You are on page 1of 6

Journal of Business Research 54 (2001) 145 – 150

Shedding some light on store atmospherics


Influence of illumination on consumer behavior
Teresa A. Summersa,*, Paulette R. Hebertb
a
Textiles, Apparel Design, Merchandising, School of Human Ecology, Louisiana State University, Agricultural Center, Baton Rouge, LA 70803, USA
b
School of Architecture, University of Louisiana at Lafayette, Lafayette, LA 70504, USA
Received 1 May 1999; accepted 1 May 1999

Abstract

The influence of display lighting, a component of store atmospherics, on consumer approach – avoidance behavior was studied.
Supplemental lighting was temporarily installed and manipulated on merchandise displays in two retail stores to test for effects on consumer
behaviors of Time at Display, Number of Items Touched and Number of Items Picked Up. Video cameras recorded the consumers’ actions.
Subjects consisted of an accidental sample of consumers in the stores who passed within the measurement zones of the cameras. A total of
2367 subjects were recorded; behaviors and demographics were coded by two raters. Data were analyzed using correlations, Analysis of
Variance, and Bonferroni paired comparisons. Supplemental lighting treatments had a positive effect on consumer behavior, as qualified by
display. Interactions between lighting and display were found to be statistically significant. Findings could be useful to retailers in
developing in-store lighting as part of a store’s atmospherics to aid in attracting and retaining consumer patronage. D 2001 Elsevier Science
Inc. All rights reserved.

Keywords: Lighting; Approach – avoidance; Store atmospherics; Consumer behavior

1. Introduction 2. Purpose

Retailers must continually strive to meet increased The purpose of our exploratory study was to empirically
competition and changing consumer preferences. Accord- measure the effect of light level manipulations in retail
ing to Kotler (1973), consumers respond to the ‘‘total display areas on the approach – avoidance behavior of
product.’’ A significant aspect of the total product is the consumers using a new methodology we developed that
place where it is bought or consumed. The atmosphere incorporated video observational techniques. We postulated
of the place can be more influential than the product itself that an increase in light levels on the test display would
in the purchase decision (p. 48). Store atmospherics increase consumers’ approach behavior, relative to the
describe the special sensory qualities of retail spaces that display, as measured by: a desire to approach (remain at)
are often designed to evoke particular consumer responses. the test display or to avoid (leave) the test display; a desire
Lighting is a significant component of store atmospherics. or willingness to explore the test display (touch merchan-
A more appealing store with better-illuminated merchan- dise) or to avoid the test display; the degree of approach
dise may entice shoppers to visit the store, linger, and (enhancement) or avoidance of the task of picking up
hopefully make a purchase. Even though the importance display merchandise.
and benefits of lighting are generally acclaimed for retail
environments, very few empirical retail lighting studies
have been conducted. 3. Rationale and previous findings

3.1. Approach – avoidance theory

* Corresponding author. Tel.: +1-225-388-2281. The Mehrabian – Russell (M – R) model, first proposed in
E-mail address: tsummer@lsu.edu (T.A. Summers). 1974, served as the theoretical framework for our study

0148-2963/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved.
PII: S 0 1 4 8 - 2 9 6 3 ( 9 9 ) 0 0 0 8 2 - X
146 T.A. Summers, P.R. Hebert / Journal of Business Research 54 (2001) 145–150

(Mehrabian and Russell, 1974). Mehrabian (1976) sug- vior utilizing a convenience sample of 171 wine store
gested that the combined effects of pleasure, arousal, and consumers over a 16-night period. Merchandise lighting
dominance influenced people’s behavior in particular en- was manipulated to be ‘‘soft’’ on different evenings by
vironments. He believed that lighting was a chief factor in replacing some of the store’s existing lamps with lower
the environment’s impact on individuals because ‘‘brightly wattages or ‘‘bright’’ by replacing lamps. The perceptions
lit rooms are more arousing than dimly lit ones’’ (p. 89). of wine store employees were used to determine that an
Mehrabian described his model as ‘‘. . . the classical ‘in- adequate difference in light levels was created. A modified
verted u’ relationship between approach – avoidance and version of an instrument, developed by Hoyer (1984), was
arousal’’ (p. 22), and theorized that people would want to used to record consumer behaviors. Areni and Kim (1994)
remain in environments that were both pleasant and found that consumers examined and handled significantly
arousing. Mehrabian proposed that, in pleasant surround- more items under ‘‘bright’’ lighting conditions than under
ings, an individual’s approach behavior would increase ‘‘soft’’ lighting conditions. While the M– R model, with its
with increases in the arousal level. Mehrabian postulated: classical ‘inverted u’ relationship between approach – avoid-
‘‘. . . why do people shop for things they don’t need or ance and arousal, implies that lighting that is ‘‘too bright’’
cannot afford? Our analyses of the various types of shops could over-stimulate consumers leading to avoidance beha-
suggest that the single common element of the shopping vior, no research was found that tested this extreme notion.
experience is arousal . . . this suggests that people shop to
increase their arousal and pleasure levels’’ (p. 293). 3.3. Measurement techniques
Donovan and Rossiter (1982) were among the first to
apply the M – R model to actual measurements of approach – Arena and Kim’s (1994) study was the only research we
avoidance behavior in retail settings. They found arousal to found that applied the M– R model to the influence of in-
be a significant predictor of approach intentions for time and store lighting on consumer behavior. Building on their
affiliation. ‘‘. . . given a pleasant retail store atmosphere, efforts, we developed and tested a new methodology in-
arousal now becomes the key mediator of intentions to spend corporating supplemental lighting with video observation to
time in the store’’ (p. 50). Donovan et al. (1994) reported the facilitate determination of the influence of lighting on
pleasantness of in-store atmospherics was a significant pre- consumer behavior. Previous studies have relied heavily
dictor of desire to approach or remain in the store and to on consumer perception and recall. However, much of
spend money while arousal was found to vary in its influ- lighting’s effects and components are difficult for the
ence. These researchers did not study the effect of lighting average consumer to describe, and his/her illumination
per se, but rather the total store environment. vocabulary is usually extremely limited. Terms, such as
brightness, may have different meanings for lighting experts
3.2. Lighting field studies and consumers. Further, the purpose of retail lighting, in
addition to making the space more aesthetic and functional,
We believe that research aimed to quantify the influence can be used to elicit particular behaviors, often without
of illumination on consumer behavior in actual retail consumers’ knowledge. The Illuminating Engineering So-
environments can be beneficial since lighting is recognized ciety of North America (IES) Handbook states that ‘‘the
as an important component of store atmospherics, affecting primary goals common in the lighting of merchandise are to
the consumers’ visual appraisal of everything in a store, attract the consumer, to initiate purchases, and to facilitate
including the merchandise (Gobe, 1990; Rea, 1993; Lopez, the completion of the sale’’ (Rea, 1993). The IES recom-
1995). Yet, we found few empirical studies in our review of mends that when illuminating merchandising spaces, light-
the literature. Most of the studies we found had been ing designers should create a ‘‘pleasant and secure
performed in controlled environments (Gardner and Siom- environment to do business’’ (Rea, 1993, p. 591).
kos, 1986; Baker et al., 1992), and only three studies had
been conducted in actual store settings. Boyce et al. (1996)
and Cuttle and Brandston (1995) surveyed consumers 4. Methodology
regarding their perceptions of in-store lighting after exten-
sive renovations were made to an existing furniture store Mehrabian (1976) theorized that brighter light increased
and to an existing grocery store. The designers – researchers, arousal and that the combination of pleasantness and arousal
Cuttle and Brandston (1995) and Boyce et al. (1996) made individuals more susceptible to influence. In our
participated in the design of the actual interior renovation study, the ‘‘influence’’ to which the individuals were sub-
of the existing stores, which were used in the studies. Both jected was the store’s merchandise. We proposed that an
studies measured the impact of the lighting on changes in increase in merchandise lighting helped to create a pleasant
electricity consumed by each store, customer, and proprietor and arousing environment at the merchandise display and
or staff’s perceptions, and merchandise sales. Areni and encouraged consumers to approach. We developed our
Kim (1994) applied the M – R model to their study of the methodology based on the number of light levels tested,
impact of in-store lighting manipulation on shopping beha- coding instrument, and observational measurements used in
T.A. Summers, P.R. Hebert / Journal of Business Research 54 (2001) 145–150 147

the Areni and Kim (1994) study. We conducted several tests developed by Hoyer (1984) and modified by Areni and Kim
on methods and a pre-test to verify the feasibility of the (1994). Inter-rater reliability tests produced highly signifi-
methodology prior to actual data collection. cant correlations between the raters ranging from 0.78 to 1.0
for all observations of demographics and dependent vari-
4.1. Research design ables. The sample included a total of 2367 consumers
recorded within the zone of measurement, which was
We selected two independently-owned stores, a hardware approximately 5 ft wide (across the display) and up to 6 ft
store, and a western apparel/feed store, located in the deep in front of the display in each store. The variable, Time
Metropolitan Statistical Area of a major southern city for at Display (Time), was determined by measuring the time (in
the study. Participating retailers allowed us to install tem- seconds) spent by consumers standing or walking in front of
porary supplemental lighting systems over one of their the display with their head and/or body turned in the general
merchandise displays; to administer On/Off treatments of direction of the test display. The other variables were
the light system alternating daily during their normal oper- measured by counting the Number of Items Touched
ating hours; and to observe and record their consumers via (# Items Touched) by consumers on the display and the
videotape. Both stores also exhibited substantial consumer Number of Items Picked Up (# Items Picked Up) by con-
activity and had a minimum of daylight intrusion, which sumers from the display. A consumer identification code was
was necessary in order to control for variations in overall assigned sequentially to each new consumer observed on the
store ambient lighting levels. videotape. If a consumer returned to the zone of measure-
The hardware store test display featured tools, and the ment within 3 min after departing the zone, his/her additional
western apparel/feed store test display comprised belts. behaviors were added to his/her initial behaviors on the same
The merchandise on the test displays in both stores Code Sheet. If more than 3 min had passed, a new Code
appeared to be non-color-critical, and therefore, unaffected Sheet was started for the new behaviors. Raters determined
by color shift due to light. Existing ambient lighting in the the Time consumers remained by watching a hand-held
hardware store was an 8-ft-long double lamp profile digital timer and recording the total number of seconds.
exposed bare lamp strips that was surface-mounted to the The # Items Touched and the # Items Picked Up by
ceiling and lamped with 75-W General Electric Cool consumers at the test display were counted and recorded
White fluorescent lamps. Existing ambient lighting in the on the Code Sheet. The sex and approximate age range of
western apparel/feed store was an 8-ft-long double lamp consumers were also noted.
profile exposed bare lamp strips surface-mounted to the
ceiling and lamped with 60-W Sylvania Super Saver 4.3. Experimental design and statistical treatment
Designer fluorescent lamps.
We installed temporary, supplemental illumination in the Correlations among the dependent variables were first
ceilings of both stores over the test displays consisting of calculated to determine the relationships. A 2 (stores)  2
inexpensive, readily obtainable, utilitarian floodlight 7  5 (lighting) Factorial Experimental Design was then utilized.
in., lensed, 500-W quartz fixtures —each lamped with one Analysis of variance (ANOVA) was used to test for sig-
quartz tubular Designers Edge 500 W T-4 lamp. We nificant differences among means of the three dependent
determined the exact quantity, location, and aiming of these variables — Time at Display (Time), Number of Items
lighting fixtures by following IES guidelines for retailers Touched (# Items Touched), and Number of Items Picked
(Recommended practice. . ., IES Merchandising Lighting Up (# Items Picked Up) — with the two independent vari-
Committee 1997). While working within the existing con- ables — stores and supplemental lighting treatments — and
straints of each store, we succeeded in achieving lighting to test for interactions. Bonferroni’s paired comparisons
quantity and quality near the light levels recommended by were calculated to determine the difference among means.
IES for hardware and apparel display spaces. The lighting Age was used as a covariate.
treatment was turned On or Off daily during the experiment
in each store, alternating the On or Off lighting treatment
each Friday and Saturday for 8 h/day per display. The 5. Results and discussion
lighting treatments were counterbalanced.
Consumers in both stores were observed to be mostly
4.2. Data collection white, with a smaller percentage of blacks, and a very small
percentage of other races represented. A large majority of
A video camera, which was positioned in the ceiling near consumers in both stores were between 20 and 59 years old.
the selected display in each store, recorded the shopping Most consumers in the hardware store appeared to be men,
behaviors of customers. The videotapes were then reviewed but an almost equal number of men and women were
by two raters, and consumer behaviors of interest were observed in the western apparel/feed store.
counted, timed, and recorded using a Consumer Behavior Not surprisingly, significant relationships existed be-
Code Sheet derived, with modifications, from the instrument tween dependent variables as indicated by the correlations
148 T.A. Summers, P.R. Hebert / Journal of Business Research 54 (2001) 145–150

between Time and # Items Touched (r = 0.60, p < 0.001); Table 2


Means and standard deviations of Time at Test Displays, Number of Items
Time and # Items Picked Up (r = 0.53, p < 0.001); as well as
Touched, Number of Items Picked Up
# Items Touched and # Items Picked Up (r = 0.67,
p < 0.001). This may be attributed to the necessity of a Supplemental lighting treatment M (S.D.)
consumer remaining at the test display in order to touch an Store Display On Off
item, and the necessity of a consumer remaining and touch- Time at test displays
Tools 5.25a (7.62) 8.58c (17.50)
ing a merchandise item in order to pick it up from the Belts 14.49be (30.98) 12.82de (25.44)
display. Areni and Kim (1994) reported similar correlations.
Number of items touched
5.1. Approach– avoidance behavior Tools 00.08a (00.45) 00.12ad (00.57)
Belts 00.63b (3.09) 00.33cd (2.06)
Results from the ANOVA, as shown in Table 1, revealed Number of items picked up
a highly significant difference in the main effect of display. Tools 00.02a (00.18) 00.04a (00.22)
Consumers remained longer at the belts (M = 13.44, S.D. = Belts 00.11b (00.57) 00.04a (00.30)
27.65) than at the tools (M = 6.75, S.D. = 13.14). No Means with different letters differ significantly at p < 0.05.
significant difference was found in Time by the main effect
of lighting treatment.
ANOVA of the interaction of display by lighting treat- The ANOVA of the interaction of display by lighting
ment revealed a highly significant difference in Time. treatment revealed a highly significant difference in the
Means and Bonferroni results are shown in Table 2. Con- # Items Touched (Table 1). Means and Bonferroni results
sumers remained longer at the tool display under the Off are shown in Table 2. Consumers touched significantly
treatment than the On treatment. No significant differences more belts under the On treatment than the Off treatment.
were found for Time at the belt display between the On and No significant differences in means were found for tools. No
Off treatments. significant differences in the # Items Touched by the
Significant differences in Time by the covariate of age covariate of age were found.
were also found (Table 1). Bonferroni results showed the A highly significant difference in # Items Picked Up
12-years-and-under age group (M = 15.33a) spent the was found by the main effect of display (Table 1). Con-
longest time at test displays in both stores followed by sumers picked up more belts (M = 0.06, S.D. = 0.42) than
13 – 19-year-olds (M = 11.59ab). The 20 – 39-year-olds tools (M = 0.03, S.D. = 0.20). A significant difference in
(M = 11.10b), 40– 59-year-olds (M = 10.77b), and 60- # Items Picked up was found by the main effect of lighting
years-and-above consumers (M = 9.51b) spent the least treatment. Consumers picked up more items under the On
time at the displays. lighting treatment (M = 0.07, S.D. = 0.45) than under the
A highly significant difference was found in # Items Off lighting treatment (M = 0.04, S.D. = 0.29).
Touched by the main effect of display (Table 1). Consumers As shown in Table 1, the ANOVA for the interaction
touched more belts (M = 0.45, S.D. = 2.50) than tools of display by lighting treatment revealed a highly sig-
(M = 0.10, S.D. = 0.51). Consumers touched significantly nificant difference in # Items Picked Up. Means and
more items under the On treatment (M = 0.41, S.D. = 2.4) Bonferroni results are shown in Table 2. Consumers
than under the Off treatment (M = 0.28, S.D. = 1.80). picked up significantly more belts under the On treatment
than the Off treatment. No significant differences in
means were found for # Items Picked Up between the
On and Off treatments for tools. No significant differ-
Table 1 ences in the # Items Picked Up by the covariate of age
ANOVA of consumer approach behaviors at test displays
were found.
F
Time at # Items # Items 5.2. Subset of approach – avoidance behavioral results
Source df display touched picked up
Main effects Following our methodology, all persons recorded by the
Store display 1 75.34*** 33.16*** 12.06** video cameras within the measurement zones were in-
Lighting 1 1.19 3.95* 4.96*
cluded in the total data set as previously described. The
Two-way interaction
Display  lighting 1 11.15** 6.84** 13.53*** majority of consumers observed paused less than 3 s at the
Covariate test displays and did not pick up or touch an item. We
Age 1 9.21** 2.96 0.12 believed it important to the integrity of this exploratory
Error 4729 573.50 4.32 0.13 study to retain data on these passersby as they served as a
* p < 0.05.
reference group, seemingly unaffected by the independent
** p < 0.01. variables. However, large variances resulted for Time,
*** p < 0.001. # Items Touched, and # Items Picked Up, and standard
T.A. Summers, P.R. Hebert / Journal of Business Research 54 (2001) 145–150 149

deviations were always higher than means for these depen- significantly more time at the displays under the On
dent variables. treatment than the Off treatment. Results support Mehra-
We were curious to learn if additional information bian’s (1976) contention that increased levels of lighting
could be gleaned by creating a subset of the total data will produce arousal and pleasure and increase the ap-
set that would include only those behaviors of consumers proach behaviors of consumers.
who remained at the test displays longer than 3 s or Differences were noted in consumer behaviors under
touched or picked up one or more items. Therefore, the lighting treatments as qualified by store display. We
additional analyses to determine differences in the effect chose to study consumers in two types of stores concur-
of lighting treatment and store display on a subset of the rently to account for extraneous community variables and
dependent variables were performed. The following three also to explore differences that might occur because of
variables, modified from the original dependent variables, store type. Results showed that variations did exist by
were analyzed: Time at Display —4 s or more (Time + 4), store type; however, because of the nature of this ex-
Number of Items Touched — one or more (# Items ploratory study, our emphasis was on differences within
Touched + 1), and Number of Items Picked Up — one or rather than between each store. Consumers’ approach
more (# Items Picked Up + 1). behavior may have been influenced by weak lighting
Results from the Subset ANOVA indicated that, like contrasts between supplemental merchandise lighting and
findings from the primary analyses for Time, there was a ambient lighting. The IES Handbook (Rea, 1993) states
highly significant difference in the modified variable, ‘‘. . . as the background luminance (and therefore the
Time + 4, by the main effect of store display, F(1) = retinal illuminance) increases, the contrast threshold de-
81.88, p < 0.000. Consumers remained longer at belts creases, rapidly at first and then more slowly . . . ,’’
(M = 20.83, S.D. = 34.36) than at tools (M = 10.52, ( p. 87). Human visual systems are designed to react to
S.D. = 17.82). In an important contrast with the primary changes in illumination within the visual field, and these
analysis, a significant difference was found in Time + 4 by changes may be thought of as contrasts. ‘‘Contrast detec-
the main effect of supplemental lighting treatment, F(1) = tion is a basic visual task from which many other visual
7.76, p < 0.005. Consumers remained longer at the test behaviors are derived,’’ (Rea, 1993, p. 86). Therefore,
display under the On lighting treatment (M = 21.76, SD = consumers could gain visual information from a retail
35.81) than under the Off lighting treatment (M = 16.11, environment based on the visual contrast between the
SD = 28.07). ambient lighting and the merchandise display lighting thus
Like the primary analyses, a highly significant difference influencing their approach behaviors. Further study is
was noted in the Subset ANOVA for the modified variable needed to test the influence of contrast threshold.
# Items Touched + 1 by the main effect of display, F(1) = In addition, the effect of store display may include
15.04, p < 0.000. Consumers touched more belts (M = 5.78, several factors that could potentially influence consumer
S.D. = 7.09) than tools (M = 1.97, S.D. = 1.28). Contrary to approach behavior. Merchandise content, need, or lack of
primary findings, no significant difference in # Items need to try on merchandise, and pricing methods may be
Touched + 1 was found by the main effect of lighting important influences in concert with light levels. Tools, as
treatment, F(1) = 0.146, p < 0.70. compared to belts, may have inherently required less time
Subset ANOVA results, like findings from the primary for consumer selection. Tools did not require size compat-
analyses, revealed a highly significant difference in the ibility while belts were often tried on in front of the test
modified variable # Items Picked Up + 1 by the main display. This required consumers to remain at the test
effect of display, F(1) = 11.46, p < 0.001. Consumers display and may have contributed to the longer Time
picked up more belts (M = 1.93, S.D. = 1.35) than tools noted for belts. Additionally, tool prices were found on
(M = 1.18, S.D. = 0.389). In contrast with primary the front face of the display fixture minimizing time
findings, no differences in # Items Picked Up + 1 by the consumers spent in determining prices. For belts, however,
main effect of lighting treatment, F(1) = 2.09, p < 0.15, pricing was indicated only on the back side of individual
were found. belts with prices missing on some of the belts. No prices
were displayed on the display fixture. In order to deter-
5.3. Discussion mine prices, consumers were required to touch or pick up
individual merchandise items.
Findings from this study indicate that light levels do Age influenced Time but did not influence # Items
contribute to consumer approach behavior. Like Areni and Touched nor # Items Picked Up. Children spent relatively
Kim’s (1994) findings of consumers examining and hand- more time at the test displays than older consumers did.
ling more items under ‘‘bright lighting’’ than under ‘‘soft They were observed playing at the test displays in both
lighting,’’ consumers touched more items and picked up stores. Children in the hardware store may have been
more belts with the addition of display lighting. When further enticed to remain within the video camera’s mea-
passersby were omitted from the analyses, the influence of surement zone due to the close proximity of gum ball
lighting was even more evident with consumers spending machines to the test display.
150 T.A. Summers, P.R. Hebert / Journal of Business Research 54 (2001) 145–150

5.4. Limitations install temporary, portable fixtures as utilized in this study.


These lighting fixtures are readily available at a relatively
Due to the complex nature of purchase behavior, the low cost from home improvement stores. Our findings
emphasis of our study was limited to the effect of lighting suggest that a retailer’s manipulation of the in-store sup-
on in-store shopping behavior. We had no control over the plemental display lighting may achieve a significant in-
items initially chosen for each display nor the frequency crease in general consumer involvement with in-store
with which the displays were replenished. In order to remain display merchandise.
unobtrusive, we did not influence the composition or layout
of the merchandise displays but only manipulated the
supplemental lighting treatments.
Data on all subjects captured by the video cameras in References
the measurement zones in this study were retained and
Areni CS, Kim D. The influence of in-store lighting on consumers’
reported. In future studies, researchers may wish to elim- examination of merchandise in a wine store. Int J Res Mark 1994;
inate the passersby or children from their analyses, espe- 11(2):117 – 25.
cially if the retail store studied does not include children in Baker J, Levy M, Grewal D. An experimental approach to making retail
its target market. store environmental decisions. J Retailing 1992;68(4):445 – 61.
Boyce PR, Lloyd CJ, Eklund NH, Brandston HM. Quantifying the ef-
fects of good lighting: the green hills farms project. Meeting of the
Illuminating Engineering Society of North America, Cleveland, OH,
6. Implications August, 1996.
Cuttle C, Brandston H. Evaluation of retail lighting. J Illum Eng Soc
1995;24(2):33 – 49.
We believe this study provides a sound methodology for
Donovan RJ, Rossiter JR. Store atmosphere: an environmental psychology
the further examination of the effect of lighting on in-store approach. J Retailing 1982;58(1):35 – 57.
consumer behavior. Based on approach – avoidance theory Donovan RJ, Rossiter JR, Marcoolyn G, Nesdale A. Store atmosphere and
(Mehrabian, 1976), the study extends Areni and Kim’s purchasing behavior. J Retailing 1994;70(3):283 – 94.
(1994) research in retail field settings. Unlike previous Gardner MP, Siomkos GJ. Toward a methodology for assessing effects of
in-store atmospherics. Adv Consum Res 1986;13:27 – 31.
studies, our methodology included a full documentation of
Gobe M. Visceral merchandising. Visual Merchandising and Store Design
physical conditions of the test stores’ interiors. The actual (December) 1990;121:16, 18.
store environments utilized in this study were realistic Hoyer WD. An examination of consumer decision making for a common
settings in which to test hypotheses about consumer beha- repeat purchase product. J Consum Res 1984;11:822 – 9.
vior. Additionally, this study also included a relatively large IES Merchandising Lighting Committee. Recommended Practice for
Lighting Merchandising Areas: A Store Lighting Guide. Manuscript
number of observations from two different store types.
in preparation.
Because of the limited amount of empirical research on Kotler P. Atmospherics as a marketing tool. J Retailing 1973;4(4):48 – 64.
the effects of lighting on consumer behavior, we recom- Lopez MJ. Retail Store Planning and Design Manual. New York:
mend additional field investigations be conducted. Wiley, 1995.
Based on findings from this study, retailers could Mehrabian A. Public Spaces and Private Spaces: The Psychology of Work,
Play, and Living Environments. New York: Basic Books, 1976.
supplement the illumination on their merchandise displays
Mehrabian A, Russell J. An Approach to Environmental Psychology.
in compliance with IES recommendations for their store Cambridge, MA: MIT Press, 1974.
type in order to attract consumers and involve them in Rea MS, editor. Lighting Handbook Reference and Application (8th edn.).
their merchandise. Retailers with limited funds could New York: Illuminating Engineering Society of North America, 1993.

You might also like