Sei sulla pagina 1di 33

Motivation and Emotion, Vol. 16, No.

L 1992

Effects of Indoor Lighting (Illuminance and


Spectral Distribution) on the Performance of
Cognitive Tasks and Interpersonal Behaviors:
The Potential Mediating Role of Positive Affect 1
Robert A. Baron, 2 Mark S. Rea, and Susan G. Daniels
Rensselaer Polytechnic Institute

Three studies examined the effects of key aspects of indoor lighting (il-
luminance, spectral distribution) on the performance of tasks that did not
primarily involve visual processing. It was hypothesized that lighting conditions
which generated positive affect among subjects would influence behavior and
cognition in ways consistent with the findings of plevious research on the in-
fluence of such affect. Results of all three studies" offered partial support for
this hypothesis. In Study 1, male and female subjects" exposed to relatively low
levels of illuminance (150 tux) assigned higher performance appraisals to a
fictitious employee and included a broader range of words in specific word
categories than subjects exposed to relatively high levels of iltuminance (1500
lux). In Study 2, subjects exposed to warm white light reported stronger
preferences for resolving interpersonal conflicts through collaboration and
weaker preferences for resolving conflicts through avoidance than subjects ex-
posed to cool-white light. Additionally, illuminance and spectral distribution
(color) interacted to influence subjects" self-set goals on a clerical coding task.
In Study 3, receipt of a small, unexpected gift and exposure to warm-white
IThis research was supported by funds from the Niagaru Mohawk Energy-Efficient Seed
Research program and from the New York State Energy Research and Development
Authority. Study t was conducted by Susan G. Daniets in partial fulfillment of requirements
for the Master of Science degree. The authors wish to express their sincere appreciation to
Mama Bronfen and Jill Thomley for their aid in collection of the data for Studies 2 and 3
and in statistical analyses. Thanks are also due to Sylvania, Inc., for supplying lamps, and to
Bob Davis and Peter Boyce for their invaluable assistance in several respects.
ZAddress all correspondence, including requests for reprints, to Robert A. Baron, Department
of Managerial Policy & Organization, Lally Management Center, Rensselaer Polytechnic
Institute, Troy, New York 12180-3590.

0146-7239/92/0300-0001506.50/0 © 1992PlenumPublishingCorporation
2 Baron, Rea, and Daniels

light both increased the amount of time subjects were willing to donate as
unpaid volunteers. In addition, in the absence of a gift, subjects volunteered
more time under low than under high iUuminance.

An extensive body of research evidence indicates that the way people


feel can strongly influence the way they think and the way they behave (cf.
Isen, 1987). More specifically, this evidence suggests that even mild shifts
in affect (especially shifts in the direction of increased positive or pleasant
feelings) can alter several aspects of cognition and several forms of inter-
personal behavior. To mention just a few of these effects, it has been found
that positive affect, induced through several different procedures (e.g.,
receipt of an unexpected gift, positive feedback; Carnevale & Isen, 1986;
Isen, Daubman, & Nowicki, 1987) can influence memory (Riskind, 1983),
decision-making (Isen & Means, 1983), risk-taking (e.g., Isen & Geva, 1987;
Isen, Nygren, & Ashby, 1987), and certain aspects of creativity (Isen &
Daubman, 1984; Isen, Johnson, Mertz, & Robinson, 1985). Corresponding-
ly, positive affect has also been found to influence important forms of in-
tel~ersonal behavior such as helping (cf. Isen, t984), aggression (cf. Baron,
1983), and bargaining (Carnevale & Isen, 1986).
As evidence has accumulated for the general and apparently robust
impact of positive affect, a growing number of researchers have sought to
determine whether such effects extend to important aspects of behavior in
work settings (e.g., George, 1989, 1990). A recent review of existing
evidence on this issue (Isen & Baron, 1991) indicates that, in several
respects, they do. For example, positive affect has been found to increase
the ratings assigned to applicants in job interviews (e.g., Baron, 1987), and
in the context of performance appraisals (Cardy & Dobbins, 1986). Similar-
ly, experiencing positive affect seems to shift individuals' strategies for deal-
ing with interpersonal conflict away from avoidance or confrontation and
toward collaboration or compromise (Baron, 1984; Baron, Fortin, Frei,
Hauver, & Shack, 1990).
Changes in individuals' affective states can, undoubtedly, be induced
by many different factors present in work settings. However, one group of
variables that may exert persistent and hence potentially important effects
in this regard involves aspects of the physical environment in which work
occurs. Previous research indicates that factors such as temperature,
humidity, noise, air quality, and crowding exert strong effects upon reported
affective states (cf. Fisher, Bell, & Baum, 1990; Sundstrom & Sundstrom,
1986). For example, very high or very low ambient temperatures induce
strong negative affect relative to more comfortable conditions (cf. Ander-
son, 1989; Baron, 1978). Similarly, certain types of noise - especially noise
Effects of Indoor Lighting 3

that is unpredictable and uncontrollable -- generate strong feelings of an-


noyance (Cohen, Evans, Stokols, & Krantz, 1986).
The present research focused primarily on another aspect of the physical
environment that may also influence the affective states of individuals in work
settings -- lighting. Research on the impact of lighting in work settings has, of
course, continued for many decades (e.g., Blackwell, 1959; Rea, 1986; Weston,
1935, 1945). Determining the effects of illuminance level on productivity was
the initial question posed in the famous "Hawthorne" studies (Roethlisberger
& Dickson, 1941). Most research on lighting, however, has examined the ef-
fects of this variable from what might be termed a visual process perspective.
It has focused primarily on determining levels of illuminance required for the
performance of various tasks (e.g., Boyce, 1981; Boyce, Berman, Collins, Lewis,
& Rea, 1989; Rea, 1988). This work has influenced recommended itlumina_nce
levels for various work settings and tasks -- standards that are now widely ap-
plied (Illuminating Engineering Society, 1987).
More recently, investigators have demonstrated increased interest in
what have been termed the indirect effects of lighting - the impact of lighting
on mood, arousal, and behavior (Belcher & Kluczny, 1987; Biner, 1991; Boyce
et al., 1989; Veitch, Gifford, & Hine, 1991). Several studies have examined
individuals' expressed preferences for different levels of illuminance in various
settings and when performing different activities (Biner, Butler, Riscsher, &
Westergren, 1989; Butler & Biner, I987; Ftynn, 1977). Additional research
has investigated the impact of levels of illuminance on several specks of in-
terpersonal behavior, such as communication (Gifford, 1988).
The present research was designed to add to this expanding body of
knowledge ha two distinct ways. First, it sought to investigate the impact
of two key aspects of lighting - illuminance (the flux density of light on a
surface measured in lux) and spectral distribution (the distribution of energy
at different wavelengths) - on the performance of a wide range of work-
related tasks. These tasks included performance appraisal, word categoriza-
tion, goal-setting, and preferred modes of resolving interpersonal conflicts.
Second, the present research sought to determine whether any impact
of lighting conditions on nonvisual work-related tasks can be understood
in terms of positive affect. Specifically, the research reported here stemmed
from the following reasoning: (1) Variations in lighting along specified
physical dimensions can induce contrasting levels of positive affect; (2) such
lighting-produced differences in positive affect may, in turn, influence
several aspects of cognition and behavior, as noted above.
In the three studies reported below, therefore, participants performed
several different cognitive and interpersonal tasks in the presence of relatively
high or relatively low illuminance (1500 or 150 lux) produced by one of several
types of fluorescent lamp types -- bulbs differing appreciably in spectral dis-
4 Baron, Rea, and Daniels

tribution. On the basis of existing literature on the impact of positive affect,


it was predicted that lighting conditions that generate positive affect would
influence cognition and behavior in a manner consistent with the findings of
previous research on the impact o.f positive affect (cf. Isen, 1987). For example,
lighting conditions that generate positive affect should raise evaluations of
neutral stimuli (Isen & Shalker, 1982), evaluations of others' performance
(Baron, 1987), and perhaps raise self-set goals (Locke & Latham, 1990) rela-
tive to lighting conditions that do not generate positive affect or generate lower
levels of such affect. Similarly, lighting conditions that generate positive affect
should lead indMduals to assign more nontypical exemplars of a given verbal
category to that category than lighting conditions that do not generate positive
affect or generate it to a lesser degree (Isen & Daubman, 1984).
Relatively little direct evidence concerning the affective impact of various
types of lighting exists. Several studies suggest that exposure to light of very
high illuminance (10,000 lux) can counter the effects of seasonal affective dis-
order (SAD) - depression experienced by approximately 5% of the adult
population during the winter months (Blehar & Rosenthal, 1989; Rosenthal,
1985). However, such findings remain somewhat controversial, and have not
been consistently verified (Kasper et al., 1989). In contrast, several additional
studies suggest that within a more normal range o f illuminance levels, in-
dividuals tend to report more positive reactions to relatively dim than to rela-
tively bright conditions (e.g., Nelson, Nilsson, & Johnson, 1984). Turning to
spectral distribution, existing evidence is also meager, but suggests that respon-
dents generally report more positive reactions to "warm" light sources (ones
with relatively more red, yellow, and orange hues) than to "cool" light sources
(ones with relatively more green or blue hues; cf. Flynn & Spencer, 1977).
Taking available evidence into account, therefore, it was predicted that subjects
would experience higher levels of positive affect in response to relatively low
(150 lux) illuminance and warm light sources than to relatively high (1500 lux)
illuminance and cool light sources. It should be emphasized, however, that
given the relative lack of pertinent evidence, these predictions were only ten-
tative in nature.

STUDY 1

Method

Subjects and Design

Subjects were 64 male and 27 female undergraduates enrolled in sec-


tions of introductory psychology at Rensselaer Polytechnic Institute. They
Effects of Indoor Lighting 5

participated in the study voluntarily, as one of several options for complet-


ing a course requirement.
The study employed a 2 x 4 factorial design based on two levels of
itluminance (approximately 150 lux, 1500 lux on the surface of the desk
where experimental tasks where performed) and four lamp types [warm
white (3000K), natural white (3600K), cool white (4200 K), and Design 50
(D50, Sylvania Trademark) (5000K)]. Physical descriptions of these lamps
may be found in the IES Handbook (Illuminating Engineering Society,
1987). Subjects were randomly assigned to experimental conditions.

Apparatus

The research was conducted in a specially prepared room ap-


proximately 3.22 m x 3.35 m in size (height = 2.45 m) in which all walls
and the ceiling were off-white in color. The only window was covered by
a light-blocking drape. Lighting was provided by four-lamp recessed fluores-
cent luminaries (lighting fixtures) with flat prismatic lenses. Illuminance was
varied by increasing or reducing the number of lamps lit, while color was
varied by changing the lamps in the ceiling fixtures. All lamps were new
when the study began, and were operated for 100 hours to stabilize the
resulting levels of illuminance.

Procedure

Overview. Subjects were informed that they would be participating in


a research project concerned with the role of individual differences in
decision-making. During the first phase of the session, they first performed
a brief time-filler task (completing questionnaires). Then, they performed
various cognitive tasks (see below) presented in counterbalanced order.
Finally, participants completed a measure of their current affective states
and evaluated various aspects of the experimental room, including lighting,
on several different dimensions. All of these procedures took place under
one of the eight lighting conditions described above.
Tasks Performed by Subjects. One of the tasks completed by subjects
involved reading the personnel folder of an imaginary employee. This
folder contained information found, in pilot research, to be quite neutral
in content (e.g., the folder contained equal amounts of positive and nega-
tive information about the employee). After reading all the information in
the folder, subjects evaluated the employee (described as a senior secretary)
on several different dimensions (qualifications for current job, job-related
skills, competence, motivation, current job performance, ability to get along
6 Baron, Rea, and Daniels

well with others, intelligence, whether this person should be hired if not
already employed, and overall rating). All ratings were made on 7-point
scales.
A second task involved word categorization (Isen & Daubman, 1984).
Subjects were presented with lists of exemplars for each of four different
categories: vehicle, furniture, clothing, vegetable. Their task was to rate each
word in terms of the extent to which it seemed to belong of not belong to
a given category. For example, for the category vegetable, exemplars included
string beans, celery, potato, pickles, and seaweed. Previous research indicates
that the last two of these words are rated by most individuals as being less
closely related to this category (i.e., they are poorer exemplars) than the
others (Rosch, 1975).
Lighting Conditions. When subjects entered the experimental room,
one of the eight lighting conditions was already present. Illumination on
the work surface was varied so that it was either 150 tux in the low-il-
luminance condition or 1500 lux in the high-illuminance condition.3 Sub-
jectively, the 150-1ux condition afforded lighting perhaps best described as
more subdued than that found in most office settings, but not excessively
dim. In contrast, the 1500-1ux condition afforded lighting best described as
somewhat brighter than that found in most offices, but, again, not exces-
sively so. Typical illuminance conditions in many offices are in the 300- to
450-1ux range. Measures of illuminance were taken in four separate areas
of the room by means of a calibrated Hagner Model $2 photometer. These
readings indicated that distinct and stable levels of illuminance were indeed
attained in the two conditions. Lamps were changed several times each day
to vary lighting color (spectral distribution). Lamps were operated a min-
imum of 1 hour prior to each session to assure a relatively constant level
of illuminance during the collection of data. Other environmental variables
(temperature, humidity) were held constant across the various lighting con-
ditions (i.e., air circulation through a combined heating-air conditioning
system was sufficient to control for any fluctuations in temperature that
might otherwise have been produced by differences in illuminance level).
Measure of Affect. After completion of the three tasks described
above, subjects completed the Positive Affect-Negative Affect Scale
(PANAS; Watson, Clark, & Tellegen, 1988). On this scale, respondents
rate the extent to which various adjectives (e.g., excited, enthusiastic)
describe their current feelings. Ratings are made on 5-point scales.
Finally, subjects evaluated the experimental room on several different
dimensions. Some of these dimensions involved general affective reactions

3The 150-1uxconditionwas produced by means of "phantom lamps." These are lamps that
permit the passage of an electric current, but do not emit light.
Effects of Indoor Lighting 7

Table I. Means and Standard Deviations for Performance Ratings


as a Function of Illuminance Level (Study 1)a
Low High
illuminance illuminance
Job qualifications 3.58a 3.33a
(0.79) (0.50)
Job-related skills 5.11a 4.50b
(0.82) (0.87)
Competence 4.41a 3.89b
(0°78) (0.96)
Motivation 3.70a 3.37a
(0.68) (0.82)
Current performance 3.83a 3.78a
(1.15) (0.83)
Ability to get along with others 3.68a 3.70a
(0.99) (0.95)
Intelligence 4.08a 3.90a
(0.79) (0.96)
Should applicant be hired 4.33a 3.59b
(1.09) (1.01)
Overall rating 4.59a 3.70b
(0,66) (0.78)
aNote: For each dependent measure, means that do not share a com-
mon subscript differ significantly (p < .05) by the Duncan Multiple-
Range Test. Higher numbers indicate more favorable ratings on each
dimension. (Range = 1-7.) Numbers in parentheses are cell standard
deviations.

to this e n v i r o n m e n t (e.g., u n a t t r a c t i v e - a t t r a c t i v e , p l e a s a n t - u n p l e a s a n t , l i k e -
dislike) w h i l e o t h e r s r e f e r r e d m o r e d i r e c t l y to t h e l i g h t i n g ( d i m - b r i g h t ,
colorless-colorful, glare-nonglare, clear-hazy).

RESULTS

Preliminary analyses indicated that there were no gender differences


on any of the dependent measures. Consequently, data from mate and
f e m a l e p a r t i c i p a n t s w e r e c o m b i n e d in all f u r t h e r a n a l y s e s . A d d i t i o n a l
a n a l y s e s i n d i c a t e d t h a t t h e r e w e r e n o s i g n i f i c a n t effects o f o r d e r o f t a s k
presentation or interactions between order and the independent variables.
8 Baron, Rea, and Danieis

Table II. Mean Numbers of and Standard Deviations of Nontypical Exemplars Included as
Category Members, as a Function of Illuminance Level and Color (Lamp Type) (Study 1)a
Color Warm white Naturalwhite Cool white D50
Low illuminance 4.55a 4,55a 3.18a 4.45a
(0.87) (0.93) (1.01) (0.97)
High illumiance 2.18b 2.55b 2.73a 3.09a
(0.68) (0.81) (1.13) (1.04)
aNote: For each color condition, means that do not share a common subscript differ sig-
nificantly (p < .05) by the Duncan Multiple-RangeTest. (Range = 1-12.) Numbers in paren-
theses are cell standard deviations,

Simulated Performance Appraisal

A multivariate analysis was performed on the data for the nine rating
scales on which subjects evaluated the imaginary employee. This analysis
yielded a significant effect for illuminance, F(9, 70) = 2.33, p < .03, and
a significant interaction between color (lamp type) and illuminance, F(27, 205)
= 1.71, p < .02, Wilks's Iambda criterion. Followup univariate analyses
indicated that the significant effect for illuminance stemmed from the fact
that subjects rated the employee significantly higher on several different
dimensions [e.g., secretarial skill, competence, overall hiring recommenda-
tion in the low-illuminance (150 tux) condition than in the high-illuminance
(1500 lux) condition (p < .05 in all cases)]. The means for these items are
shown in Table I. It should be noted that in previous research (e.g., Baron,
1987), subjects experiencing positive affect have assigned more favorable
evaluations to ratees than individuals not experiencing positive affect.
Followup univariate analyses indicated that the significant interaction
between color and illuminance stemmed primarily from ratings on two
items, not qualified-qualified, responsible-irresponsible, F(3, 78) = 3.24,
2.99, p < .03, .05, respectively. The pattern of the interaction was identical
in both cases, and indicated that, for all lamp types except the cool white,
ratings were higher in the low-illuminance than in the high-illuminance con-
dition. This pattern was reversed for the cool-white condition.

Word Categorization

Following procedures employed by Isen and Daubman (1984), an


analysis of variance was performed on the number of nontypical exemplars
rated as being category members (assigned a rating of 6 or above on the
10-point scale). This analysis yielded a significant effect for illuminance,
Effects of Indoor Lighting 9

F(1, 80) = 9.75, p < .002. As shown in Table II, subjects in the low-il-
luminance condition generally assigned higher ratings to the poor ex-
emplars than those in the high-illuminance condition. However, despite the
absence of an interaction between illuminance and color, followup com-
parisons (by the Duncan Multiple-Range Test) indicated that differences
between the low- and high-illuminance conditions were significant only in
the case of the warm-white and natural-white lamp conditions.
An additional analysis was performed on the ratings assigned to the
twelve poor exemplars. This analysis, too, yielded a significant main effect
for illuminance, F(1, 80) = 5.85, p < .03. Subjects in the low-illuminance
conditions assigned higher mean ratings to poor exemplars (e.g., elevator
for the category vehicle, telephone for the category furniture; overall M =
4.26) than subjects in the high-illuminance condition (overall M = 3.64).
In previous research, subjects experiencing positive affect have assigned
higher ratings to nontypical exemplars of a given category than subjects
not experiencing positive affect.

Room and Lighting Evaluation

A multivariate analysis of variance ( M A N O V A ) was performed on


subjects' ratings of the room and its lighting. This analysis yielded a sig-
nificant effect for level of illuminance, F(13, 68) = 4.06, p < .001 (by
Wilks's lambda criterion), and a significant effect for color (lamp type),
F(39, 202) = 1.69, p < .01. Followup univariate analyses indicated that, as
expected, subjects rated the room as being brighter in the 1500-tux condi-
tion (3// = 5.82) than in the 150-tux condition (3// = 4.11), F(1, 80) =
43.54, p < .001. Similarly, those in the 1500-1ux condition rated the room
higher in clarity (M = 3.07) than those in the 150-1ux condition (M = 1.77),
F(1, 80) = 25.59, p < .001.
With respect to color, followup univariate tests indicated that subjects
in the cool-white condition rated the room as less relaxing (M = 4.27) than
those exposed to the other three lamps (M = 3.09, 3.05, 3.41, respectively,
for the warm-white, natural-white, and D50 bulbs), F(3, 80) = 3.84, p <
.01. Conversely, subjects rated the room as being more pleasant in the
warm-white condition and as being least pleasant in the cool-white condi-
tion, F(3, 80) = 2.79, p < .05. Subjects in the warm-white condition also
evaluated the room as less glaring (M = 5.50) than those in the D50 con-
dition [M = 4.05; F(3, 80) = 3.88, p < .01].
Taken as a whole, findings for the room evaluation measure suggest
that subjects in the different lighting conditions did in fact have contrasting
perceptions of the experimental room and its lighting, despite the fact that
10 Baron, Rea, and Danieis

they were exposed to only one of these conditions in the between-subjects


design employed. Further, there was some indication that the low-il-
luminance condition induced more favorable affective reactions from sub-
jects than the higher-illuminance condition and that, at least in some
respects, the warm-white lamps induced somewhat more favorable reactions
than the cool-white lamps.

Positive and Negative Affect Scale

Separate analyses of variance were performed on the scores for posi-


tive and negative affect derived from the PANAS scale (Watson et al.,
1988). No significant effects emerged from these overall analyses. Thus,
there was no evidence, on scores for this measure, that subjects experienced
differential affective reactions to the various lighting conditions.

DISCUSSION

The results of this experiment suggest that variations in indoor light-


ing can influence performance on several work-related tasks - tasks that
can be viewed as primarily cognitive rather than visual in nature. Subjects'
performance on several tasks was significantly affected by illuminance and,
to a somewhat lesser degree, by lamp color. With respect to illuminance,
it was found that subjects exposed to a relatively low level of itluminance
(150 lux) assigned higher ratings to an imaginary employee and included
more nontypical exemplars in specific word categories than subjects ex-
posed to relatively high levels of illuminance (1500 lux). These findings
directly parallel those obtained in previous studies concerned with the ef-
fects of positive affect (cf. Isen & Baron, 1991). In those studies it was
found that subjects induced, through various procedures, to experience
positive affect also assigned higher ratings to strangers (cf. Baron, 1987)
and included more nontypical exemplars in categories (Isen & Daubman,
1984) than subjects not induced to experience positive affect. Thus, subjects
exposed to relatively low levels of illuminance behaved, in several respects,
like those experiencing positive affect in previous research.
It should be noted, of course, that only two levels of illuminance were
employed. Thus, it cannot be determined, with any degree of certainty,
whether low levels of iUuminance reduced scores on the various dependent
measures collected or whether high levels of iUuminance increased them.
Further research in which additional, intermediate levels of illuminance are
Effects of Indoor Lighting I1

employed (e.g., Veitch et al., 1991) must be conducted before any firm
conclusions in this respect can be reached.
Results with respect to color (lamp type) were less consistent. How-
ever, a significant interaction between illuminance and color was obtained
for the performance-appraisal task. This stemmed from the fact that sub-
jects exposed to the cool-white lamp failed to rate the employee higher in
the 150-1ux than in the 1500-1ux condition; subjects exposed to all other
lamp types showed this difference. Thus, exposure to the cool-white lamp
seemed, in a sense, to counter or reduce the positive effects induced by
low levels of illuminance.
To recapitulate, results indicate that (1) lighting influences perfor-
mance on several diverse tasks, and (2) the pattern of these effects closely
resembles that observed in previous research concerned with the influence
of positive affect (e.g., Fiedter & Forgas, 1988; Isen, 1987). Whether the
effects of illuminance and lamp color on participants' behavior were ac-
tually mediated by positive affect, however, remains uncertain. Some find-
ings are consistent with such an interpretation. For example, subjects
ex]~osed to warm-white light reported more favorable reactions to the ex-
perimental room along some dimensions than did those in the cool-white
condition (e.g., higher ratings on the dislike-like dimension). However,
findings for the PANAS did not provide confirmation for this suggestion.
Several factors may have contributed to the absence of significant results
for this measure.
First, of course, it is possible that the lighting conditions employed
did not produce differential levels of positive and negative affect among
subjects. Perhaps discernible shifts in affective state are produced only
when illuminance and color are varied over a somewhat wider ranger than
that employed here (cf. Biner, 1991).
Second, it is possible that variations in lighting did induce shifts in
subjects' affective states, but that these shifts were too small or subtle for
them to recognize and report. (Recall that the study employed a between-
subjects design in which participants were exposed to only one lighting con-
dition.) Considerable evidence suggests that human beings have only a
limited capacity for monitoring and verbally describing subtle shifts in af-
fective states (e.g., Nisbett & Wilson, 1977). It seems possible that such
limitations played a role in the present findings.
Third, it is possible that the instrument used to measure positive and
negative affect (the PANAS) was not sufficiently sensitive to assess shifts
in affective state produced by the different lighting conditions. While the
PANAS has been used to measure affect in several previous studies, its
suitability for measuring modest environmentally induced shifts in affect
such us those considered here remains to be established.
12 Baron, Rea, and Danieis

Finally, in this investigation, participants completed the PANAS only


after performing various tasks. It is possible that any effects of lighting
conditions on affective states were masked or reduced by this fact.
In sum, the present findings suggest that variations in lighting can
indeed influence performance on several tasks. Moreover, these effects are
highly similar, in several respects, to those observed among individuals in-
duced to experience positive affect through a wide range of procedures.
The potential mediating role of positive affect in these results, however,
remains uncertain. In order to address this important issue, and to obtain
further evidence on the impact of lighting on work-related tasks, two ad-
ditional studies were conducted.

STUDY 2

Considerable evidence exists for the contention that the PANAS


provides valid and reliable measures of positive and negative affect (cf.
Watson et al., 1988). However, comments by several participants in Study
1 indicated that persons for whom English is a second language are un-
certain of the meaning of some of the adjectives it employs (e.g., enthusias-
tic, jittery, inspired, distressed). In view of this possibility, it was reasoned
that a measure of current affect that employs somewhat simpler terms
might provide additional evidence on the potential mediating effect of posi-
tive affect with respect to the impact of indoor lighting. In this study, there-
fore, subjects were asked to complete a questionnaire consisting of items
similar to those successfully employed to measure current affect in several
previous experiments (e.g., Baron, 1987; Baron et al., 1990).
A second purpose of this experiment was that of extending (and hence
conceptually replicating) the results obtained in Study 1. In the initial study,
attention was directed primarily to what might be termed cognitive tasks
-- decision-making, performance appraisals. In this followup investigation,
the focus was shifted to the potential impact of lighting on important forms
of interpersonal behavior. Previous research indicates that positive affect
can influence many aspects of such behavior, from helping on the one hand
(Isen, 1970, 1987) to aggression on the other (e.g., Baron, 1983). Recent
research has extended this finding to the topic of organizational conflict
(cf. Baron, 1990). Several studies have reported that exposure to conditions
serving to induce positive affect among subjects (e.g., humorous materials,
a small gift, flattery) enhances their tendency to seek integrative rather
than distributive resolutions to ongoing conflicts (Baron, 1984; Baron et
al., 1990; Pruitt & Rubin, 1986). Subjects in whom positive affect has been
induced offer more concessions to opponents and express stronger
Effects of Indoor Lighting 13

preferences for resolving conflicts through collaboration and compromise


(but weaker preferences for resolving conflicts through avoidance and com-
petition) than subjects in whom positive affect has not been induced (cf.
Baron et al., 1990; Carnevale & Isen, 1986). The present study sought to
determine whether certain types of lighting could induce similar effects.
Specifically, it examined the hypothesis that subjects exposed to lighting
conditions that induce positive affect (low illuminance; warm-white light)
would demonstrate patterns of behavior similar to those observed in the
previous research just described.
A final purpose of this study was examining the impact of lighting con-
ditions on self-set goals and related processes (cf. Locke & Latham, t990).
Previous findings indicate that individuals experiencing negative affect tend to
set lower goals on various tasks and to express less confidence in attaining
them than those in a neutral or more positive state (cf. Baron, 1988, 1990).
Extending these findings to the present context, it was hypothesized that sub-
jects exposed to low-illuminance, warm-white light would set higher goals on
a clerical coding task and express higher confidence in their ability to perform
this task than subjects exposed to other lighting conditions.

Method

Subjects and Design

Subjects were 59 male and 13 female undergraduates enrolled in sections


of introductory psychology at Rensselaer Polytechnic Institute. They par-
ticipated in the study voluntarily in order to complete a course requirement. 4
A 2 x 2 factorial design based on two levels of illuminance (150 lux, 1500 lux)
and two lamp types (warm white, cool white) was employed. Subjects were
randomly assigned to each of these four conditions as they appeared for their
appointments. Apparatus was identical to that employed in Study i.

Procedure

Lighting Conditions. Subjects were exposed to one of four lighting con-


ditions based on low (150 lux) or high (1500 lux) illuminance and warm-
white or cool-white fluorescent bulbs.

4Because males constitute more than 80% of the student population from which participants
were drawn, it was impossible, for practical reasons, to obtain equal numbers of male and
female subjects.
14 Baron, Rea, and Daniels

Tasks Performed by Subjects. On entering the laboratory, subjects first


completed an initial measure of current affect (Current Feelings Survey).
This form requested that they rate their current feelings along 10 dimen-
sions (sad-happy, bad-good, negative-positive, unpleasant-pleasant, tired-
energetic, anxious-calm, dull-alert, irritable-relaxed, sleepy-awake,
bored-interested). The items on this questionnaire were highly similar to
those used in many previous studies (cf. Baron, 1987). Next, subjects com-
pleted the Thomas-Kilmann MODE Scale (Thomas & Kilmann, 1971).
This instrument provides separate measures of five distinct personal styles
of resolving interpersonal conflict: avoidance, compromise, collaboration,
accommodation, competition (cf. Thomas, 1992).
Subjects were then asked to perform a clerical coding task employed
in several previous studies (e.g., Baron, 1988). This task involved several
steps (e.g., finding the prices of coded items in an array of codes, entering
the price of each item, determining a discount). Before actually performing
the task, subjects indicated how many items they would complete (a self-set
goal), and rated their confidence in attaining that goal, their ability to per-
form this task, and the amount of effort they expected to expend on it.
(All ratings were made on 7-point scales.) After completing these ratings,
participants performed the clerical coding task for 5 rain.
An additional task asked subjects to consider a conflict situation --
one in which another person in their company had failed to complete some
work by a promised date. Subjects read three versions of this incident,
each presenting a different explanation for the person's failure to meet the
deadline. In one version, the cause for this failure was ambiguous; in
another, it stemmed from the fact that this person was overworked (an
external cause); in the third, it stemmed from irresponsibility on the part
of this person (an internal cause). After reading each incident, subjects
rated (on 7-point scales) the likelihood that they would follow each of five
distinct courses of action. These were the five basic patterns of conflict
resolution described by Thomas (1992) and measured by the MODE
(Thomas & Kilmann, 1971): compromise, accommodation, competition,
avoidance, collaboration. Compromise involves splitting any differences
down the middle. Accommodation, in contrast, involves surrendering to the
demands of one's adversary -- giving in to this person's demands. Com-
petition reflects a "win-lose" approach in which individuals attempt to max-
imize their own outcomes at the expense of their opponents. Avoidance
involves walking away from the situation - refusing to participate in it.
Finally, collaboration reflects a "win-win" or integrative approach to con-
flict. Individuals adopting this approach attempt to maximize the joint out-
comes for themselves and the other party to the conflict (cf. Thomas, 1992).
Effects of Indoor Lighting 15

Table IIL Mean Reported Affect and Standard Deviations as a Function of Lighting Con-
ditions (Study 2)a
Lighting Cool white, Cool white, Warm white, Warmwhite,
condition 1500 lux 150 lux 1500 lux 150 lux
Affect dimension
Anxious-calm 4.67ab 4.06ab 3.94a 4.9%
(1.09) (0.78) (1.78) (1.37)
Sleepy-awake 4.56bc 3.83ab 3.39a 4.28b~
(1,56) (1.50) (0.78) (1.08)
aNote: For each affect dimension (i.e., dependent measure), means that do not share a com-
mon subscript differ significantly (p < .05) by the Duncan Multiple-Range Test. (Range =
1-70 Higher numbers indicate greater tendencies to report feeling calm and awake. Numbers
in parentheses are cell standard deviations.

At this point subjects completed a second measure of affect (the Cur-


rent Feelings Survey). Finally, they completed a questionnaire designed to
assess their subjective reactions to the lighting in the experimental room
(dim-bright, cold-warm, not glaring-glaring, hazy-clear), and the extent to
which lighting in the room was similar to that in various locations (home,
offices, restaurants, classrooms, stores, etc.).

Results

Because of the small number of female participants, it was not pos-


sible to analyze for differences due to gender. However, removal of all data
for females did not alter the obtained findings in any respect. For this
reason, data for both sexes were combined in all subsequent analyses.

Measures of Current Affect

A multivariate analysis of variance performed on the first measure


of affect yielded no significant effects. H o w e v e r , a c o r r e s p o n d i n g
analysis performed on the data for the second affect measure yielded
an interaction between color and illuminance that closely approached
significance, F(10, 59) = 1.98, p < .06. Although the overall interaction
was not significant, in view of the exploratory nature of the current re-
search it seemed appropriate to consider the univariate analyses per-
f o r m e d concurrently with the multivariate analysis. These univariate
analyses indicated that the interaction between color and illuminance
was significant for two dimensions: anxious-calm, sleepy-awake, F(1, 59)
= 5.97, 4.34, p < .02, .05. In both cases, the form of the interaction
16 Baron~ Rea, and Daniels

Table IV. Mean Scores and Standard Deviations on Scales of the MODE Scale as a Func-
tion of Lighting Conditions (Study 2)a
Lighting Cool white, Cool white, Warm white, Warm white,
condition 1500 lux 150 lux 1500 lux 150 lux

MODE Scale
Collaboration 5.56a 5.06a 6.00at 7.28c
(1.13) (1.57) (1.83) (1.78)
Avoidance 6.56b 7.78c 6.4% 4.67a
(1.04) (1.62) (1.23) (1.42)
Competition 5.84a 5.5% 4.87a 4.92a
(1.21) (1.38) (1.42) (1.30)
Cooperation 6.00a 6.33a 6.50a 7,08a
(1,06) (1.00) (1.10) (1.13)
Accommodation 5.47a 5.71a 6.75a 6.35a
(0.99) (1.0t) (1.34) (1.26)
Note: For each dependent measure, means that do not share a common subscript differ sig-
nificantly (p < .05) by the Duncan Multiple-Range Test. (Range = 0-12.) Higher numbers
indicate greater reported likelihood of using each mode of conflict resolution. Numbers in
parentheses are cell standard deviations.

Table V. Mean Ratings and Standard Deviations of Ability


to Perform the Clerical Coding Task as a Function of II-
luminance and Colora
Warm Cool
Color white white

Low illuminance 6.06a 5.21b


(1.38) (0.95)
High illuminance 5.28b 5.89b
(0.99) (0.81)
aNote: Means that do not share a subscript differ significantly Q9
< .05) by the Duncan Multiple-Range Test. Higher numbers
indicate greater reported ability to perform the task. (Range
= 1-7.) Numbers in parentheses are all standard deviations.

w a s t h e s a m e : in t h e c o o l - w h i t e c o n d i t i o n , t h e 150- a n d 1500-1ux c o n -
d i t i o n s d i d n o t d i f f e r s i g n i f i c a n t l y . H o w e v e r , in t h e w a r m - w h i t e c o n d i -
t i o n , t h e 1 5 0 - a n d 1500-1ux c o n d i t i o n s d i f f e r e d s i g n i f i c a n t l y o n b o t h
m e a s u r e s . S u b j e c t s r e p o r t e d b e i n g m o r e c a l m a n d m o r e a w a k e in t h e
150-1ux t h a n in t h e 1500-1ux c o n d i t i o n ( r e f e r to T a b l e I I I ) .
Effects of Indoor Lighting 17

Table Vl. Mean Reported Preferences and Standard Deviations for Five Modes of Con-
flict Resolution as a Function of Apparent Causes of Conflicta
Apparent cause of conflict
Ambiguous External Internal
Compromise 4.398 5.638 2.76b
(1.65) (1.20) (2.01)
Accommodation 2.618 4.76b 1.29b
(1.09) (1.03) (0.59)
Competition 3.33a 2.29a 5.82b
(1.49) (1.31) (1.13)
Avoidance 3.33a 4.00a 3.71a
(1.57) (1.80) (2.39)
Collaboration 5.398c 5.82a 4.24bc
(1.09) (0.88) (2.19)
aNote: For each mode of conflict resolution, means that do not share a common subscript
differ significantly (p < .05) by the Duncan Multiple-Range Test. (Range = 1-7.) Higher
numbers indicate greater preference for each conflict resolution mode. Numbers in paren-
theses are cell standard deviations.

MODE Scores

Separate analyses of variance were p e r f o r m e d on subjects' scores on


each o f the five dimensions o f the M O D E . Significant effects were f o u n d
for two o f these dimensions: collaboration and avoidance. With respect to
collaboration, the effect o f color was significant, F(1, 68) = 6.42, p < .02.
This finding reflected the fact that subjects r e p o r t e d stronger tendencies
to resolve interpersonal conflicts t h r o u g h collaboration in the warm-white
(M = 6.63) than in the cool-white condition (M = 5.31).
With respect to avoidance, b o t h the main effect of color, F(1, 68) =
6.91, p < .02, and the interaction o f color and illuminance, F(1, 68) =
5.99, p < .02, were significant. T h e main effect for color derived f r o m the
fact that subjects r e p o r t e d w e a k e r tendencies to resolve interpersonal con-
flicts t h r o u g h avoidance in the warm-white (M = 5.55) than in the cool-
white ( M = 7.17) condition. T h e interaction between color and illuminance
again reflected the fact that the w a r m white-low illuminance g r o u p differed
significantly" f r o m all the others; subjects in this g r o u p r e p o r t e d the weakest
t e n d e n c y to resolve conflicts t h r o u g h avoidance (M = 4,67). M e a n s for the
interactions between color and illuminance for all five M O D E dimensions
are presented in Table IV.
18 Baron, Rea, and Daniels

Goal-Setting Measures

Before performing the clerical coding task, subjects were asked to indi-
cate how many items they thought they would complete. While subjects in the
low-illuminance, warm-white condition set higher goals than those in any other
group (M = 16.53), an analysis of variance on these data failed to yield any
significant effects. However, analyses on the data for subjects' ratings of their
ability to perform the clerical task yielded a significant interaction between
color and illuminance, F(1, 67) = 8.93, p < .005. The means involved in this
interaction are shown in Table V. As can be seen from this table, in the warm-
white condition, subjects reported significantly higher ability in the low-il-
luminance than in the high-illurninance condition (M = 6.06 vs. 5.28, p <
.05). In the cool-white condition, however, the difference between the low-
and high-illuminance groups was not significant (M = 5.21 vs. 5.89).
There were no differences in actual performance on the clerical
coding task as a function of either color F(1, 66) = .27, p > .20, or il-
luminance F(1, 66) = .08, p > .20, nor was the interaction between these
variables significant, F(1, 66) = 1.70, p < .20.

Conflict Situation

A multivariate analysis of variance was performed on the data for the


hypothetical conflict situation (in which a co-worker had failed to provide
promised work by an agreed-upon deadline). This analysis yielded only one
significant a f f e c t - that for the apparent cause behind this failure (am-
biguous, excessive workload, laziness), F(14, 924) = 40.72, p < .001. Subjects
reported the most negative reactions (greater preferences for dealing with
the situation through competition, weaker preferences for handling it through
collaboration or compromise) when the delay stemmed from laziness than
when it stemmed from an external cause (too much work) or the cause of
the delay was ambiguous (see Table VI). Interestingly, for several dependent
measures (compromise, competition, collaboration) subjects' reactions in the
ambiguous and "overworked" (i.e., external cause) conditions were more
similar to each other than either of these conditions was to the "lazy" (in-
ternal cause) condition. No significant effect for color or illuminance were
obtained in this analysis.

Subjective Reactions to Lighting Conditions

A multivariate analysis of variance was performed on subjects'


ratings of lighting in the experimental room. This analysis yielded a sig-
Effects of Indoor Lighting 19

Table VII. Mean Ratings and Standard Deviations of Similarity of Lighting Conditions in
the Experimental Room to Those in Various Locations (Study 2)a
Business
Location Home Restaurant office Hospital
Low illuminance 2.92 2.40 4.77 4.51
(0.43) (0.52) (0.7t) (0.68)
High illuminance 2.17 1.90 5.58 5.47
(0.61) (0.38) (0.68) (0.89)
aNote: For each location, differences between the low- and high-illuminance condition are
significant (p < .05) by the Duncan Multiple-Range Test. Higher numbers indicate greater
reported similarity to the locations named. (Range = I-7.) Numbers in parentheses are cell
standard deviations.

nificant effect for illuminance, F(4, 65) = 9.12, p < .001. Followup
univariate analyses indicated that this effect s t e m m e d primarily f r o m
ratings along the d i m - b r i g h t and hazy-clear dimensions, F(1, 68) = 34.97,
6.08, p < .001, < .02, respectively. As expected, subjects rated the high-
i l l u m i n a n c e c o n d i t i o n as b r i g h t e r a n d m o r e c l e a r t h a n t h e low-il-
l u m i n a n c e c o n d i t i o n . T h e e f f e c t o f i l l u m i n a n c e also a p p r o a c h e d
significance for the not glaring-glaring dimension, F(1, 68) = 3.46, p <
.07. Subjects rated the r o o m as m o r e glaring in the high-illuminance than
in the tow-illuminance condition.
Additional items on the final questionnaire asked subjects to indi-
cate the extent to which lighting conditions in the e x p e r i m e n t a l r o o m
w e r e s i m i l a r to t h o s e in o t h e r settings. A m u l t i v a r i a t e analysis o f
variance on these data yielded a highly significant effect for itluminance,
F(7, 61) = 3.02, p < .002. Followup univariate analyses indicated that
this e f f e c t s t e m m e d p r i m a r i l y f r o m ratings of similarity to subjects'
h o m e , businesses offices, restaurants, and hospitals. As expected, sub-
jects in the low-illuminance condition rated lighting in the e x p e r i m e n t a l
r o o m as m o r e similar to that in their h o m e s and in restaurants than did
subjects in the high-illuminance condition. 5 Conversely, subjects in the
high-illuminance condition rated lighting in the experimental r o o m as
m o r e similar to businesses offices and hospitals than did subjects in the
low-illuminance condition (refer to T a b l e VII).

5Even in the low-illuminance condition, however, subjects rated lighting conditions as below
the neutral point on the scale of similarity. Thus, in an absolute sense, they did not perceive
lighting conditions in the experimental rooms as very similar to those in their homes or
restaurants.
20 Baron, Rea, and Daniels

Discussion

The results of this experiment extend those of Study 1 in several


respects. First, they indicate that variations in illuminance level and color
can influence important forms of interpersonal behavior and certain aspects
of goal-setting as well as performance on tasks that are primarily cognitive
in nature. Subjects reported significantly stronger tendencies to resolve in-
terpersonal conflicts through collaboration, but weaker tendencies to resolve
them through avoidance, in the warm-white than in the cool-white condi-
tion. Further, with respect to avoidance, an interaction between color and
illuminance indicated that both tendencies were especially pronounced in
the warm-white, low-illuminance group.
Turning to goal-related processes, an interaction between color and
iUuminance was obtained for subjects' ratings of their ability to perform a
clerical coding task. This interaction reflected the fact that, in the warm-
white condition, subjects in the low-illuminance group reported significantly
higher ability than those in the high-illuminance group.
Both sets of findings (those with respect to the MODE and those for
measures related to goal setting) are generally consistent with the
hypothesis that lighting conditions expected to induce positive affect among
subjects (specifically, the warm-white, low-illuminance condition) would
result in patterns of behavior similar to those obtained in previous studies
focused on the impact of positive affect (cf. Baron, 1984; Baron et al.,
1990). In these previous studies, subjects induced to experience positive
affect (through receipt of a small gift, flattery, etc.) reported stronger
preferences for collaboration and compromise, and higher confidence in
their own abilities, than subjects not induced to experience positive affect.
The results of this investigation also provide some support for the
suggestion that the lighting conditions employed would exert significant ef-
fects on participants' subjective reactions to the experimental room and on
their reported affective states. In the warm-white condition, subjects
reported feelings significantly more calm and more alert in the low-il-
luminance than in the high-illuminance condition. Assuming that these
reactions are more indicative of positive affect than their polar opposites
on these scales (tense, sleepy), this finding provides some indication that
lighting conditions influenced positive affect. Other items on the Current
Feelings Survey, however, did not yield significant findings. Thus, as in
Study 1, color and illuminance significantly influenced several aspects of
participants' behavior but did not exert strong, consistent effects upon their
self-reported affect. Self-reports of affective states are often of questionable
usefulness or validity (cf. Isen & Baron, 1991). However, Study 3 was
designed, in part, to address this issue further.
Effects of Indoor Lighting 21

STUDY 3

Self-report measures such as those employed in Studies 1 and 2 are


widely used in research on the impact of affective states. However, they
are only one of several means for investigating the potential mediating role
of affect with respect to behavior and cognition. Isen (1987, 1990) has
repeatedly noted that another, and perhaps preferable approach, involves
a strategy known as convergingoperations (Garner, Hake, & Eriksen, 1956).
In this approach, different groups of subjects in an experiment are exposed
to treatments that differ in external form, but which are assumed to in-
fluence the intervening variable of interest (in this case, positive affect).
To the extent that these contrasting but conceptually related treatments
produce similar effects on dependent measures known to be sensitive to
the intervening variable, it can be asserted that they are indeed exerting
their impact through this variable.
In the present context, the use of converging operations would involve
exposing subjects to various lighting conditions as well as to other, unre-
lated variables -- ones found, in previous research, to influence positive af-
fect. To the extent lighting conditions exert their impact on behavior and
cognition through positive affect, they would be expected to influence
relevant dependent measures in the same manner as these other inde-
pendent variables.
The present study adopted such an approach. In this experiment, par-
ticipants were exposed or not exposed to a procedure found to generate
positive affect in many previous studies: receipt of a small, unexpected gift
(cf. Isen, 1987, 1990). In addition, subjects performed various tasks pre-
viously shown to be sensitive to positive affect under one of four different
lighting conditions. It was reasoned that if lighting conditions do influence
behavior and cognition through their impact on positive affect, then they
would influence affect-sensitive dependent measures in the same manner
as the receipt of a small, unexpected gift.
Two tasks were employed to assess these predictions: simulated per-
formance appraisals (one of the tasks used in Study 1), and donating extra
time to the researcher without pay or credit (i.e., serving as an unpaid
volunteer). Many previous studies have found that persons experiencing
positive affect are more willing to offer such assistance to others than per-
sons not experiencing positive affect (Cunningham, Shaffer, Barbee, Wolff,
& Kelley, 1990; Isen, 1970; Isen & Levin, 1972). It was tentatively predicted,
therefore, that both receipt of a small gift and exposure to certain lighting
conditions (low illuminance, warm-white light) would increase subjects' will-
ingness to volunteer and would enhance the ratings they assigned to an
imaginary employee.
22 Baron, Rea, and Daniels

Method

Subjects and Design

Subjects were 64 male and 16 female undergraduates enrolled in sec-


tions of introductory psychology at Rensselaer Polytechnic Institute. They
participated in the study voluntarily to complete a course requirement. A
2 x 2 x 2 factorial design based on two levels of illuminance (150 lux, 1500
lux), two lamp types (warm white, cool white), and the presence or absence
of a small, unexpected gift was employed. Subjects were randomly assigned
to each of these eight conditions. As in Studies 1 and 2, various lighting
conditions were in effect throughout the session. Apparatus was identical
to that employed in Studies 1 and 2.

Procedure

Gift. Delivery of a small, unexpected gift was accomplished through


the same procedures as those used in many previous studies (cf. Isen &
Geval, 1987; Isen et al., 1985). After bringing subjects to the experimental
room, the experimenter commented: "I know that you are receiving credit
for being a subject, but to show you our appreciation, we've also arranged
for each subject to receive a small gift. Please accept it with our thanks."
The experimenter then handed subjects a small plastic bag containing
several pieces of candy. (All bags contained an identical assortment.) In
the no-gift condition, no gift was mentioned and none was given to par-
ticipants.
Tasks Performed by Subjects'. After receiving or not receiving a gift,
subjects completed an expanded version of the Current Feeling Scale. After
that, they performed the performance appraisal task employed in Study 1.
On this task, they read the personnel folder of an imaginary employee,
and then rated this person on a number of different dimensions.
Request for Help. After completing the performance appraisal task,
the experimenter requested extra help from all participants. The ex-
perimenter commented as follows: "The main part of the study is now com-
plete. Actually, though, I have some Other tasks I'd like you to do. t can't
increase the amount of credit I give you -- you'll still only get i hour. But
I'd really appreciate your help. Are you willing to stay some extra time?"
If subjects agreed, the experimenter continued: "About how long are you
willing to stay?" Following these procedures, participants were debriefed
and released.
Effects of Indoor Lighting 23

Table VIII. Mean Reported Affect and Standard Deviations on Two Dimensions as a
Function of Illuminance, Color, and Receipt of a Small Gift (Study 3)a
No Gift Gift
150 lux 1500 tux 150 lux 1500 lux
Nervous-calm
Warm white 5.78a 4.83bc 5.00ac 5.33ac
(1.27) (1.30) (1.37) (1.56)
Cool white 4.33a 5.00a 5.50a 5.00a
(0.78) (1.78) (1.54) (1.09)
Tense-relaxed
Warm white 5.8% 4.58bc 4.700c 4.75bc
(1.78) (1.05) (1.45) (1.80)
Cool white 4.t7a 4.91a 5.08a 4.58a
(1.06) (1.42) (t.78) (.3t)
~Vote: Within each color condition (for each dependent measure), means that do not share
a subscript differ significantly (p < .05) by the Duncan Multiple-Range Test. Higher numbers
indicate greater tendencies to report feeling calm and relaxed. (Range = t-7.) Numbers in
parentheses are cell standard deviations.

Results

Current Feelings Survey

Previous research (e.g., Baron, 1990) indicates that items on the Cur-
rent Feelings Survey may measure several different factors. To take account
of such findings, data for this survey were subjected to a principal-com-
ponents factor analysis. This analysis yielded three distinct factors. Factor
1, which accounted for 49.7% of the variance, loaded on four scales per-
taining to positive or negative feelings (bad-good, sad-happy, negative-
positive, unpleasant-pleasant). Factor 2, which accounted for 14.1% of the
variance, loaded on four scales pertaining to activation (tired-energetic,
slow-quick, drowsy-alert, bored-interested). Factor 3, which accounted for
12.4% of the variance, loaded on two scales related to anxiety or tension
(nervous-calm, tense-relaxed).
Separate multivariate analyses of variance were then conducted on
the data for the items relating to these three factors. For factors 1 and 2,
only the effect of the gift variable was significant, F(4, 79) = 2.01, p < .05,
F(4, 79) = 3.23, p < .01. Followup univariate analyses indicated that in-
dividuals who received a gift reported more positive feelings (factor 1) and
greater activation or alertness (factor 2) than those who did not receive a
gift.
24 Baron, Rea, and Daniels

Table IX. Performance Ratings and Standard Deviations as a Func-


tion of Receipt of a Small Gift (Study 3)a
No gift Gift
Deserves merit raise 2.91a 3.41b
(0.91) (1.24)
Deserves promotion 2.58a 3.9%
(1.23) (0.97)
Qualified for current job 4.17a 4.55a
(1.19) (0.75)
Job-related skills 4.75a 4.42a
(0.75) (0.67)
Intelligence 4.33a 4.50a
(0.89) (0.67)
Careful in work 3.254 3.44a
(0.75) (0.88)
Motivation 3.25a 3.330
(1.06) (0.78)
Work attitudes 3.58a 3.75a
(0.79) (0.62)
Current performance 3.83a 4.08a
(1.11) (0.79)
Gets along with others 3.6% 3.83a
(0.99) (0.84)
Overall rating 4.10a 4.25a
(0.82) (0.97)
aNote: For each dimension, means that do not share a subscript differ
significantly (p < .05) by the Duncan Multiple-Range Test. Higher
numbers indicate more favorable ratings. (Range = 1-7,) Numbers
in parentheses are cell standard deviations.

O n f a c t o r 3, the interaction b e t w e e n illuminance, color, a n d gift


was significant, F(2, 81) = 3.65, p < .05. This i n t e r a c t i o n s t e m m e d f r o m
the fact that a m o n g subjects w h o did n o t receive a gift, those e x p o s e d
to w a r m - w h i t e light r e p o r t e d f e e l i n g s i g n i f i c a n t l y c a l m e r a n d m o r e
relaxed in the low-illuminance t h a n in the high-illuminance condition.
I n contrast, subjects e x p o s e d to cool-white light r e p o r t e d feeling slightly
(but n o t significantly) c a l m e r and m o r e relaxed in the high- t h a n in the
low-illuminance c o n d i t i o n (see T a b l e V I I I ) . Similar findings w e r e n o t
o b s e r v e d a m o n g subjects w h o received a gift, however.
Effects of Indoor Lighting 25

Table X. Mean Number of Minutes and Standard Deviations Volunteered as a Function


of Illuminance, Color, and Receipt of a Gift (Study 3)a
No gift Gift
Color 150 lux 1500 lux 150 lux 1500 lux
Warm white 22.22bd 15.91bc 19.38cd 20.56cd
(6.18) (8.96) (10.51) (8.46)
Cool white 13.57ac 7.14a 15.50Ocd 20.50cd
(8.52) (4.88) (8.96) (10.12)
aNote: Means that do not share a subscript differ significantly(p < .05) by the Duncan Mul-
tiple-Range Test. Numbers in parentheses are cell standard deviations.

Simulated Performance Appraisal

A multivariate analysis of variance performed on the data for the


scales on which subjects evaluated the imaginary employee yielded only a
significant main effect for illuminance, F(11, 71) = 2.04, p < .04. This find-
ing, which replicated results of Study 1, indicated that, overall, subjects as-
signed higher ratings to the imaginary employee in the low-illuminance (M
=- 4.86) than in the high-illuminance condition (M = 3.71). While the main
effect for the gift variable was not significant in the overall multivariate
analysis of variance, followup univariate analyses indicated that the effect
of this variable was significant for two of the dependent measures: whether
the imaginary employee deserved a merit raise and a promotion, F(1, 79)
= 4.60, 3.72, p < .05 in both cases. As anticipated, subjects who had
received a small gift rated the employee higher on both dimensions than
those who had not received a gift (see Table IX). In sum, there was some
evidence that lighting variables and receipt of a small gift influenced per-
formance on this task in a similar manner.

Time Volunteered

Virtually all participants agreed to the experimenter's request for


help. Thus, there were no differences in proportion of subjects volunteering
in the various experimental conditions. However, an analysis of variance on
the amount of time participants volunteered revealed a main effect for
color, F(1, 70) = 5.96, p < .02, and a main effect for gift, F(1, 70) = 3.91,
p < .05. In addition, the interaction between gift and illuminance, F(1, 70)
= 4.68, p < .05, was also significant.
The main effect of color stemmed from the fact that subjects exposed
to warm-white light volunteered more time (M = 19.51 min) than those
26 Baron, Rea, and Daniels

exposed to cool-white light (M = 14.17 min). Similarly, the main effect of


the gift variable stemmed from the fact that participants who received a
gift (M = 18.98) volunteered more time than those who did not (M =
14.71).
The interaction between illuminance and receipt of a gift reflected
the fact that, in the absence of a gift, subjects in the 150-1ux condition
donated more time (M = 17.90) than those in the 1500-1ux condition (M
= 11.53). In the presence of a gift, however, these differences were not
apparent (M = 17.44, 20.53, for the 150- and 1500-1ux conditions, respec-
tively; see Table X). Thus, it appeared that receipt of a small, unexpected
gift eliminated any impact of illuminance on subjects' willingness to help.

Discussion

The results of this study provide further evidence for the conclusion
that indoor lighting can influence important forms of work-related be-
havior. As in Study 1, subjects assigned higher ratings to an imaginary
employee under conditions of low compared with relatively high il-
luminance. In addition, the amount of time they volunteered as uncom-
pensated participants was influenced by color, but only when there was no
gift.
Perhaps the most important finding of Study 3 involves the fact that
for two dependent measures - performance appraisal and helping -- ex-
posure to specific lighting conditions and receipt of a small, unexpected
gift produced parallel effects. Low iltuminance and receipt of a gift en-
hanced ratings of a fictitious employee. Similarly, both receipt of a gift and
exposure to warm-white light increased subjects' willingness to serve as un-
paid volunteers. Finally, in the absence of a gift, subjects volunteered more
of their time under low than high levels of illuminance. Thus, the results
of study 3 provide some evidence, through converging operations, for the
conclusion that lighting conditions influence at least some aspects of be-
havior through their impact on positive affect. To summarize the pertinent
reasoning, receipt of a gift has been found, in previous studies, to induce
positive affect and to influence several aspects of cognition and behavior
through this mechanism (Isen, 1987). In the present study, certain lighting
conditions were found to influence the same aspects of cognition and be-
havior in a corresponding manner. Together, these facts suggest that light-
ing conditions did indeed influence performance through the intervening
variable of positive affect.
It should be noted that results for the Personal Feelings Survey, too,
provided some evidence for the view that effects of lighting stem, at least
Effects of Indoor Lighting 27

in part, from shifts in positive affect. Subjects who received a gift reported
higher levels of positive affect on factors 1 and 2 than subjects who did
not. Since small, unexpected gifts have been found to generate positive
affect in many previous studies (cf. Isen, 1990), this finding suggests that
the Current Feelings survey is sensitive to shifts in affect. In this context,
the three-way interaction between illuminance, color, and gift for factor 3
of this scale - a factor relating to feelings of anxiety or tension - is reveal-
ing. This interaction indicated that, among subjects who did not receive a
gift, those exposed to warm-white light reported experiencing more positive
affect (i.e., feeling calmer and more relaxed) under low iUuminance than
under high illuminance. Among subjects exposed to cool-white light, how-
ever, the opposite was true: Subjects reported less tension under high than
under low illuminance. While these findings are complex, taken together
with the main effect for receipt of a gift, they do suggest that the impact
of lighting conditions may stem, at least in part, from lighting-induced shifts
in positive affect.
The finding that illuminance influenced helping behavior only in the
absence of a gift may also be interpreted as evidence for the view that this
aspect of indoor lighting influences affective states. Under these conditions,
modest increments in positive affect induced by low levels of illuminance
could, apparently, increase the amount of time donated by subjects. In the
presence of a gift, in contrast, relatively high levels of positive affect were
already present. Thus, further increments in positive affect could not readily
be induced by low levels of illuminance. In other words, the presence of
a gift may have generated ceiling effects with respect to lighting-induced
positive affect. This possibility can be examined in future studies by reduc-
ing the size or attractiveness of the gift employed, or through the use of
other treatments known to generate somewhat lower levels of positive af-
fect.

GENERAL CONCLUSIONS

Together, the three studies conducted offer support for the hypothesis
that lighting conditions can influence performance on several different
tasks. Across the three investigations, variations in two aspects of indoor light-
ing -- illuminance and lamp color -- were found to exert significant effects on
all of the following: (1) evaluations of a fictitious employee, (2) word
categorization, (3) preference for resolving interpersonal conflicts through
collaboration vs. avoidance, (4) reported expected ability to perform a cleri-
cal task, and (5) willingness to offer help to others. In addition, participants
reported contrasting subjective reactions to the experimental rooms and to
28 Baron, Rea, and Daniels

the lighting conditions in them under varying combinations of illuminance


and color. The tasks listed above are quite diverse in nature, and involve
interpersonal processes (Thomas, 1992) as well as several aspects of cog-
nition and information processing (Fiedler & Forgas, 1988; Isen, 1987).
Thus, it appears that indoor lighting can influence performance on a wide
range of tasks aside from ones that primarily involve the processing of
visual information (cf. Boyce, 1981).
It should be noted, however, that variations in spectral distribution
and illuminance did not significantly influence several other dependent
m e a s u r e s - for example, responses to the PANAS and reactions to a
hypothetical conflict situation in which provocative actions by another per-
son appeared to stem from contrasting causes. These negative findings sug-
gest the need for further research designed to identify those aspects of
behavior and cognition that are, and are not, influenced by these aspects
of indoor lighting.
It is also important to note that while spectral distribution and il-
luminance did indeed influence several aspects of participants' behavior,
the mechanisms underlying such effects remain somewhat uncertain. Shifts
in positive affect have been suggested as one possibility, and evidence per-
taining to this suggestion will be considered below. Here, we wish to em-
phasize that several alternatives exist and should ibe given careful attention
in future studies. One of these involves the potential role of arousal (cf.
Veitch et al., 1991). In a recent study, Biner (1991) found evidence indicat-
ing that levels of itluminance somewhat higher than those used here (ap-
proximately 1750 lux vs. 1500 lux in the present study) can induce
increments in arousal relative to more moderate levels of illuminance (ap-
proximately 350 lux), and that levels of illuminance lower than those used
here (approximately 32 lux vs. 150 lux) can produce reductions in arousal.
These findings suggest that some portion of the effects of indoor lighting
may stem from shifts in arousal as well as from shifts in affective states.
This possibility - and especially the possibility that warm-white, low-il-
luminance conditions produce reductions in arousal - should be carefully
examined in future studies. However, the fact that receipt of a small gift
and such lighting conditions enhanced both helping and performance ap-
praisals argues against an interpretation based on reduced arousal. This is
the case because receipt of a small gift has been found in both the present
and previous research (e.g., Isen & Daubman, 1984) to generate increased
rather than reduced arousal. Yet, in the present study, a gift produced
similar effects on several dependent measures as did low-illuminance and
warm-white light. Further research is clearly needed, therefore, to inves-
tigate the potential role of arousal as a factor in the impact of indoor light-
ing.
Effects of Indoor Lighting 29

Returning to the suggestion that the effects of lighting observed in


the present study stemmed at least in part from shifts in affective states,
several findings should be emphasized. First, subjects reported more posi-
tive reactions to the experimental rooms or the lighting conditions in them
when exposed to warm-white than to cool-white light and to low rather
than to high illuminance (Studies 1 and 2). Second, as already noted, sub-
jects reported feeling significantly more calm and less tense in the presence
of warm-white, low-illuminance light than in the presence of warm-white,
high-illuminance light (Study 3). Such reports can be interpreted as reflect-
ing shifts in affect rather than simple decrements in arousal. Third, and
perhaps most important, subjects' behavior on several dependent measures
was influenced in the same manner by receipt of a small, unexpected gift
(a treatment known to induce positive affect) and by lighting conditions to
which they reported positive reactions. While none of these findings, by
itself, provides definitive evidence for the mediating role of positive affect
with respect to the behavioral impact of illuminance or lamp color, taken
together they do offer some support for the view that positive affect plays
a role in this process. This conclusion agrees with the findings of a large
number of recent studies indicating that positive affect can influence many
aspects of work-related behavior (George, 1989, 1990; Isen & Baron, 1991).
There is no intention of suggesting here, however, that positive affect
is the only mechanism through which lighting conditions might influence
behavior. On the contrary, as noted earlier, it seems possible that other
processes, such as arousal and the priming of specific memories or memory
categories, may also be involved (Baddeley, 1990). Additional research is
needed to determine the extent to which these, and perhaps other,
mechanisms play a role in the behavioral effects of indoor lighting.
It should also be noted that quite apart from illuminance level and
color, relative familiarity with various lighting conditions may play a rote
in determining affective reactions to them. A large body of evidence indi-
cates that, at least up to a point, increasing exposure to (and therefore
familiarity with) a wide range of stimuli enhances affective reactions to
these stimuli (e.g., Bornstein, 1989). In this context, it would be expected
that lighting conditions to which indMduals have been most frequently ex-
posed -- for example, coot-white light at relatively high-illuminance levels
(as in offices, public buildings, etc.), or warm-white light at relatively low-
illuminance levels (as in homes, restaurants, etc.) - might generate higher
levels of positive affect than less familiar lighting conditions. Further re-
search is necessary to evaluate this possibility, and to determine the relative
contribution of familiarity and various aspects of lighting conditions them-
selves to affective reactions to indoor illumination.
30 Baron, Rea, and Daniels

At this point, it should be noted that the present findings were ob-
tained under restricted laboratory conditions in experiments of relatively
short duration. Thus, the extent to which they are generalizable to actual
work settings remains to be established. It should be noted, however, that
the effects of several other environmental variables (e.g., noise, uncomfor-
tably high or low temperatures, crowding) have been found to increase
rather than decrease with extended periods of exposure (Anderson, 1989;
Fisher et al., 1990). Further, these have often been found to exert similar
effects in both field and laboratory settings (Bell & Fusco, 1989). Thus,
while generatizability of the present findings can only be established
through further research, there are some grounds for suggesting that the
effects observed here may also occur in many work settings. In any case,
given that large numbers of employees are exposed to indoor lighting for
several hours each day, it seems reasonable to suggest that the potential
effects of this aspect of work environments merit further, systematic inves-
tigation.

REFERENCES

Anderson, C. A. (1989). Temperature and aggression: Ubiquitous effects of heat on occur-


rence of human violence. Psychological Bulletin, 106, 74-96.
Baddeley, A. (1990). Human memory. Boston: Allyn & Bacon.
Baron, R. A. (1978). Aggression and heat: The "long hot summer" revisited. In A. Baum, S.,
Valins, & J. E. Singer (Eds.), Advances in environmental research (VoL 1). Hillsdale, NJ:
Erlbaum.
Baron, R. A. (1983). The control of human aggression: A strategy based on incompatible
responses. In R. G. Geen & E. I. Donnerstein (Eds.), Aggression: Theoreticaland empirical
reviews. New York: Academic Press.
Baron, R. A. (1984). Reducing organizational conflict: An incompatible response approach.
Journal of Applied Psychology, 69, 272-279.
Baron, R. A. (1987). Mood of interviewer and the evaluation of job candidates. Journal of
Applied Social Psychology, 17, 911-926.
Baron, R. A. (1988). Negative effects of destructive criticism: Impact on conflict, self-efficacy,
and task performance. Journal of Applied Psychology, 73, 199-207.
Baron, R. A. (1990). Conflict in organizations. In K. R. Murphy & F. E. Saal (Eds.). Psychology
in organizations: Integrating science and practice. Hillsdale, NJ: Erlbaum.
Baron, R. A., Fortin, S. P., Frei, R. L., Hauver, L. A., & Shack, M. L. (1990). Reducing
organizational conflict: The potential role of socially-induced positive affect. International
Journal of Conflict Management, 1.
Belcher, M. C., & Kluczny, R. (1987). The effects of light on decision making: Some ex-
perimental results (pp. 354-357). Proceedings of CIE. Venice.
Bell, P. A., & Fusco, M. E. (1989). Heat and violence in the Dallas field data: Linearity,
curvilineary, and heteroscedasticity. Journal of Applied Social Psychology, 19, 1479-1482.
Biner, P. M. (1991). Effects of lighting-induced arousal on the magnitude of goal valence.
Personality and Social Psychology Bulletin, 17, 219-226.
Biner, P. M., Butler, D. L., Riscsher, A. R., & Westergren, A. J. (1989). An arousal optimiza-
tion model of lighting level preferences: An interaction of social situation and task
demands. Environment and Behavior, 21, 3-16.
Effects of Indoor Lighting 31

Blackwell, H. R. (t959). Development and use of a quantitative method for specification of


interior illuminating levels on the basis of performance data. Illuminat#zg Engineerin~ 54,
317-353.
Blehar, M. C, & Rosenthal, N. E. (Eds.). (1989). Season affeetive disorders and phototherapy.
New York: Guilford Press.
Boyce, P. R. (1981). Human factors in lighting. New York: MacmilIan.
Boyce, P. R , Berman, S. M., Collins, B. L., Lewis, A. L., & Rea, M. S. (1989). Lighting and
human performance: A review. Lighting Equipment Division, National Electrical Manufac-
ture~ Association and Lighting Research Institute.
Buffer, D. L, & Biner, P. M. (1987). Preferred lighting levels: Variability among settings,
behaviors, and individuals. Environment and Behavior; 19, 695-721.
Cardy, R. L., & Dobbins, G. H. (1986). Affect and appraisal accuracy: Liking as an integral
dimension in evaluating performance. Journal of Applied Psychology, 71, 672-678.
Carnevale, P. J. D., & Isen, A. M. (1986). The influence of positive affect and visual access
on the discovery of integrative solutions in bilateral negotiation. Organizational Behavior
and Human Decision Processes, 37, 1-13.
Cohen, S., Evans, G. W., Krantz, D. S., & Stokols, D. (1980). Physiological, motivational, and
cognitive effects of aircraft noise on children. American Psychologist, 35, 231-243.
Cohen, S., Evans, G. W., Stokols, D., & Krantz, D. (1986). Behavior, health, and environmental
stress. New York: Plenum Press.
Cunningham, M. R., Sharer, D. R., Barbee, A. P., Wolff, P. L. & Kelley, D. J. (1990). Separate
processes in the relation of elation and depression to helping: Social versus personal
concerns. Journal of Experimental Social Psychology, 26, 13-33.
Davis, R. G., & Ginthner, D. (1990). Correlated color temperature, illuminance level, and
Kruithof curve. Journal of the Illuminating Engineering Society, 19, 27-38.
Fiedler, K., & Forgas, J. P. (Eds.). (1988). Affect, cognition, and social behavior. Toronto:
Hogrefe.
Fisher, J. D., Bell, P. A., & Baum, A. (1990). Environmentalpsychotogy (3rd ed.). New York:
Holt, Rinehart, & Winston.
Flynn, J. E. (1977). A study of subjective responses to low energy and nonuniform lighting
systems. Lightil N Design and Application, 7, 6-15.
Flynn, J. E., Spencer, T. J., Martyniuk, O., & Hendrick, C. (1973). Interim study of procedures
for investigating the effect of light on impression and behavior. Journal of the Illuminating
Engineering Society, 3, 87-94.
Garner, W. R., Hake, H. W., & Eriksen, C. W. (1956). Operations and the concept of per-
ception. Psychological Review, 63, 149-159.
George, J. M. (1989). Mood and absence. Journal of Applied Psychology, 74, 317-324.
George, M. M. (1990). Personality, affect, and behavior in groups. Journal of Applied Psychol-
ogy, 75, 107-I16.
Gifford, R. (1988). Light, decor, arousal, comfort and communication. Journal of Environmen-
tal Psychology, 8, 177-189.
Illuminating Engineering Society of North America (1987). Illuminating Engineering Sociefy
Handbook: Application Volume. New York: Kaufman, J. E. and Chfistensen, J. F. (eds.).
Isen, A. M. (1970). Success, failure, and reaction to others: The warm glow of success. Journal
of Personality and Social Psychology, 15, 294-301.
Isen, A. M. (1984). Toward understanding the role of affect in cognition. In R. Wyer & T.
SruI1 (Eds.), Handbook of social cognition. Hillsdale, NJ: Erlbaum.
Isen, A. M. (1987). Positive affect, cognitive processes, and social behavior. In L. Berkowitz (Ed.),
Advances in experimental socialpsychology (Vol. 20, pp. 203-253). New York: Academic Press.
Isen, A. M. (1990). The influence of positive and negative affect on cognitive organization:
Some implications for development. In N. Stein, B. Leventhal, & T. Trabasso (Eds.).
Psychological and biological approaches to emotion (pp. 75-94). Hillsdale, NJ: Erlbaum.
Isen, A. M., & Baron, R. A. (1991). Affect as a factor in organizational behavior. In B. M.
Staw & L. L Cummings (Eds.), Research in organizational behavqor (Vol. 14, pp. 1-53).
Greenwich., CT: JAI Press.
32 Baron, Rea, and Daniels

Isen, A. M., & Daubman, K. A. (1984). The influence of affect on categorization. Journal of
Personality and Social Psychology, 47, 1206-1217.
Isen, A. M, Daubman, K. A., & Nowicki, G. P. (1987). Positive affect facilitates creative
problem solving: When we are glad, we feel as if the light has increased. Journal of Per-
sonality and Social Psychology, 51, 1122-1131.
Isen, A. M., & Geva, N. (1987). The influence of positive affect on acceptable level of risk:
The person with a large canoe has a large worry. Organizational Behavior and Human
Decision Processes, 39, 145-154.
Isen, A. M., Johnson, M. M. S., Mertz, E., & Robinson, G. (1985). The influence of positive
affect on the uniqueness of word association. Journal of Personality and Social Psychology,
48, 1413-1426.
Isen, A. M., & Levin, P. F. (1972). Effects of feeling good on helping: Cookies and kindness.
Journal of Personality and Social Psychology, 21, 354-358.
Isen, A. M., & Means, B. (1983). The influence of positive affect on decision-making strategy.
Social Cognition, 2, 18-31.
Isen, A. M., Nygren, T. E , & Ashby, F. G. (1988). Influence of positive affect on the subjective
utility of gains and losses: It is just not worth the risk. Journal of Personality and Social
Psychology, 55, 710-717.
Isen, A. M., & Shalker, T. E. (1982). Do you "accentuate the positive, eliminate the negative"
when you are in a good mood? Social Psychology Quarterly, 41, 345-349.
Kasper, S., Rogers, S. L. B., Yancey, A., Skwerer, R. G., Schultz, P. M., & Rosenthal, N. E.
(1989). Psychological effects of light therapy in normals. In M. C. Blehar & N. E. Rosen-
thai (Eds.), Seasonal affective disorders and phototherapy (pp. 260-269). New York: Guil-
ford Press.
Kaye, S. S. (1988). Variations in the luminous and sonic environment: Proofreading and visual
search effects: Mood states and frustration tolerance aftereffects (Project 84:SP:2). Winnipeg:
Lighting Research Institute, University of Manitoba.
Liden, R. C., & Mitchell, T. R. (1988). Ingratiatory behaviors in organizational settings.
Academy of Management Review, 13, 572-587.
Locke, E. A., & Latham, G. P. (1990).A theory of goal setting and task performance. Englewood
Cliffs, NJ: Prentice-Hall.
Nagar, D., & Pandey, J. (1987). Affect and performance on cognitive task as a function of
crowding and noise. Journal of applied Social Psychology, 17, 147-157.
Nelson, T. M., Nilsson, T. H., & Johnson, M. (1984). Interaction of temperature, illuminance
and apparent time on sedentary work fatigue. Ergonomics, 27, 89-101.
Nisbett, R. E.., & Wilson, T. D. (1977). Telling more than we can know: Verbal reports on
mental processes. Psychological Bulletin, 84, 231-259.
Oldham, G. R., & Fried, Y. (1987). Employee reactions to workplace characteristics. Journal
of Applied Psychology, 72, 75-80.
Organ, D. W. (1988). Organizational citizensh!p behavior: The good soldier syndrome. Lexington,
MA: Lexington Books.
Pruitt, D., & Rubin, J, Z. (1986). Social conflict. New York: Random House.
Rea, M, S. (1986). Toward a model of visual performance: Foundations and data. Journal of
the IES, 15 (2), 41-57.
Rea, M. S. (1988). Proposed revision of the IESNA illuminance selection procedure. Journal
of the Illumination Engineerb~gSociety, 17 (1), 20-28.
Riskind, J. H. (1983), Nonverbal expressions and the accessibility of life experience memories:
A congruence hypothesis. Social Cognition, 2, 61-86.
Roethlisberger, F. J., & Dickson, W. J. (1941). Management and the worker:An account of a
research program conducted by the Western electric Company, Hawthorne Works, Chicago.
Cambridge, MA: Harvard University Press,
Rosch, E, (1975). Cognitive representations of semantic categories. Journal of Experimental
Psychology: General, 104, 192-233.
Rosenthal, N. E. (1985). Antidepressant effects of light in seasonal affective disorder.
American Journal of Psychiatry, 142, 163-170.
Effects of Indoor Lighting 33

Sundstrom, E., & Sundstrom, M. G. (1986). Workptaces: The psychology of the physical en-
vironment in offices and factories. London: Cambridge University Press.
Thomas, K. W. (1992). Conflict and negotiation processes. In M. Dunnette (Ed.), Handbook
of industrial and organizational psychology (2nd ed.). Palo Alto, CA: Consulting
Psychologists Press.
Veitch, J. A., Gifford, R., & Hine, D. W. (t991). Demand characteristics and full spectrum
lighting effects on performance and mood. Journal of Environmental Psychology, 11, 87-95.
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures
of positive and negative affect: The PANAS scales. Journal of Personality and Social
Psychology, 54, 1063-1070.
Weston, H. C. (1935). The relation between illumination and industrial efficiency, 1. The
effect of size of work. HMSO, pp. 1-14.
Weston, H. C. (1945). The relation between illumination and visual efficiency - The effect of
brightness contrast (Report No. 87, pp. 1-35). Industrial Health Research board, Great
Britain Medical Research Council.

Potrebbero piacerti anche