You are on page 1of 13

LEUKOS

The Journal of the Illuminating Engineering Society

ISSN: 1550-2724 (Print) 1550-2716 (Online) Journal homepage: https://www.tandfonline.com/loi/ulks20

Subjective Assessments of Lighting Quality: A


Measurement Review

Alicia C. Allan, Veronica Garcia-Hansen, Gillian Isoardi & Simon S. Smith

To cite this article: Alicia C. Allan, Veronica Garcia-Hansen, Gillian Isoardi & Simon S. Smith
(2019) Subjective Assessments of Lighting Quality: A Measurement Review, LEUKOS, 15:2-3,
115-126, DOI: 10.1080/15502724.2018.1531017

To link to this article: https://doi.org/10.1080/15502724.2018.1531017

© 2019 The Author(s). Published with


license by Taylor & Francis Group, LLC.

Published online: 04 Feb 2019.

Submit your article to this journal

Article views: 5291

View related articles

View Crossmark data

Citing articles: 16 View citing articles

Full Terms & Conditions of access and use can be found at


https://www.tandfonline.com/action/journalInformation?journalCode=ulks20
LEUKOS
2019, VOL. 15, NOS. 2–3, 115–126
https://doi.org/10.1080/15502724.2018.1531017

Subjective Assessments of Lighting Quality: A Measurement Review


Alicia C. Allana, Veronica Garcia-Hansena, Gillian Isoardib, and Simon S. Smithc
a
School of Design, Queensland University of Technology, Brisbane, Australia; bLight Naturally, Brisbane, Australia; cInstitute for Social Science
Research, University of Queensland, Brisbane, Australia

ABSTRACT ARTICLE HISTORY


Lighting is an important component of indoor environmental quality that can affect occupant Received 31 January 2018
satisfaction, well-being and productivity. Lighting quality is a broad abstract concept and this has Revised 11 September 2018
implications for its assessment. Subjective evaluations of lighting are an important complement to Accepted 28 September 2018
objective photometric information; however, there is limited existing guidance for the selection of KEYWORDS
such measures. We review and highlight the advantages and limitations associated with measures Glare; lighting quality;
of general lighting quality and discomfort glare. Existing measures of lighting quality have broad lighting satisfaction;
coverage of individual lighting features but do not always clearly form cohesive scales measuring postoccupancy evaluation;
an underlying construct. Questions used in experimental glare research focus narrowly on glare visual discomfort
severity, with ambiguous response rating scales. There is a need for the development of reliable
and valid tools to assess lighting quality and its components, with clearly defined definitions and
constructs, and explicit reporting of psychometric scale properties. The development of rigorous
self-report tools will improve the understanding and design of quality lighting environments.

1. Introduction individual completing the tasks and activities), sug-


gesting that lighting quality depends on an interac-
Indoor environmental quality has implications for
tion between lighting, place, and person. Evaluation
the satisfaction, well-being, and productivity of
of lighting typically involves detailed photometric
building occupants (Humphreys and Nicol 2007;
measures of the environment; however, photometric
Newsham et al. 2009; Singh et al. 2010). Together
measures alone do not fully capture the “person”
with thermal, acoustic, and air quality components,
component of lighting quality. Established protocols
lighting is a key aspect of indoor environmental
for describing lighting (CIE 213: 2014; International
quality. Lighting can affect occupant well-being and
Energy Agency [IEA] 2016) acknowledge the impor-
productivity, and dissatisfaction with lighting can
tance of capturing subjective evaluations in addition
influence operational costs if it results in redesign
to objective photometric information; however, they
or unanticipated occupant intervention into the
do not provide any summary, comparison, or gui-
environment. Although a desirable outcome, light-
dance for selecting measures. Therefore, it is not
ing quality remains an abstract concept without
clear how existing subjective evaluation measures
a widely agreed definition. Boyce (2014) suggests
overlap and what gaps remain. This review aims to
that lighting quality is best considered in terms of
summarize existing measures of subjective lighting
its impact on outcomes such as visual performance
quality, identify advantages and limitations of exist-
and behavior, and Veitch and Newsham (1996)
ing surveys to inform measure selection, and provide
define lighting quality more broadly, as lighting
recommendations for future measure development.
that supports visual performance, task and beha-
vioral performance, social interactions, mood, health
and safety, and aesthetic judgments. Integral to both 2. Review of measures
definitions are acknowledgement of the purpose of
To identify measures, we conducted database
the space (the tasks or activities that take place in
searches, examined established protocol and
a particular setting) and the user of a space (the
guideline documents from international lighting

CONTACT Alicia C. Allan alicia.allan@qut.edu.au School of Design, Queensland University of Technology, GPO Box 2434, Brisbane 4000, Australia.
Open Access for this manuscript was supported by the Illuminating Engineering Society and International Commission on Illumination.
© 2019 The Author(s). Published with license by Taylor & Francis Group, LLC.
This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-
nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built
upon in any way.
116 A. C. ALLAN ET AL.

bodies, and examined methods and reference lists flicker) over the long term. It includes statements that
of articles reporting subjective evaluation. We assess quality (e.g., “The lights flicker throughout
identified four types of subjective measures: gen- the day”), comfort (e.g., “Overall, the lighting is com-
eral evaluations of lighting quality, questions fortable”), and performance (e.g., “Reflections from
examining discomfort glare (one specific feature the light fixtures hinder my work”). It has been used
of the lighting environment), indoor environmen- in both field and experimental settings (Borisuit et al.
tal quality scales, and affective response scales. Of 2014; Boyce et al. 2006; Eklund and Boyce 1996).
these, general lighting quality measures and mea-
sures of glare were the most prevalent and are
2.1.1.1. Advantages. The survey allows diagnosis
reviewed in detail.
of issues resulting from individual lighting fea-
tures, as well as an overall assessment of lighting
2.1. Measures of general lighting quality quality. Convergent validity is thoroughly demon-
strated via associations with previously developed
Subjective assessments form an important compo- tools, as is test–retest reliability (the stability of
nent of postoccupancy evaluations (POEs) of the a test over time) over one week. The authors
built environment. There are a number of occupant clearly identify the aim of their measure and ratio-
self-report measures that include general lighting nale for its development. Normative data for this
quality as either a substantial component or the sole survey are reported, providing a practical compar-
focus. Tools that have not been published in a peer- ison point for those using the tool in the field.
reviewed source or have an associated fee were not
included, because these restrictions limit utility to
researchers, widespread adoption, and comparability 2.1.1.2. Limitations. The survey does not evaluate
of results. The CIE guide to protocols for describing daylighting; however, an additional daylighting
lighting (CIE 2014) provides examples of existing question is described by Veitch (2010). The item
measures, two of which are reviewed here (the evaluating glare (“The light fixtures are too bright”)
Thorn Lighting Work Space Appraisal was excluded does not explicitly address discomfort and excludes
because it is a proprietary unpublished tool, and the daylighting. One question requires a comparative
Ergonomic Lighting indicator [Tralau et al. 2009] is quality judgment (i.e., “How does the lighting com-
a lighting design tool that does not capture occupant pare to similar workplaces in other buildings?”),
ratings). In addition to the CIE protocol examples, which may produce unreliable responses. Some
three other measures were included. scale properties such as internal reliability are not
Table 1 summarizes the key details of the five reported, although many aspects of reliability and
reviewed measures, including the number of ques- validity are examined. Finally, the norms provided
tions, whether non-lighting questions are included, are based on surveys of offices in the mid-1990s.
the type of response format used, the time frame Given the changes in building design and lighting
participants are asked to consider when responding, technology in the intervening period, these norms
and whether psychometric scale properties (e.g., may no longer be a useful reference point.
Cronbach’s alpha for evaluation of internal consis-
tency of scale items or results of structural validity 2.1.2. Lighting conditions survey
via factor analysis) were reported. The remaining Developed as part of the Daylight in Buildings task
columns indicate what specific features of the light- for the IEA Solar Heating and Cooling Programme
ing environment are addressed by the tool. (IEA 1999), the Lighting Conditions Survey con-
sists of 37 items that assess long-term lighting
2.1.1. Office lighting survey satisfaction, quality, and other features of the
The office lighting survey includes evaluations of workplace (i.e., office layout, orientation) in
broad (overall satisfaction and comfort; for example, a field setting. It includes broad questions relating
“Overall, the lighting is comfortable”) and specific to overall satisfaction, as well as questions about
characteristics of lighting (illuminance, brightness individual aspects of lighting (e.g., “Does the light-
ratio, veiling reflections, discomfort glare, color, and ing cause reflections in your work material?”).
LEUKOS 117

Table 1. Characteristics of existing lighting quality surveys.a


Lighting characteristics

Artificial light

Task lighting
Natural light
Performance

Distribution
Reflections
Comfort

Brightness
Windows

Shading
Control
Flicker

Color
Glare

View
Number Scale
of Other Response Time properties Overall
Survey questions areas? type frame reported? satisfaction
Office Lighting 10 No Agree/disagree LT Selected x P x x x x x x
Survey (Eklund three-point
and Boyce 1996) rating scale
Lighting Conditions 37 Yes Mixed (rating LT No F x x x x x x x x x x
Survey (IEA 1999) scales, multiple
choice, free-
text)
National Research 10 No Five-point CC Yes x P, x x
Council (NRC) rating scale L, S
Canada Lighting
Quality Scale
(Veitch and
Newsham 2000)
IEA retrofit 36 Yes Mixed, (rating LT & No F x x x x x x x x x
monitoring user scale and free- CC
assessment text)
survey (IEA 2016)
Satisfaction with 18 Yes Seven-point CC Yes S x x x
Environmental rating scale
Features Veitch
et al. (2007)
a
P = presence, L = location, S = severity, LT = long-term, CC = current conditions.

2.1.2.1. Advantages. Selected items from this sur- and the presence of reflections). The phrasing of the
vey have been used or adapted in subsequent questions relates to lighting during the current day.
research (Aries et al. 2010; Velds 2002), providing
the potential for comparison across different set- 2.1.3.1. Advantages. The authors performed fac-
tings. The survey covers many different individual tor analysis to identify two subcomponents of their
lighting characteristics, including a particular scale, lighting quality and glare, each of which
focus on daylighting (see Table 1 for a summary consists of multiple questions that can be averaged
of individual lighting characteristics covered). to give an overall subscale score and has high
internal reliability (as demonstrated via
Cronbach’s alpha). Multiple characteristics of
2.1.2.2. Limitations. A small number of specific
glare are recorded in this survey.
lighting features are not addressed (e.g., flicker and
control), nor is an overall rating of visual comfort.
2.1.3.2. Limitations. This survey was designed for
Questions focus on identification of individual
use in a mock-up experimental setting. Therefore,
problematic lighting features, and questions
it measures only selected lighting features, uses
describing the environment and environmental
a day-long reference period, and does not include
aspects (e.g., “Do you ever notice cold draughts
any daylighting questions.
near the windows?”) are interspersed with judg-
ments of satisfaction and quality, without an
attempt to clearly distinguish how questions 2.1.4. IEA retrofit monitoring user assessment survey
group together or to measure defined constructs. The lighting monitoring protocol for retrofits
described by the Solar Heating and Cooling
Programme of the International Energy Agency
2.1.3. NRC Canada Lighting Quality Scale includes a “user assessment” portion where occupant
Veitch and Newsham (2000) developed this scale as ratings are captured (IEA 2016). A case report using
part of an office simulation experiment. It includes this tool has also been described (Gentile et al. 2016).
broad overall satisfaction questions, as well more It assesses individual features but does not include
specific assessment of individual features (i.e., visual a broad question assessing overall satisfaction with
performance, presence, location and severity of glare, lighting. Assessments are made about both current
118 A. C. ALLAN ET AL.

and long-term conditions, and it includes judgments psychometric development approach and describe
of satisfaction (e.g., “General satisfaction with the the rationale, constructs assessed, scale factor ana-
following aspects of this room”) and quality (e.g., lysis, and relevant scale properties.
“Do you ever experience flickering from the electric
light sources in this room?”). 2.2.1.2. Limitations. Because it is part of a broader
questionnaire on environmental features, this survey
2.1.4.1. Advantages. This tool captures occupant is brief and measures few aspects of lighting quality.
satisfaction and evaluations of a range of specific
lighting features (see Table 1 for a summary of
2.3. Subjective ratings of discomfort glare
coverage), including daylighting. The inclusion of
questions relating both to current conditions and All of the general lighting quality scales reviewed
to long-term experience provides multiple per- included questions relating to glare, indicating its
spectives to the administrator. importance for any occupant assessment of lighting.
Glare is one specific component of subjective light-
2.1.4.2. Limitations. Some lighting features are ing quality that has received considerable experi-
assessed only as a point-in-time rating. Although mental attention. Glare occurs when the range of
the authors report using scale design methodology, simultaneous luminances within the field of view
environmental feature questions are interspersed causes discomfort or loss of visual performance
with lighting questions, and it is not clear how (Boyce 2014). Discomfort glare causes annoyance
multiple items could be combined. As such, the or discomfort to the viewer, and a large body of
constructs measured are not clearly identified, nor experimental work has attempted to describe the
are psychometric properties reported. physiological mechanisms involved. Although it is
a subjective process, there has been substantial work
in the development of glare indices in an effort to
2.2. Indoor environmental quality surveys
quantify the likelihood of glare occurring in a given
Broader surveys of indoor environmental quality gen- luminous scene. Glare indices such as the daylight
erally include one or two questions addressing lighting glare probability index (Wienold and Christoffersen
in addition to items regarding acoustics, noise, and 2006) assess the luminances in a visual field to esti-
other characteristics of the indoor environment. mate the likelihood of glare occurring. However,
Indoor environmental quality scales are generally glare prediction indices do not entirely explain the
reported as an overall environmental satisfaction or perception of glare, and their performance degrades
utility score without an individually identifiable light- the further they are applied from highly controlled
ing section and therefore most are not included in this settings (Jakubiec and Reinhart 2015; Van Den
review. However, the Satisfaction with Environmental Wymelenberg and Inanici 2016).
Features survey includes a substantial lighting assess- As Boyce (2014) comments, even with the best
ment component reported as a distinct construct and characterization of the physiological processes asso-
is therefore included in Table 1. ciated with glare “… there would still be the funda-
mental fact that discomfort glare involves
2.2.1. Satisfaction with environmental features a psychological element in the form of the different
This survey assesses occupants’ satisfaction with criteria used by individuals” (179). Glare indices
three separate environmental features: privacy/ represent only a measure of glare potential rather
noise, ventilation/temperature, and lighting (Veitch than being a direct measure of the experience of
et al. 2007). Five of the 181 items relate to lighting, glare itself, which requires assessment from
specifically satisfaction with brightness, reflections, a viewer. Further, subjective observer ratings are
glare, and view. Participants are asked to respond used to validate glare prediction indices and there-
based on conditions at the time of evaluation. fore directly inform their efficacy. As a result, it is
important to consider how subjective assessments
2.2.1.1. Advantages. The authors report on field are obtained. The majority of studies assessing glare
validation results and use an appropriate are completed in highly controlled experimental
LEUKOS 119

Table 2. Items used to assess glare severity in experimental settings.


Source Question details Examples of subsequent use
De Boer (1967) Nine-point rating scale, cues at odd numbers: 1 = Bullough et al. (2008); Balk and Tyrrell (2011)
unbearable, 3 = disturbing, 5 = just acceptable, 7 =
satisfactory, 9 = just noticeable
Osterhaus and Exact question wording not always reported Christoffersen and Wienold (2004); Wienold and Christoffersen
Bailey (1992) 4-point rating scale: imperceptible, noticeable, disturbing, (2006); Jakubiec and Reinhart (2015); Karlsen et al. (2015);
intolerable Rodriguez et al. (2015)
Definitions quantifying the amount of time individuals
could work with glare present often provided
Iwata et al. (1992) Exact question wording not always reported Kent, Altomonte et al. (2017); Tuaycharoen and Tregenza
Four-point response scale: just perceptible, just acceptable, (2007; with additional scale points); Velds (2002; with
just uncomfortable, just intolerable behavioural definitions and a “no glare” response)
Van Den Visual comfort measured using the question: “This is Van Den Wymelenberg and Inanici (2016); Mahić et al. (2017)
Wymelenberg a visually comfortable environment for the intended work”
and Iancini (2014) Seven-point rating scale (strongly disagree to strongly agree)

settings, where specific lighting conditions or indivi- which may not be consistently interpreted by par-
dual fixtures are manipulated carefully in order to ticipants. It was designed for the assessment of the
establish photometric thresholds for glare response. glare following direct exposure to luminaires and
Table 2 describes individual items that have been therefore may not transfer well to capturing the
developed for this purpose. Although some are admi- overall experience of glare in a field setting, parti-
nistered alongside other questions relating to lighting cularly where daylight components are present.
and glare, single items are used as the outcome vari-
able capturing glare severity. All of the items in Table 2.3.2. Imperceptible–intolerable four-point scale
2 were developed for use in experimental settings to This four-point rating scale, as originally described by
gather point-in-time ratings of current conditions. Osterhaus and Bailey (1992), is frequently used to
capture subjective assessments of glare in experimen-
2.3.1. De Boer scale tal settings. The exact wording of the question varies
One of the oldest and most widely used glare scales is across studies; however, the verbal cues for the four-
that developed by De Boer (1967). This question asks point rating scale are generally imperceptible, notice-
participants to rate currently experienced glare on able, disturbing, and intolerable. Many studies, but not
a nine-point scale, with anchors at 1 = unbearable, 3 all, provide participants with an extended descriptive
= disturbing, 5 = just acceptable, 7 = satisfactory, and definition for each scale point describing the length of
9 = just noticeable. The scale was originally designed time that glare could be tolerated. Ratings on this four-
for the evaluation of visual perception in road traffic point scale are often dichotomized at the point
in motorists. between imperceptible and noticeable or between
noticeable and disturbing (Rodriguez et al. 2017; Van
2.3.1.1. Advantages. This scale has been used in Den Wymelenberg 2014). An adaptation of this scale
research assessing glare from indoor and outdoor has been used by Jakubiec and Reinhart (2015) to
luminaire sources. assess the experience of glare over the long term (an
academic semester), asking participants to rate the
2.3.1.2. Limitations. The limitations associated typical amount of discomfort glare they experienced
with the De Boer scale (and other glare scales) in the morning, midday, and afternoon, with addi-
are summarized thoroughly by Fotios (2015). Of tional questions to assess glare source and intervention
most concern is its failure to provide an option for strategies. Some subsequent reports use this in con-
reporting the absence of glare—the response scale junction with questions assessing other aspects of glare
presumes some level of glare. This is a major and lighting (Wienold and Christoffersen 2006).
drawback—asking participants to rate the extent
of the glare they are currently experiencing is 2.3.2.1. Advantages. This rating scale has been
inappropriate if there is none present. In addition, widely used, which facilitates the comparability and
the scale has very ambiguous descriptive anchors, statistical comparison and combination of studies
120 A. C. ALLAN ET AL.

conducted in different settings. Additionally, glare descriptors of the response scale do not by them-
indices used to estimate the glare potential of the selves communicate a clear order of increasing
luminous environment (such as the daylight glare severity.
probability index) have been validated against these
scale anchors in experimental settings. Therefore, if 2.3.4. Visual comfort rating
a main aim is to compare and combine results with Van Den Wymelenberg (2014) proposes that glare
other studies, this scale is desirable. The adaptation may be better captured by phrasing questions not in
of this question to a long-term rating of the environ- terms of discomfort but in terms of comfort. A series
ment (Jakubiec and Reinhart 2015) provides a useful of studies (Mahić et al. 2017; Van Den Wymelenberg
bridge between the measures used in experimental and Inanici 2014, 2016) have used a positively
glare research and field settings. worded question with a seven-point response scale
to assess glare. This question has been rated along-
2.3.2.2. Limitations. The question used to elicit side more general questions of lighting quality that
participant responses is not always the same, which address visual comfort, visual appearance, bright-
may impede comparability between studies. The ness, reflections, and distribution.
phrasing often implies the presence of glare, and the
response scale points are ambiguous. The lowest rat- 2.3.4.1. Advantages. This question is easy to
ing (imperceptible glare) is not plain English and is an understand and does not presume the existence
oxymoron—by definition, glare is a perceptual process of glare in the phrasing of the question. The verbal
that requires a sensation of annoyance or discomfort anchors are clearly ordered, unambiguous, and do
to exist. This could be rectified by substituting imper- not require extended behavioral definitions.
ceptible with another scale anchor that clearly com-
municates that no glare is present. The use of 2.3.4.2. Limitations. The construct of visual com-
unnoticeable by Rodriguez et al. (2015) as the lowest fort can be interpreted more broadly than the
response scale point improves but does not entirely absence of glare. This question is suitable for asses-
remove this issue. The use of extended behaviorally sing glare in experimental settings where the lighting
anchored definitions is likely to provide more reliable conditions are carefully managed and manipulated
judgments; however, including definitions may not be (the setting this item was originally applied in).
feasible or desirable in a field setting. However, visual comfort can be influenced by other
aspects of lighting (e.g., flicker), and in a field setting
this question may not be specific to glare.
2.3.3. Glare sensation vote
A glare sensation vote subjective rating (Hopkinson
1972; further adapted by Iwata et al. 1992) is similar to 2.4. Other subjective measures of lighting
the above imperceptible–intolerable scale (Osterhaus
2.4.1. Descriptive scales
and Bailey 1992), with slightly different scale presenta-
Flynn et al. (1979) provide a series of semantic
tion and labels. It has also been used with extended
differential scales to evaluate user impressions
behavioral definitions (e.g., Kent, Fotios et al. 2017).
resulting from a series of lighting scenarios.
Descriptive prompts such as “clarity,” “order,” “spa-
2.3.3.1. Advantages. Versions of this scale have ciousness,” “relaxation,” and “privacy” are rated by
been used in numerous studies, sometimes with occupants to characterize their affective response to
the addition of a no glare response and behavioral the environment. These scales allow a more impres-
definitions for each of the scale points, which sionistic description of the lighting in a space in
improves clarity for respondents. a way that the previously discussed surveys do not.
They have been used to help understand how light-
2.3.3.2. Limitations. As with the previous scale, ing design affects occupants (Durak et al. 2007;
there is not always an option to indicate that Manav and Yener 1999). However, this descriptive
there is no glare present. The question is not approach does not provide any indication of
always clearly described, and the semantic whether these characteristics match the observer’s
LEUKOS 121

aesthetic preferences, making them difficult to trans- environments. Future protocol documents could
late into a measure of lighting quality or satisfaction. recommend not only that subjective evaluations
should be gathered but also specify what lighting
2.4.2. Lighting preferences, beliefs, and behavioral characteristics should be assessed by occupants.
consequences The lighting characteristics listed in Table 1
Finally, there are also studies of general lighting could serve as a starting point for this discussion.
preferences (Haans 2014; Veitch et al. 2010) or We suggest that a broad coverage of characteristics
beliefs (Veitch and Gifford 1996); however, we would be of most value and that questions relating
believe that these are better conceptualized as indi- to daylighting are critical inclusions for field sur-
vidual differences that predict subjective assessments veys, given the heavy reliance on daylighting stra-
of lighting quality, rather than forming part of the tegies in modern buildings.
lighting quality assessment itself. Additionally, beha-
vioral consequences of poor lighting quality (e.g., use 2.5.3. Evaluation type
of blinds) are a related yet conceptually distinct set of Judgments of lighting quality (rating from bad to
behaviors. As a result, none of these have been good), judgments of satisfaction (unsatisfactory to
included in this review. satisfactory), or endorsement of a problem or issue
(e.g., “The lighting causes deep shadows,” as
included in the Office Lighting Survey) were all
2.5. Gaps and recommendations for future
used within surveys. Authors should think criti-
measure development
cally about the relevance of the type of judgments
2.5.1. General lighting quality scales they are asking occupants to make about the light-
The reviewed general lighting quality surveys dif- ing and whether or not it is useful to combine
fered considerably in terms of type and breadth of different types of evaluations. Depending on con-
content (i.e., lighting characteristics), the time period text, satisfaction ratings may be more meaningful
they asked respondents to consider (i.e., current than quality ratings, which rely on knowledge and
conditions vs. long-term assessments), and the type may not always be suitable for nonexpert occupant
of judgment they asked people to make. Readers can respondents (Veitch et al. 2007).
use Table 1 in combination with the in-text descrip-
tions to select or adapt a measure that may be useful 2.5.4. Articulation of constructs and use of
for their particular setting. We identified a number psychometric scale development methods
of overall limitations with existing subjective assess- There is a focus on asking respondents to rate or
ments of lighting and provide some suggestions for identify problems with individual lighting charac-
the future development of measures. teristics at the expense of broader evaluations (e.g.,
of satisfaction or comfort). It is not always clear
2.5.2. Consensus on lighting characteristics how items group together or how existing ques-
There is not currently agreement on which indivi- tionnaires could map onto broader components
dual lighting characteristics should be included in and existing definitions of lighting quality (e.g.,
a subjective assessment of lighting quality. Though including visual comfort, performance, and aes-
all addressed glare, questions regarding daylighting/ thetics). Two of the measures reviewed reported
natural lighting, flicker, lighting control, and task no psychometric properties of their scale (such as
lighting were inconsistently included in measures internal reliability, factor structure, or correlations
(see Table 1). None of the measures included items with other related measures) to assess validity and
relating to aesthetic judgments or satisfaction, nor reliability, despite previous calls for authors to use
did any address variability of light across the day, psychometric methods for measure development
which may be relevant in heavily daylit spaces or as (CIE 2014). In these measures, lighting appraisal
a result of variable or smart lighting schemes. questions are presented intermixed with questions
The field would benefit from some consensus assessing other aspects of the physical environ-
on what specific lighting characteristics should be ment (e.g., type of office or proximity of windows).
included in subjective assessments of lighting Although this broader environmental information
122 A. C. ALLAN ET AL.

is critical to collect in a POE evaluation, these 2.6. Glare scales


questions do not themselves directly address light-
Glare is a more specific construct representing just
ing quality and should be clearly differentiated.
one component of lighting quality. The single-item
There are conventional statistical approaches for
questions shown in Table 2 have generally been
the creation of self-report measures that are
applied in more controlled research settings. As
described extensively in psychological literature
such, there is more clarity regarding what is
and have been widely adopted within other disci-
being measured and how it is being measured.
plines (see Cronbach and Meehl 1955; DeVellis
Despite this, there remains a lack of agreement
2012). Because subjective experiences are intangi-
on a universal rating scale (Fotios 2015; Van Den
ble and not directly observable, we rely on self-
Wymelenberg), which limits comparability
report instruments to measure underlying con-
between experimental studies assessing glare.
structs—in this case, lighting quality—and their
Fotios (2015) notes that despite the creation of
relationships. At a minimum, scale development
glare indices based on the results of subjective
requires clear articulation of the theoretical con-
ratings, there is still considerable ambiguity in
structs to be measured and a sound argument for
the methods used to elicit those ratings.
why items relate to that construct. Following this,
It may be that measurement of behavioral or
a comprehensive pool of items (larger than the
physiological responses is a more useful indicator
eventual intended length of the measure) is gen-
of glare in some settings than stated preferences
erally administered to an initial group and reduced
(Fotios 2018). However, behaviors are not them-
based on initial validation. Statistical psychometric
selves preferences and can be influenced by other
methods are used to refine the item pool. Factor
factors, particularly in a field setting. For example,
analysis can be used to identify and confirm the
the use of blinds may serve as a practical indicator
underlying structure of a scale (and possible sub-
of glare presence in a field setting but is also
scales) that estimates levels of a more abstract
dictated by external factors such as view desirabil-
theoretical construct (e.g., satisfaction with light-
ity and social dynamics. Therefore, rating scales
ing). Future measures should document the ratio-
will likely remain an important component of
nale and conceptual basis for their development
glare evaluation.
and overtly label the variable being measured by
each question or group of questions. Methods and
rationale for measure development should be pub- 2.6.1. Dichotomization of multipoint scales
lished with the results of psychometric analyses Multipoint ratings of glare are often dichotomized
such as internal consistency of the items in to create categories of comfort and discomfort;
a scale, factor analysis, or item response analyses. however, there is disagreement on the placement
Rather than being a perfunctory exercise, enga- of this border, creating inconsistencies and lack of
ging in formal statistical analysis of scale proper- comparability between studies (Van Den
ties helps authors determine what construct is Wymelenberg). The dichotomization of continu-
being measured, reveal the structure of any sub- ous variables can be problematic for both concep-
components, explore whether questions group tual and statistical reasons. Dichotomization
together effectively, assess similarity to and differ- considerably reduces statistical power and
ence from other related constructs and measures, increases the likelihood of false-positive results
and identify whether there are particular items (i.e., detecting discomfort where none exists).
that are not performing as intended. These ana- Dichotomized results do not accurately represent
lyses help the authors to demonstrate the quality the differences between groups, because they treat
of their measures and identify relationships and respondents straddling the borderline as being
hierarchies between the individual items and very different, when in fact they are very similar
broader constructs and theoretical frameworks. (Altman 2006). However, the ongoing difficulty in
They can also be used to identify which questions identifying a clean threshold that differentiates
are most valuable to retain in shorter forms or in luminous scenes that are comfortable from those
broader surveys of indoor environmental quality. that are problematic suggests that imposing a cut-
LEUKOS 123

point does not solve these issues and instead runs Across this broader time frame, other glare char-
the risk of creating spurious results. Analyzing acteristics may be of interest, such as glare source,
interval or multipoint data requires different sta- which is often controlled in an experimental setting
tistical approaches but is very feasible, and we (Christoffersen and Wienold 2004; Hirning et al.
recommend increased analysis of these scales in 2013), glare frequency (IEA 2016), glare duration,
their gathered format. and glare timing (Jakubiec and Reinhart 2015),
both across the day and across the year. The general
lighting quality surveys all include glare questions.
2.6.2. Question construction
Some ask occupants to rate the frequency rather than
Most existing glare questions do not provide
severity of glare; for example, “Does the artificial light
respondents with an option to indicate absence
ever cause glare strong enough to bother you?” (IEA
of discomfort or glare, and some response scale
1999) or, “In general, how often do you experience
anchors (i.e., the descriptive words associated with
glare from direct sunlight?” (IEA 2016). These addi-
each scale point) are ambiguous and not clearly
tional glare characteristics have not been systemati-
ordered. Further, the exact wording of questions
cally described and are therefore not reviewed
used to elicit ratings is not always reported in full,
separately. However, describing and defining these
leaving room for procedural discrepancies and
additional glare characteristics and measuring them
reduced comparability between studies. A sizeable
in field settings will allow us to better determine their
body of literature has shown that the wording and
importance and better predict the experience of
order of questions has a demonstrable impact on
glare. We recommend that authors consider which
participant responses, and question wording has
glare characteristic they are assessing and explicitly
implications for readability, translations, and study
state this when describing their measures.
replication.
The exact wording of questions posed to parti-
cipants and the associated response scale anchors 2.7. Broad recommendations for lighting
should always be reported in full. Any changes to research
the wording of previously published measures
2.7.1. Consistency across research areas
should be explicitly described, and questions
The field would benefit from greater consistency
should be phrased to avoid inferring the presence
in measures used to assess glare and lighting qual-
of discomfort glare. Response options must
ity in general. Questions designed for administra-
include a clear option for participants to indicate
tion in experimental contexts for moment-in-time
that they are not experiencing discomfort or glare,
ratings may differ from those designed for long-
and response scale descriptors should have an
term assessments in field settings. However, there
unambiguous order that indicates increasing or
could be greater consistency in the structure or
decreasing severity.
phrasing of questions to allow both types of mea-
surements to be meaningfully integrated and com-
2.6.3. Additional glare characteristics pared. Additionally, the discomfort glare literature
Glare research has focused heavily on luminous often uses the terms visual discomfort and glare
conditions that lead to the acute experience of interchangeably. Visual comfort can be influenced
glare. When glare is assessed at a point in time by a range of lighting issues, including insufficient
(i.e., observers are asked to rate current conditions) light and flicker (Boyce 2014), and may represent
in a highly controlled experimental setting, the a slightly broader construct. Consideration of the-
primary interest is the severity of glare, and the oretical definitions and careful construction of
questions in Table 2 address glare severity (or lighting quality scales could help to resolve this
simply glare presence when they are dichoto- hierarchy of constructs.
mized). However, in POE or field settings,
researchers, practitioners, and building managers 2.7.2. General advice for scale construction
may be more interested in occupants’ experiences We strongly recommend that scale development
of glare over the long term. texts are consulted in the development of future
124 A. C. ALLAN ET AL.

measures, because there is a wealth information they have narrow applicability. Increased transfer-
available regarding the construction of questions ability to a wider variety of settings could be
for self-report scales (see DeVellis 2012; Kline achieved by writing questions that are task and
1986). In general, questions should avoid using setting agnostic (i.e., use of language such as “this
technical jargon and ambiguity, remain relatively space” or “usual tasks or activities”) or by phrasing
short, and ensure that they address only one aspect questions such that the named task or setting can
or issue. Questions should avoid leading the respon- be easily substituted without changing question
dent to a particular conclusion (e.g., inferring the structure. More broadly applicable questions may
presence of glare), and negatively worded questions encourage increased examination of lighting qual-
should be written with care, because they can be ity outside of an office building context.
confusing to readers. Response rating scales should
be labeled in a way that implies approximately equal
3. Conclusions
weighting across the scale, is unambiguous, and has
a clear increasing or decreasing order. There are Subjective evaluations of lighting are an important
mixed views regarding the optimal number of scale complement to photometric measures when deter-
points in Likert-type responses; however, five- to mining lighting quality. Existing protocols acknowl-
seven-point scales are generally desirable, because edge the role of subjective measures of lighting, and
higher numbers can become impractical and lower this review reveals that although a number of such
numbers can provide inadequate variation. measures exist, there is room for further develop-
Although single-item measures are desirable for ment of reliable and valid tools to assess lighting
their brevity, asking multiple questions measuring quality and its components. Therefore, development
the same underlying variable reduces the likelihood of future measures should be based on clear defini-
that the specific phrasing of the question results in tions of lighting quality and its component parts,
misinterpretation by the respondent. Further, multi- follow established methods for scale development,
item measures allow researchers to assess the inter- and explicitly report the psychometric scale.
nal consistency of responses using Cronbach’s alpha. The lack of consensus on scales assessing lighting
Single items that effectively measure a construct can quality greatly limits the comparability of results and
emerge from and be validated against longer mea- the pooling of data from multiple studies to compare
sures, although even carefully designed single-item and contrast lighting quality in different settings.
responses are typically less stable than those Greater consensus and consistency in measures
obtained from multi-item measures (DeVellis 2012). would enable broader examination of how different
In experimental settings where it is practical to environmental, architectural, and lighting features
obtain only one or a handful of ratings per condi- influence occupants’ judgments of lighting quality
tion, identical conditions can be repeated and and satisfaction. Further, consistent implementation
responses averaged to give a more stable subjective of standard measures would allow establishment of
estimate of the measured variable. However, even norms as originally undertaken with the Office
well-designed subjective scales cannot rectify over- Lighting Survey (Eklund and Boyce 1996). This
sights in experimental design that allow for the would dramatically improve the utility of such scales
influence of confounding variables, order effects, for practitioners, who could compare results from
and anchoring biases (see Kent, Fotios et al. 2017; individual postoccupancy evaluations to existing pub-
Kent et al. 2018). These effects can be avoided lished norms and report benchmarks to clients.
through careful experimental design (e.g., via ran- Gathering subjective assessments is inherently
domization or counterbalancing of conditions) less precise than capturing physical measurements
and are not specific to rating scales—they can but is no less important. The careful design of
also affect behavioral indicators or physiological subjective measures can greatly reduce measure-
responses. ment uncertainty. By developing reliable and valid
Finally, though questions that name a specific tools, we can improve our understanding of light-
task (e.g., computer work) or setting (e.g., offices) ing quality and establish design solutions that lead
increase the tangibility of questions to participants, to more satisfied building occupants.
LEUKOS 125

Disclosure statement DeVellis R. 2012. Scale development: theory and applications.


3rd ed. Thousand Oaks (CA): SAGE.
Gillian Isoardi is a consultant for Light Naturally, who have Durak A, Camgöz Olguntürk N, Yener C, Güvenç D, Gürçinar Y.
provided support to an associated project. Light Naturally has 2007. Impact of lighting arrangements and illuminances on
no business or financial benefits directly relating to the contents different impressions of a room. Build Environ. 42:3476–-
of this review. The other authors have no interests to declare. 82. doi:10.1016/j.buildenv.2006.10.048
Eklund NH, Boyce PR 1996. The development of a reliable, valid,
and simple office lighting survey. J. Illum. Eng. Soc. 25:25
Funding Flynn J, Hendrick C, Spencer TOM. 1979. A guide to the
This work was supported by Australian Research Council methodology procedures for measuring subjective impres-
Linkage grant “Designing Healthy and Efficient Lighting sions in lighting. Journal of the Illuminating Engineering
Environments in Green Buildings”(LP150100179) in partner- Society 1979;8:95-110.). J Illum Eng Soc. 8:95–110.
ship with AECOM and Light Naturally. Fotios S. 2015. Research Note : uncertainty in subjective eva-
luation of discomfort glare. Light Res Technol. 47:379–83.
doi:10.1177/1477153515574985
Fotios S. 2018. Correspondence: New methods for the eva-
References luation of discomfort glare. Light Res Technol. 50:489–491.
doi:10.1177/1477153518773577
[CIE] Commission Internationale de l’Eclairage. 2014. Guide Gentile N, Dubois M, Osterhaus W, Stoffer S, Naves C,
to protocols for describing lighting. Vienna (Austria): CIE. Amorim D, Geisler-Moroder D, Jakobiak R. 2016.
Publication No: CIE 213:2014. A toolbox to evaluate non-residential lighting and day-
Altman DG. 2006. The cost of dichotomising continuous lighting retrofit in practice. Energy Build. 123:151–61.
variables. BMJ. 332:1080–1080. doi:10.1136/bmj.332.7549. doi:10.1016/j.enbuild.2016.04.026
1080 Haans A. 2014. The natural preference in people’s appraisal
Aries MBC, Veitch JA, Newsham GR. 2010. Windows, view, of light. J Environ Psychol. 39:51–61. doi:10.1016/j.
and office characteristics predict physical and psychologi- jenvp.2014.04.001
cal discomfort. J Environ Psychol. 30:533–41. doi:10.1016/ Hirning MB, Isoardi GL, Coyne S, Garcia Hansen VR,
j.jenvp.2009.12.004 Cowling I. 2013. Post occupancy evaluations relating to dis-
Balk S, Tyrrell R. 2011. The Accuracy of Drivers’ Judgments comfort glare: A study of green buildings in Brisbane. Build
of the Effects of Headlight Glare: Are We Really Blinded Environ. 59:349–57. doi:10.1016/j.buildenv.2012.08.032
by the Light?. In: Proceedings of the Sixth International Hopkinson RG. 1972. Glare from daylighting in buildings.
Driving Symposium on Human Factors in Driver Appl Ergon. 3:206–15.
Assessment, Training and Vehicle Design, June 27-30, Humphreys MA, Nicol JF. 2007. Self-Assessed Productivity
Olympic Valley — Lake Tahoe, California. Iowa City, IA: and the Office Environment. ASHRAE Trans. 113:606–16.
Public Policy Center, University of Iowa, 2011:510–517. International Energy Agency. 1999. Post occupancy evalua-
Borisuit A, Linhart F, Scartezzini J-L, Munch M. 2014. Effects tion of daylight in buildings. Galve (Sweden): Centre for
of realistic office daylighting and electric lighting condi- Built Environment.
tions on visual comfort, alertness and mood. Light Res. International Energy Agency. 2016. Monitoring protocol for
1–18. doi:10.1177/1477153514531518 lighting and daylighting retrofits. Stuttgart (Germany):
Boyce P, Veitch JA, Newsham GR, Jones CC, Heerwagen J, Fraunhofer-Institut für Bauphysik.
Myer M, Hunter CM. 2006. Lighting quality and office work: Iwata T, Shukuya M, Somekawa N, Kimura K. 1992.
two field simulation experiments. Light Res Technol. Experimental study on discomfort glare caused by
38:191–223. doi:10.1191/1365782806lrt161oa windows. J Archit Plan Environ Eng. 432:21–33.
Boyce PR. 2014. Human factors in lighting. 2nd ed. Boca Jakubiec JA, Reinhart CF. 2015. A concept for predicting
Raton (FL): CRC Press. occupants’ long-term visual comfort within daylit spaces.
Bullough JD, Brons JA, Qi R, Rea MS. 2008. Predicting Leukos. 2724:1–18. doi:10.1080/15502724.2015.1090880
discomfort glare from outdoor lighting installations. Karlsen L, Heiselberg P, Bryn I, Johra H. 2015. Verification of
Light Res Technol. 40:225–42. simple illuminance based measures for indication of dis-
Christoffersen J, Wienold J 2004. ECCO-build Report ECCO- comfort glare from windows. Build Environ. 92:615–26.
DBUR-0303-01 Monitoring Procedure for assessment of doi:10.1016/j.buildenv.2015.05.040
user reaction to glare. October:1–70. Kent MG, Altomonte S, Wilson R, Tregenza PR. 2017.
Cronbach LJ, Meehl PE. 1955. Construct validity in psycho- Temporal effects on glare response from daylight. Build
logical tests. Psychol Bull. 52:281–302. Environ. 113:49–64. doi:10.1016/j.buildenv.2016.09.002
De Boer JB. 1967. Visual perception in road traffic and the Kent MG, Fotios S, Altomonte S. 2017. Discomfort glare
field of vision of the motorist. In Public Lighting. Edited evaluation: the influence of anchor bias in luminance
by JB de Boer. Eindhoven (The Netherlands): Philips adjustments. Light Res Technol. 1–16. doi:10.1177/
Technical Library. 1477153517734280
126 A. C. ALLAN ET AL.

Kent MG, Fotios S, Altomonte S. 2018. Order effects when Tuaycharoen N, Tregenza PR. 2007. View and discomfort
using Hopkinson’s multiple criterion scale of discomfort glare from windows. Light Res Technol. 39:185–98.
due to glare. Build Environ. 136:54–61. doi:10.1016/j. doi:10.1177/1365782807077193
buildenv.2018.03.022 Van Den Wymelenberg K. 2014. Visual comfort, discomfort
Kline P. 1986. A handbook of test construction: glare, and occupant fenestration control: Developing a
Introduction to psychometric design. New York (NY): research agenda. Leukos. 10:207–21. doi:10.1080/
Methuen & Co. 15502724.2014.939004
Mahić A, Galicinao K, Van Den Wymelenberg K. 2017. A pilot Van Den Wymelenberg K, Inanici M. 2014. A critical investiga-
daylighting field study: testing the usefulness of tion of common lighting design metrics for predicting human
laboratory-derived luminance-based metrics for building visual comfort in offices with daylight. Leukos. 10:145–64.
design and control. Build Environ. 113:78–91. doi:10.1016/ doi:10.1080/15502724.2014.881720
j.buildenv.2016.11.024 Van Den Wymelenberg K, Inanici M. 2016. Evaluating a new
Manav B, Yener C. 1999. Effects of different lighting arrange- suite of luminance-based design metrics for predicting
ments on space perception. Archit Sci Rev. 42:43–47. human visual comfort in offices with daylight. Leukos.
doi:10.1080/00038628.1999.9696847 12:113–38. doi:10.1080/15502724.2015.1062392
Newsham G, Brand J, Donelly C, Veitch JA, Aries M, Veitch JA, Charles KE, Farley KMJ, Newsham GR. 2007.
Charles K. 2009. Linking indoor environment conditions A model of satisfaction with open-plan office conditions:
to job satisfaction: a field study. Build Res Inf. COPE field findings. J Environ Psychol. 27:177–89.
37:129–47. doi:10.1080/09613210802710298 doi:10.1016/j.jenvp.2007.04.002
Osterhaus W, Bailey IL 1992. Large area glare sources and their Veitch JA, Gifford R. 1996. Assessing beliefs about lighting
effect on discomfort and visual performance at computer work effects on health, performance, mood and social behaviour.
stations. In: Proceedings of the 1992 IEEE Industry Environ Behav. 28:446–70.
Applications Society Annual Meeting, 4-9 October. Houston, Veitch JA, Newsham GR 1996. Determinants of lighting
Texas. quality II: research and Recommendations. In: Annual
Rodriguez RG, Garretón JAY, Pattini AE. 2015. Glare and cogni- Meeting of the American Psychological Association.
tive performance in screen work in the presence of sunlight. Toronto, Canada. p. 58. doi:10.1006/jevp.1999.0169
Light Res Technol. 1477153515577851. doi:10.1177/ Veitch JA, Newsham GR. 2000. Exercised control, lighting
1477153515577851 choices, and energy use: an office simulation experiment.
Rodriguez RG, Yamín Garretón JA, Pattini AE. 2017. An J Environ Psychol. 20:219–37.
epidemiological approach to daylight discomfort glare. Veitch JA, Newsham GR, Mancini S, Arsenault CD. 2010.
Build Environ. 113:39–48. doi:10.1016/j.buildenv. Lighting and office renovation effects on employee and orga-
2016.09.028 nizational well-
Singh A, Syal M, Grady SC, Korkmaz S. 2010. Effects of green being. Ottowa (Canada): National Research Council of
buildings on employee health and productivity. Am Canada. doi:10.4224/2037/4532
J Public Health. 100:1665–68 [accessed 2018 Jan 19]. Velds M. 2002. User acceptance studies to evaluate discom-
http://www.ncbi.nlm.nih.gov/pubmed/20634460. fort glare in daylit rooms. Sol Energy. 73:95–-
Tralau B, Dehoff P, Schierz C. 2009. Introduction of the ergo- 103. doi:10.1016/S0038-092X(02)00037-3
nomic Lighting Indicator as an objective measure for lighting Wienold J, Christoffersen J. 2006. Evaluation methods and
quality. In: Experiencing Light 2009 - International Conference development of a new glare prediction model for daylight
on the Effects of Light on Wellbeing. Eindhoven, The environments with the use of CCD cameras. Energy Build.
Netherlands. 38:743–57. doi:10.1016/j.enbuild.2006.03.017

You might also like