.084 0.803 0.445
Tint 0.030 0.054 0.029 0.051 0.560 0.581
Air quality satisfaction
Intercept (Blackout
shades)
4.869 0.419 −0.099 0.293 11.615 < .001
Sheer shades 0.172 0.137 0.120 0.096 1.256 0.242
Tint 0.325 0.175 0.227 0.122 1.861 0.096
Air movement work satisfaction
Intercept (Blackout
shades)
4.661 0.449 −0.067 0.295 10.378 < .001
Sheer shades 0.128 0.119 0.084 0.078 1.076 0.295
Tint 0.228 0.208 0.150 0.137 1.095 0.303
Odors in your area satisfaction
Intercept (Blackout
shades)
5.288 0.309 −0.102 0.258 17.105 < .001
Sheer shades 0.103 0.135 0.086 0.112 0.768 0.461
Tint 0.131 0.134 0.109 0.111 0.980 0.352
Temperature in your area satisfaction
Intercept (Blackout
shades)
4.932 0.314 −0.134 0.263 15.725 < .001
Sheer shades 0.196 0.148 0.164 0.124 1.326 0.223
Tint 0.264 0.175 0.222 0.147 1.510 0.168
Visual privacy satisfaction
Intercept (Blackout
shades)
5.373 0.396 0.073 0.225 13.552 < .001
Sheer shades −0.231 0.163 −0.131 0.093 −1.419 0.187
Tint −0.399 0.258 −0.227 0.147 −1.548 0.156
Table 4 (continued)
B SEB ß SEß t p
Acoustic privacy satisfaction
Intercept (Blackout
shades)
4.728 0.478 −0.018 0.287 9.893 < .001
Sheer shades 0.131 0.153 0.078 0.092 0.855 0.418
Tint 0.017 0.122 0.010 0.073 0.141 0.891
Noise from conversations satisfaction
Intercept (Blackout
shades)
4.742 0.391 −0.055 0.293 12.144 < .001
Sheer shades 0.204 0.133 0.153 0.099 1.538 0.166
Tint 0.102 0.170 0.076 0.127 0.599 0.562
Frequency of distraction from other people satisfaction
Intercept (Blackout
shades)
4.824 0.361 −0.062 0.271 13.371 < .001
Sheer shades 0.181 0.136 0.136 0.102 1.333 0.215
Tint 0.120 0.154 0.090 0.115 0.782 0.452
Background noise satisfaction
Intercept (Blackout
shades)
5.033 0.437 −0.098 0.323 11.515 < .001
Sheer shades 0.169 0.093 0.125 0.068 1.822 0.098
Tint 0.109 0.106 0.081 0.079 1.025 0.333
Size of work area satisfaction
Intercept (Blackout
shades)
5.116 0.352 −0.107 0.250 14.521 < .001
Sheer shades 0.226 0.121 0.161 0.086 1.866 0.092
Tint 0.060 0.143 0.043 0.102 0.422 0.682
Degree of enclosure satisfaction
Intercept (Blackout
shades)
4.874 0.565 −0.085 0.302 8.630 < .001
Sheer shades 0.172 0.139 0.092 0.074 1.236 0.247
Tint 0.185 0.115 0.099 0.062 1.609 0.141
Distance from coworkers satisfaction
Intercept (Blackout
shades)
5.258 0.362 0.043 0.204 14.516 < .001
Sheer shades −0.156 0.204 −0.088 0.115 −0.764 0.462
Tint −0.276 0.246 −0.156 0.138 −1.126 0.287
My department/agency is a good place to work
Intercept (Blackout
shades)
4.470 0.488 0.006 0.308 9.150 < .001
Sheer shades −0.151 0.079 −0.095 0.050 −1.900 0.088
Tint −0.048 0.067 −0.031 0.042 −0.725 0.478
Job satisfaction
Intercept (Blackout
shades)
4.184 0.504 −0.045 0.308 8.298 < .001
Sheer shades −0.107 0.067 −0.065 0.041 −1.596 0.140
Tint 0.011 0.059 0.007 0.036 0.193 0.848
Environmental conditions support my personal productivity
Intercept (Blackout
shades)
4.753 0.407 −0.195 0.274 11.682 < .001
Sheer shades 0.435 0.206 0.293 0.139 2.107 0.064
Tint 0.448 0.216 0.302 0.146 2.070 0.071
Personal productivity satisfaction
Intercept (Blackout
shades)
5.142 0.470 −0.086 0.324 10.937 < .001
Sheer shades 0.162 0.119 0.112 0.082 1.358 0.203
Tint 0.288 0.163 0.198 0.113 1.760 0.117
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
10
3.2.3. Eyestrain and additional behavioral outcomes
The Headache & Eyestrain questionnaire was used to assess eye-
strain and related symptoms. For these analyses, the dependent variable
was the survey rating for a given symptom and the independent vari-
able was experimental condition, see S23 for descriptive statistics.
Participants reported less eyestrain, eye fatigue, and difficulty focusing
in the motorized Mesh Shades condition and Dynamic Tint condition vs.
Baseline, see Table 5. They also reported less eye discomfort and less
difficulty concentrating in the Dynamic Tint condition vs. Baseline.
Please see S22 for descriptive statistics.
The effects of the Dynamic Tint and Mesh Shades conditions on self-
reported ratings of eyestrain were comparable, and there were no sta-
tistical differences between these conditions, see S22.
See Supplementary materials S23 for results on positive and nega-
tive affect, alertness, and sick building syndrome.
4. Discussion
The goal of this research was to test the efficacy of two modern
methods to provide office occupants access to daylight and view while
minimizing glare: manually-controlled motorized mesh shades (Mesh
Shades) and windows with electrochromic tint (Dynamic Tint). Office
occupant cognitive function performance, satisfaction, and eyestrain in
Mesh Shades and Dynamic Tint conditions were compared to the same
measures in a baseline condition lacking daylight and view in which
blackout shades covered windows (Blackout Shades).
Compared to baseline, participants' cognitive function performance
improved in both the Mesh Shades and the Dynamic Tint conditions.
However, not all aspects of performance improved. This research
identified that access to daylight and view can improve people's ability
to hold and manipulate items in memory (Working memory) and in-
hibit responses when it is necessary to do so (Inhibition). While we did
not observe a difference in people's ability to switch between tasks with
different constraints (Task switching) across daylight and view condi-
tions, future research could further test this relationship, perhaps using
a larger sample size, longer testing period, more condition repetitions,
or a more sensitive task switching measure.
How might access to daylight and view improve cognitive perfor-
mance? According to the attention restoration theory [63], directed
attention—a concept closely associated to executive function
[64]—becomes fatigued with use, but can be restored through certain
experiences that capture involuntary attention without drawing on di-
rected attention. An example is interacting with nature, such as walking
in nature or viewing nature images, which has been shown to improve
directed attention [63,65,66]. In city environments, like that of parti-
cipants in this study, a view out the window can allow for nature in-
teractions such as observing the sky and clouds, shadows that change
over time, and green space.
Natural environments can also have a restorative effect on stress and
enhance positive emotion [67]. Future research could examine the re-
lationship between office occupants' access to view, cognitive and work
performance, stress and broader well-being. Non-intrusive measures,
such as heart rate variability collected from wearables, or the amount of
face-to-face interaction with co-workers collected from sociometric
badges, could allow for the measurement of occupants’direct beha-
vioral reactions.
Access to either motorized Mesh Shades or Dynamic Tint improved
occupants' satisfaction with light and view, and reduced their perceived
eyestrain symptoms, compared to baseline. In addition to satisfaction
with lighting, the motorized Mesh Shades and Dynamic Tint conditions
improved people's satisfaction with other aspects of the environment
such as aesthetic appearance and the ability to alter physical condi-
tions, as well as the environment overall.
When asked what environmental factors should be improved to
support people's work in the Mesh Shades, Dynamic Tint, and Blackout
Shades conditions, participants overwhelmingly chose “window access”
as the most important improvement in the baseline condition. This is in
line with previous work suggesting that people who do not have access
to a windows are the ones who most desire access [68]. Similarly, there
is a negative relationship between satisfaction and perception of im-
portance of a feature—those occupants who are least satisfied with an
environmental feature may also believe it to be most important [69].
While two modern shading methods were tested in the current re-
search, there are others that remain to be tested in the future, including
automated window shade systems and passive design elements like
external façade shading or lattice. These initial findings highlight the
general cognitive, satisfaction, and well-being benefits that well-shaded
windows can provide to office occupants.
5. Conclusions
Windows provide access to daylight and view but can also increase
discomfort and eyestrain from glare. This study tested the occupants
impacts of two modern shading systems designed to provide daylight
and view while minimizing glare: manually-controlled motorized
shades (Mesh Shades) and windows with automatic tinting (Dynamic
Tint), against a baseline condition with no access to daylight and view
(Baseline/Blackout Shades). The study was conducted in a living lab,
which allowed for participants' workplace environmental conditions to
Table 5
Eyestrain symptoms by environmental condition. B = fixed effect estimate, SE
B = standard error of fixed effect estimate.
Coefficient B SEB ß SEß t p
H&ES Eyestrain
Intercept (Blackout
shades)
1.741 0.179 0.348 0.269 9.713 < .001
Sheer shades −0.400 0.149 −0.601 0.224 −2.686 0.025
Tint −0.383 0.135 −0.576 0.203 −2.841 0.019
H&ES Eye fatigue
Intercept (Blackout
shades)
1.717 0.178 0.283 0.259 9.637 < .001
Sheer shades −0.347 0.153 −0.505 0.223 −2.268 0.049
Tint −0.342 0.133 −0.498 0.193 −2.578 0.030
H&ES Eye discomfort
Intercept (Blackout
shades)
1.601 0.148 0.305 0.238 10.810 < .001
Sheer shades −0.257 0.139 −0.414 0.224 −1.845 0.097
Tint −0.348 0.123 −0.560 0.198 −2.829 0.019
H&ES Blurred vision
Intercept (Blackout
shades)
1.441 0.198 0.218 0.325 7.279 < .001
Sheer shades −0.281 0.125 −0.460 0.206 −2.237 0.0507
Tint −0.277 0.125 −0.455 0.206 −2.212 0.0525
H&ES Irritability
Intercept (Blackout
shades)
1.705 0.222 0.232 0.290 7.682 < .001
Sheer shades −0.243 0.125 −0.317 0.163 −1.945 0.081
Tint −0.195 0.095 −0.254 0.125 −2.042 0.068
H&ES Difficulty focusing
Intercept (Blackout
shades)
1.820 0.212 0.233 0.268 8.566 < .001
Sheer shades −0.323 0.117 −0.408 0.148 −2.764 0.021
Tint −0.362 0.116 −0.457 0.147 −3.120 0.011
H&ES Difficulty concentrating
Intercept (Blackout
shades)
1.757 0.216 0.150 0.272 8.129 < .001
Sheer shades −0.169 0.104 −0.212 0.131 −1.615 0.1396
Tint −0.273 0.093 −0.343 0.117 −2.936 0.0123
H&ES Headache
Intercept (Blackout
shades)
1.212 0.113 0.065 0.247 10.735 < .001
Sheer shades −0.012 0.112 −0.026 0.245 −0.107 0.917
Tint −0.088 0.124 −0.192 0.271 −0.709 0.496
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
11
be carefully controlled and varied while measuring participants’reac-
tions to changes in their work environment using validated cognitive
performance tests and questionnaires.
Cognitive function performance improved with access to daylight
and view. Working Memory improved in the Mesh Shades and Dynamic
Tint conditions vs. Baseline, and Inhibition (our ability to deliberately
inhibit automatic responses when it is necessary to do so) improved in
Mesh Shades and in Dynamic Tint vs. Baseline. Satisfaction with
lighting, with the overall environment, and with other outcomes im-
proved with access to daylight and view. Eyestrain symptoms lessened
with access to daylight and view.
There were no differences in people's performance, satisfaction, or
eyestrain symptoms between settings with motorized Mesh Shades and
Dynamic Tint.
Modern shading methods that provide access to daylight and view
while limiting glare can improve occupants’performance and satisfac-
tion, and reduce eyestrain.
Acknowledgements
We would like to thank the participants for their time and com-
mitment to the study. Thank you to the staffand interns of the Well
Living Lab who kept the facility running. Thank you also to the staffand
leadership teams of Delos Living LLC, the Mayo Clinic, and to the Well
Living Lab leadership team, Joint Steering Committee, Scientific
Advisory Council, and corporate-alliance members.
This research was supported by a donation from View Inc. as cor-
porate-alliance members of the Well Living Lab. View Inc. was not in-
volved in study recruitment, data collection, analysis, or interpretation,
or writing the manuscript.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.buildenv.2019.106379.
References
[1] N.E. Klepeis, W.C. Nelson, W.R. Ott, J.P. Robinson, A.M. Tsang, P. Switzer,
J.V. Behar, S.C. Hern, W.H. Engelmann, The National Human Activity Pattern
Survey (NHAPS): a resource for assessing exposure to environmental pollutants, J.
Expo. Anal. Environ. Epidemiol. (2001), https://doi.org/10.1038/sj.jea.7500165.
[2] J.A. Leech, W.C. Nelson, R.T. Burnett, S. Aaron, M.E. Raizenne, It's about time: a
comparison of Canadian and American time-activity patterns, J. Expo. Anal.
Environ. Epidemiol. 12 (2002) 427–432, https://doi.org/10.1038/sj.jea.7500244.
[3] C. Schweizer, R.D. Edwards, L. Bayer-Oglesby, W.J. Gauderman, V. Ilacqua,
M. Juhani Jantunen, H.K. Lai, M. Nieuwenhuijsen, N. Künzli, Indoor time-micro-
environment-activity patterns in seven regions of Europe, J. Expo. Sci. Environ.
Epidemiol. (2007), https://doi.org/10.1038/sj.jes.7500490.
[4] Average Usual Weekly Hours Worked on the Main Job, (2018) https://stats.oecd.
org/index.aspx?DataSetCode=ANHRS.
[5] Y. Al Horr, M. Arif, A. Kaushik, A. Mazroei, M. Katafygiotou, E. Elsarrag, Occupant
productivity and office indoor environment quality: a review of the literature,
Build. Environ. 105 (2016) 369–389, https://doi.org/10.1016/j.buildenv.2016.06.
001.
[6] S. Leder, G.R. Newsham, J.A. Veitch, S. Mancini, K.E. Charles, Effects of office
environment on employee satisfaction: a new analysis, Build. Res. Inf. 44 (2016)
34–50, https://doi.org/10.1080/09613218.2014.1003176.
[7] M. Frontczak, P. Wargocki, Literature survey on how different factors influence
human comfort in indoor environments, Build. Environ. 46 (2011) 922–937,
https://doi.org/10.1016/j.buildenv.2010.10.021.
[8] A.D. Galasiu, J.A. Veitch, Occupant preferences and satisfaction with the luminous
environment and control systems in daylit offices: a literature review, Energy Build.
38 (2006) 728–742 http://linkinghub.elsevier.com/retrieve/pii/
S0378778806000624.
[9] C. Cuttle, People and windows in workplaces, Proc. People Phys. Environ. Res.
Conf., Wellington, New Zealand, 1983, pp. 203–212.
[10] D.L. Butler, P.M. Biner, Effects of setting on window preferences and factors asso-
ciated with those preferences, Environ. Behav. (1989), https://doi.org/10.1177/
0013916589211002.
[11] M.C. Finnegan, L.Z. Solomon, Work attitudes in windowed vs. windowless en-
vironments, J. Soc. Psychol. (1981), https://doi.org/10.1080/00224545.1981.
9711674.
[12] M. Boubekri, F. Haghighat, Windows and environmental satisfaction: a survey study
of an office building, Indoor Built Environ. (1993), https://doi.org/10.1177/
1420326X9300200305.
[13] J.A. Veitch, R. Gifford, Assessing beliefs about lighting effects on health, perfor-
mance, mood, and social behavior, Environ. Behav. 28 (1996) 446–470, https://
doi.org/10.1177/0013916596284002.
[14] B.W.P. Wells, Subjective responses to the lighting installations in a modern office
building and their design implications, Build. Environ. 1 (1965) 57–68.
[15] K.M. Beauchemin, P. Hays, Sunny hospital rooms expedite recovery from severe
and refractory depressions, J. Affect. Disord. (1996), https://doi.org/10.1016/
0165-0327(96)00040-7.
[16] F. Benedetti, C. Colombo, B. Barbini, E. Campori, E. Smeraldi, Morning sunlight
reduces length of hospitalization in bipolar depression, J. Affect. Disord. (2001),
https://doi.org/10.1016/S0165-0327(00)00149-X.
[17] J.H. Choi, L.O. Beltran, H.S. Kim, Impacts of indoor daylight environments on pa-
tient average length of stay (ALOS) in a healthcare facility, Build. Environ. 50
(2012) 65–75, https://doi.org/10.1016/j.buildenv.2011.10.010.
[18] M.B.C. Aries, M.P.J. Aarts, J. Van Hoof, Daylight and health: a review of the evi-
dence and consequences for the built environment, Light. Res. Technol. 47 (2015)
6–27, https://doi.org/10.1177/1477153513509258.
[19] M. Boubekri, I.N. Cheung, K.J. Reid, C.H. Wang, P.C. Zee, Impact of windows and
daylight exposure on overall health and sleep quality of office workers: a case-
control pilot study, J. Clin. Sleep Med. 10 (2014) 603–611, https://doi.org/10.
5664/jcsm.3780.
[20] J.J. Romm, W.D. Browning, Greening the Building and the Bottom Line: Increasing
Productivity through Energy-Efficient Design, Rocky Mt. Inst., 1998, p. 16 http://
www.rmi.org/Knowledge-Center/Library/D94-27_GreeningBuildingBottomLine.
[21] P. Leather, M. Pyrgas, D. Beale, C. Lawrence, Windows in the workplace: sunlight,
view, and occupational stress, Environ. Behav. (1998), https://doi.org/10.1177/
001391659803000601.
[22] J. Christoffersen, K. Johnsen, Windows and daylight. A post-occupancy evaluation
of Danish offices, Light (2000) 112–120 2000.
[23] R.S. Ulrich, View through a window may influence recovery from surgery, Science
80 (1984), https://doi.org/10.1126/science.6143402.
[24] S.R. Kellert, J.H. Heerwagen, M.L. Mador (Eds.), Biophilic Design: the Theory,
Science, and Practice of Bringing Buildings to Life, Wiley, Hoboken, NJ, 2008.
[25] C.M. Tennessen, B. Cimprich, Views to nature: effects on attention, J. Environ.
Psychol. (1995), https://doi.org/10.1016/0272-4944(95)90016-0.
[26] M.D. Velarde, G. Fry, M. Tveit, Health Effects of Viewing Landscapes - Landscape
Types in Environmental Psychology, Urban For. Urban Green., 2007, https://doi.
org/10.1016/j.ufug.2007.07.001.
[27] A. Borisuit, F. Linhart, J.L. Scartezzini, M. Munch, Effects of realistic office day-
lighting and electric lighting conditions on visual comfort, alertness and mood,
Light. Res. Technol. 47 (2015) 192–209 http://lrt.sagepub.com/cgi/doi/10.1177/
1477153514531518.
[28] Heschong Mahone Group - California Energy Commision, Windows and Offices: A
Study of Office Worker Performance and the Indoor Environment, (2003), https://
doi.org/10.1175/1520-0450(1998)037<0414:TDFBIM>2.0.CO;2.
[29] M.G. Figueiro, M.S. Rea, A.C. Rea, R.G. Stevens, Daylight and Productivity –A Field
Study, ACEEE Summer Study Energy Effic. Build., 2002, pp. 69–78.
[30] M.B.C. Aries, J.A. Veitch, G.R. Newsham, Windows, view, and office characteristics
predict physical and psychological discomfort, J. Environ. Psychol. 30 (2010)
533–541, https://doi.org/10.1016/j.jenvp.2009.12.004.
[31] Z. Kong, D.M. Utzinger, K. Freihoefer, T. Steege, The impact of interior design on
visual discomfort reduction: a field study integrating lighting environments with
POE survey, Build. Environ. 138 (2018) 135–148, https://doi.org/10.1016/j.
buildenv.2018.04.025.
[32] E. Wotton, An Investigation of the Effects of Windows and Lighting in Offices,
Health and Welfare Canada, 1982.
[33] K. Van Den Wymelenberg, Patterns of occupant interaction with window blinds: a
literature review, Energy Build. 51 (2012) 165–176, https://doi.org/10.1016/j.
enbuild.2012.05.008.
[34] W. O'Brien, K. Kapsis, A.K. Athienitis, Manually-operated window shade patterns in
office buildings: a critical review, Build. Environ. (2013), https://doi.org/10.1016/
j.buildenv.2012.10.003.
[35] A.I. Rubin, B.L. Collins, R.L. Tibbott, Window blinds as a potential energy saver: a
case study, US Department of Commerce, National Bureau of Standards, 1978.
[36] T. Inoue, T. Kawase, T. Ibamoto, S. Takakusa, Y. Matsuo, The development of an
optimal control system for window shading devices based on investigations in office
buildings, ASHRAE Transact. (1988), https://doi.org/10.1007/s13398-014-0173-
7.2.
[37] M.S. Rea, Window blind occlusion: a pilot study, Build. Environ. (1984), https://
doi.org/10.1016/0360-1323(84)90038-6.
[38] Y. Sutter, D. Dumortier, M. Fontoynont, The use of shading systems in VDU task
offices: a pilot study, Energy Build. (2006), https://doi.org/10.1016/j.enbuild.
2006.03.010.
[39] B. Meerbeek, T. van Druenen, M. Aarts, E. van Loenen, E. Aarts, Impact of blinds
usage on energy consumption: automatic versus manual control, Eur. Conf.
Ambient Intell. 2014, pp. 158–173.
[40] C.F. Reinhart, K. Voss, Monitoring manual control of electric lighting and blinds,
Light. Res. Technol. (2003), https://doi.org/10.1191/1365782803li064oa.
[41] L.G. Bakker, E.C.M. Hoes-van Oeffelen, R.C.G.M. Loonen, J.L.M. Hensen, User sa-
tisfaction and interaction with automated dynamic facades: a pilot study, Build.
Environ. (2014), https://doi.org/10.1016/j.buildenv.2014.04.007.
[42] A. Piccolo, F. Simone, Effect of switchable glazing on discomfort glare from win-
dows, Build. Environ. (2009), https://doi.org/10.1016/j.buildenv.2008.08.013.
[43] J.-M. Dussault, L. Gosselin, Office buildings with electrochromic windows: a
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
12
sensitivity analysis of design parameters on energy performance, and thermal and
visual comfort, Energy Build. (2017), https://doi.org/10.1016/j.enbuild.2017.07.
046.
[44] N.L. Sbar, L. Podbelski, H.M. Yang, B. Pease, Electrochromic dynamic windows for
office buildings, Int. J. Sustain. Built Environ. (2012), https://doi.org/10.1016/j.
ijsbe.2012.09.001.
[45] M. Zinzi, Office worker preferences of electrochromic windows: a pilot study, Build.
Environ. (2006), https://doi.org/10.1016/j.buildenv.2005.05.010.
[46] A. Jamrozik, C. Campanella, B.A. Bauer, Well Living Lab: a new tool for measuring
the human experience in the built environment, Conscious Cities 2 (2017).
[47] A. Jamrozik, C. Ramos, J. Zhao, J. Bernau, N. Clements, T. Vetting Wolf, B. Bauer, A
novel methodology to realistically monitor office occupant reactions and environ-
mental conditions using a living lab, Build. Environ. (2018), https://doi.org/10.
1016/j.buildenv.2017.12.024.
[48] A. Miyake, N.P. Friedman, M.J. Emerson, a H. Witzki, A. Howerter, T.D. Wager, The
unity and diversity of executive functions and their contributions to complex
“Frontal Lobe”tasks: a latent variable analysis, Cogn. Psychol. 41 (2000) 49–100,
https://doi.org/10.1006/cogp.1999.0734.
[49] A. Diamond, Executive functions, Annu. Rev. Psychol. 64 (2013) 135–168, https://
doi.org/10.1146/annurev-psych-113011-143750.
[50] J.E. Fisk, C. a Sharp, Age-related impairment in executive functioning: updating,
inhibition, shifting, and access, J. Clin. Exp. Neuropsychol. 26 (2004) 874–890,
https://doi.org/10.1080/13803390490510680.
[51] J.A. Horne, O. Ostberg, A self-assessment questionnaire to determine morningness-
eveningness in human circadian rhythms, Int. J. Chronobiol. 4 (1975) 97–110
http://europepmc.org/abstract/med/1027738.
[52] K.E. West, M.R. Jablonski, B. Warfield, K.S. Cecil, M. James, M.A. Ayers, J. Maida,
C. Bowen, D.H. Sliney, M.D. Rollag, J.P. Hanifin, G.C. Brainard, Blue light from
light-emitting diodes elicits a dose-dependent suppression of melatonin in humans,
J. Appl. Physiol. 110 (2011).
[53] N. Unsworth, R.P. Heitz, R.W. Engle, An automated version of the operation span
task, Behav. Res. Methods 37 (2005) 498–505.
[54] J.L. Foster, Z. Shipstead, T.L. Harrison, K.L. Hicks, T.S. Redick, R.W. Engle,
Shortened complex span tasks can reliably measure working memory capacity,
Mem. Cogn. 43 (2014) 226–236 http://link.springer.com/10.3758/s13421-014-
0461-7.
[55] D. Besner, J.A. Stolz, C. Boutiuer, The Stroop effect and the myth of automaticity,
Psychon. Bull. Rev. 4 (1997) 221–225.
[56] C.M. Arrington, G.D. Logan, Voluntary task switching: chasing the elusive ho-
munculus, J. Exp. Psychol. Learn. Mem. Cogn. 31 (2005) 683–702, https://doi.org/
10.1037/0278-7393.31.4.683.
[57] W. Kool, J.T. McGuire, Z.B. Rosen, M.M. Botvinick, Decision making and the
avoidance of cognitive demand, J. Exp. Psychol. Gen. 139 (2010) 665–682, https://
doi.org/10.1037/a0020198.
[58] J. Park, Are Humans Good Sensors? Using Occupants as Sensors for Indoor
Environmental Quality Assessment and for Developing Thresholds that Matter,
Carnegie Mellon University, 2015.
[59] J.A. Veitch, K.E. Charles, K.M.J. Farley, G.R. Newsham, A model of satisfaction with
open-plan office conditions: COPE field findings, J. Environ. Psychol. 27 (2007)
177–189, https://doi.org/10.1016/j.jenvp.2007.04.002.
[60] A.U. Viola, L.M. James, L.J.M. Schlangen, D.J. Dijk, Blue-enriched white light in the
workplace improves self-reported alertness, performance and sleep quality, Scand.
J. Work Environ. Health 34 (2008) 297–306, https://doi.org/10.5271/sjweh.1268.
[61] S. Burge, A. Hedge, S. Wilson, J.A. Bass, A. Robertson, Sick building syndrome, a
study of 4373 office workers, Ann. Occup. Hyg. 31 (1987) 493–504.
[62] S.G. Luke, Evaluating significance in linear mixed-effects models in R, Behav. Res.
Methods (2017), https://doi.org/10.3758/s13428-016-0809-y.
[63] S. Kaplan, The restorative benefits of nature: toward an integrative framework, J.
Environ. Psychol. (1995), https://doi.org/10.1016/0272-4944(95)90001-2.
[64] S. Kaplan, M.G. Berman, Directed attention as a common resource for executive
functioning and Self-Regulation, Perspect. Psychol. Sci. (2010), https://doi.org/10.
1177/1745691609356784.
[65] M.G. Berman, J. Jonides, S. Kaplan, The cognitive benefits of interacting with
nature, Psychol. Sci. 19 (2008) 1207–1212, https://doi.org/10.1111/j.1467-9280.
2008.02225.x.
[66] R. Berto, Exposure to restorative environments helps restore attentional capacity, J.
Environ. Psychol. (2005), https://doi.org/10.1016/j.jenvp.2005.07.001.
[67] R. Berto, The role of nature in coping with psycho-physiological stress: a literature
review on restorativeness, Behav. Sci. 4 (2014) 394–409, https://doi.org/10.3390/
bs4040394.
[68] T.A. Markus, The function of windows- A reappraisal, Build. Sci. (1967), https://
doi.org/10.1016/0007-3628(67)90012-6.
[69] E. Ne’eman, G. Sweitzer, E. Vine, Office worker response to lighting and daylighting
issues in workspace environments: a pilot survey, Energy Build. (1984), https://doi.
org/10.1016/0378-7788(84)90071-9.
A. Jamrozik, et al. Building and Environment 165 (2019) 106379
13
