Archives of Design Research

The effects of attention and spatial complexity can cause unique visual experiences or feelings in an environment. Attention refers to the ability to focus or concentrate (Kopec, 2006). In a lighting environment, attention is involuntarily drawn toward areas of brightness that contrast with the visual background (Gordon, 2003). In a retail store, the pattern of light posted in certain locations, such as on the wall at eye level or above, can influence consumers’ emotional responses (Boyce, 2003). According to findings of lighting studies, in general, bright light attracts people’s attention better than dim light; however, appropriate combinations of bright and dim light can produce a visual contrast that can create stronger images to attract attention and produce attraction to a specific lighting setup (Nuckolls, 1983).

Spatial complexity refers to visual richness, ornamentation, information rate, diversity, and variety in an environment (Custers, Kort, & Ijsselsteijn, 2010; Nasar, 2000). Berlyne (1971) explored the effects of environmental complexity on subjects’ emotion and confirmed a linear relationship between environmental complexity and interest, and a curvilinear relationship between environmental complexity and preference. Nasar (1987) and Health, Smith, and Lim (2000) found a positive relationship between spatial complexity and subjects’ arousal levels, while Wohlwill (1976) and Nasar (1997) found a negative link between spatial complexity and pleasure. Gilboa and Rafaeli (2003) studied the effects of environmental complexity in a retail environment by testing relationships between complexity, order, and three emotions— pleasure, arousal, and dominance—that mediate approach and avoidance behaviors. They found a positive relationship between complexity and arousal, and a curvilinear relationship to consumers’ approach intention.

Flynn et al., (1973) tested combinations of intensity and patterns of light (the images and effect of light on the walls and ceilings) in a conference room and found that various combinations of lighting, including intensity and patterns of light, can affect viewers’ impressions of and satisfaction with visual clarity and spatial complexity.

Pleasure refers to the feeling of being pleased and gratified with surrounding environments (Steffy, 2002). Arousal refers to the levels of a person’s feelings by stimulation or excitement (Baker et al., 1992). Much research shows that emotion and motivation for shopping are inextricably linked such that many people make purchases for emotional reasons (Runyon & Stewart, 1987). Therefore, emotional responses frequently elicit various consumer behaviors. Some researchers claim that dimmer lighting can induce more pleasure (Custers et al, 2010), while others report more pleasant emotions with higher lighting levels (Fleischer et al., 2001; Hygge & Knez, 2001; Liao, 2011). Fleischer et al. (2001) found that high intensity lighting was more pleasant than a low intensity lighting. Similarly, Sorcar (1987) found that people experience more pleasantness in a brightly lit environment with interesting light patterns created by spot lighting and wall washing.

For arousal, Steffy (2002), Gifford (1988) and Mehrabian (1976) found that bright lighting can elicit active feeling and arousal, whereas dim lighting can induce little to no arousal. With a high contrast lighting environment that renders patterns of light and shade, a hierarchy between foreground and background can increase stimulation in a space (Gordon, 2003). Sorcar (1987) reported that brightly lit spaces with various shadows and lighting patterns can form a dramatic and exciting atmosphere.

Consumers’ satisfaction in a store is an important determinant of approach-avoidance intention (Cheng et al., 2003; Voss et al., 1998). Consumers’ performance assessments are based not only on physical attributes of products or retail environments, but also on their emotional responses to physical cues (Burns & Neiser, 2006). According to Spreng and Chiou (2002), there is generally a strong link between satisfaction and repurchase intentions. In addition, many consumer studies also indicate that there is a positive association of satisfaction and intentions (Bigné, Andreu, & Gnoth, 2005; Mittal, Kumar, & Tsiros, 1999). According to Reddy, Reddy, and Azeem (2011), successful image-oriented lighting can create special scenes with varied light and shadow, thus inducing higher consumer satisfaction with merchandise and the retail environment. These findings suggest that lighting design in a retail store should be more oriented toward creating attractive images and atmospheres of the store rather than creating a bright ambient environment for efficient product search tasks.

Regarding effects of lighting on approach-avoidance intention, Summers and Hebert (2001) examined the effects of lighting contrast on display areas, and showed that the stronger contrast of display lighting and background lighting facilitates shoppers with higher approach intention; they stayed longer and examined more products. Some lighting studies suggest that if people have higher levels of pleasure in a bright retail space, they should also have higher approach intentions (Baker et al., 1992; Spies, Hesse, & Loesch, 1997). Other studies show that arousal has a positive effect on approach behavior (Baker et al., 1992; Donovan & Rossiter, 1982). According to Gordon (2003), a higher contrast level was found to attract consumers’ attention to merchandise and encourage them to stay longer in a retail environment.

Traditional measurement and experiments of lighting are often held in the field or laboratories with different light sources (Boyce et al., 2006). In order to satisfy needs for different lighting experiments, conducting traditional lighting experiments can be very costly with various lights, equipment and spaces to set up (Pinto et al., 2010). Another reliable method, mock-up, is often used to simulate the performance of lighting. However, mockups still have the same problems as traditional lighting laboratories: higher costs and longer setting and experimental time.

In recent years, computer simulation has become increasingly popular as a tool to explore various issues of lighting, human perception, and environmental phenomena (Newsham et al., 2005). Although computer lighting simulations probably cannot produce 100% realistic light, they are still greatly useful to help researchers conduct lighting experiments and also helping designers pre-test and determine the setting for lighting (Newsham et al., 2005).

Recently, 3D computer animation simulation has been widely used in the architecture and interior design fields including exploring issues of physical environments, human behavior, and psychology. Bateson and Hui (1992) state that environmental psychologists and marketers have shown simulated environments can lead to similar results that would be found in real environments. The advantages of using 3D computer simulation include the ability to derive more visual and emotional reactions from subjects and produce collected data with higher value for professionals and designers to refer to and use (Mania & Andrew, 2004). In addition, 3D computer simulation is ideal for creating controlled conditions for observing the psychological reactions of people to their environment and in many situations, the reactions of people to virtual environments are similar to those in real environments (Hwang et al., 2012; De Kort et al., 2003). According to the findings from Lin and Yoon’s (2011) study with an exhibition space, lighting simulation using computer simulation can successfully replicate physical lighting conditions. Findings indicated that there was no significant difference in perceived visual clarity, pleasure, and satisfaction between the real and virtual exhibition environments. The reliability of 3D computer simulation to test lighting effects and investigate emotional and psychological reactions has gained growing support.

Consumer electronic retail stores are popular places for shopping for computers, laptops, tablets, and smartphones. To successfully create an interesting atmosphere and shopping environment for such products it is necessary to explore the relationship between electronics retail environments, electronic products, and shoppers. In particular, young shoppers, such as college students, are one of the main consumer groups. However, there is a lack of empirical evidence available for designers and retailers to base their decisions when attempting to provide optimal consumer experiences and addressing various needs in an electronics retail setting. In this study, Mehrabian and Russell’s model (M-R model) was used to test the effects of retail lighting on consumers’ emotions and intentions. Among the many factors to consider, lighting contrast and color temperature in a retail environment were selected for aforementioned reasons to study consumer responses in terms of: 1) perceptions –attention and spatial complexity, 2) emotional responses – pleasure and arousal, 3) store satisfaction and 4) approach and avoidance intentions. In order to conduct different lighting tests with contrast and color temperature of lighting, a 3D walk-through simulation was used to capture subjects’ perceptions, emotions, and psychological responses.

Eighty subjects were recruited from the University of Missouri in Columbia, Missouri. The criteria used to select the appropriate subjects include: (1) 19 to 40 years old, (2) no vision impairment that affects recognition of objects, and (3) without experience of working in lighting fields. The age range of participants was kept from 19 to 40 years old for three reasons (Park, 2001; Pae, 2009). The first reason is to avoid possible age effects from different age groups, such as the elderly or children. The second reason is that the young student group is one of the main shopping groups in the electronic products market because most of them need computers and tablets for their academic courses and assignments. The third reason is that student groups have similar behavior and social activities on campus; therefore, using student subjects for this study can achieve a higher consistency of sample behavior and experiences.

The experiments were conducted in a laboratory at the University of Missouri in Columbia. The size of the laboratory is about 15’ by 12’, and the height is 8’. During the experiment, the interior lights were turned off. Except for the images presented on the screen, the entire room – including the space behind the screen – was kept dark. There are no windows in the experimental environment.

A between-subjects design was chosen for the study to avoid carry-over effects. The main assumption of the experiment is that the combinations of contrast and color temperature of lighting have direct influence on subjects’ perception and psychological responses in an electronics retail store. For the retail environment stimulus, an electronics retail store was selected because electronic goods such as computers and tablets are popular amongst college students and they can relate to various electronic devices and the environment. 3D computer animations were used to simulate a life-like browsing experience in the space for the study. The main reason for choosing a 3D computer animation over a real-time walk-through for this study is that doing so enables us to expose participants to the identical experience with a pre-defined path and duration. The vision of all participants followed the default pathway, and they all saw the same images and spots in the virtual electronics retail environment during the animation. Autodesk 3D Studio Max Design 2012 is a mainstream software package widely used in the fields of architecture and interior design. The program features a varying photorealistic rendering engine that offers an advanced lighting simulation function. The software was used to create virtual retail environments and the 3D scenes were rendered to animation.

The size and style of the experimental virtual retail space were identified through retail space design and interior design guide books (Custers, et al., 2010; Karlen & Benya, 2004). The size of the virtual retail space is approximately 1,300 sq. ft. with a rectangular floor plan, a layout that reflects a typical medium-sized retail store that is freestanding or in a mall (Karlen & Benya, 2004). The ceiling height is about 14’. Layout of furniture, materials of the walls, ceilings, and floors, color scheme, and lighting solutions were chosen to fit current trends of retail store design. Interior design and architecture professionals reviewed the design of the environment during the production stage. The virtual retail environment was projected on a large screen (60” x 60”). The size of the image on the screen was around 57” x 43”. A chair and table were placed in front of the big screen, and the participants sat 6’ from the screen for optimal viewing. The length of the animation was 40 seconds. After watching the entire animation without interruption, participants were asked to answer a series of questions using a written questionnaire. As participants answered the questions, the animation replayed continuously on the screen. This study did not restrict the time allotted for completing the questionnaires so that participants could simulate a more natural shopping experience.

The lighting settings for the study were based on the guidelines from the handbook (IESNA, 2011) and the practical report of retail lighting (IESNA, 2001) published by the Illustrating Engineering Society of North America. According to the IESNA handbook, the suggested lux of lighting in interior circulations and a retail store’s whole environment ranges from 100 to 300 lux. For merchandise areas such as the surfaces of display tables, the suggested light levels are from 300 to 1000 lux. Based on these guidelines and the findings from the pilot lighting tests, this study adopted 20 lux as ambient illumination and 1000 lux for merchandise illumination for the high-contrast lighting condition, as well as 300 lux as ambient illumination and 1000 lux for merchandise illumination for the low-contrast lighting condition. The color temperature variable is also tested at two levels: high color temperature conditions, and low color temperature conditions. The high color temperature condition has 5300K cool lighting and the low condition has 3200K warm lighting.

The present study’s retail display space had 68 lights including 20 down lights, 36 spotlights, and 12 wall washers. Spotlights were put in at a 45-degree angle in front of display tables and the surrounding walls including one back wall and two side walls. The back wall had 5 wash lights, the right side wall had 4 wash lights, and the left side wall had 3 wash lights. Two lighting researchers who are university faculty members and two lighting experts from IESNA reviewed the arrangements and conditions of lighting for this study. Figure 1 shows which lights were turned on in low-contrast and high-contrast lighting conditions, and Table 1 presents four 3D virtual lighting environments.

Figure 1 Low- and high-contrast lighting conditions

	Contrast of light
Color Temp.	Low contrast Ambient light: 300lux Merchandise illumination: 1000lux	High contrast Ambient light: 20lux Merchandise illumination: 1000lux
Cool Color 5300K
Warm Color 3200K

High-contrast lighting: luminance ratio of 50 to 1 (between merchandise illumination and ambient illumination)

Low-contrast lighting: luminance ratio of 3 to 1 (between merchandise illumination and ambient illumination)

A self-report questionnaire was developed based on pilot studies and other related studies. Most question items were directly adopted from instruments with established validity and reliability from previous studies, but some items were modified for the study. The questionnaire consisted of two main sections. The first section included demographic variables such as gender, age, visual impairment, whether or not the participant has worked as a lighting professional, and how often the participant goes shopping in stores. In the second section, the dependent variables for the study were the subjective experience of four lighting treatments. There were questions on six variables: attention, spatial complexity, pleasure, arousal, store satisfaction, and approach-avoidance intentions. All questions for the six dependent variables were directly selected or have been modified from related studies to fit the needs of this study examining retail store environmental settings with four different lighting conditions.

The items for attention were modified from lighting report IESNA (1993) and from cognition studies of Nisbett et al. (2001) and Nisbett and Miyamoto (2005). The questions for spatial complexity variable were selected from lighting studies of Flynn et al. (1973), IESNA (1993), Mullen and Johnson (1988), and cognition studies of Nisbett et al. (2001). The question items for pleasure were selected from studies by Baker et al. (1992) and Pae (2009). Arousal questions were selected from previous studies by Baker et al. (1992) and Bearden et al. (2011). Store satisfaction questions were selected from Pae (2009)’s study. The approach-avoidance intentions questionnaire was developed based on Baker et al (1992)’s study. Participants were asked to indicate their answers on a 7-point Likert scale ranging from 1 (strongly disagree) to 7 (strongly agree).

Appointments for participation in the virtual retail environment were scheduled with one to six subjects at a time. Data was collected through a survey questionnaire given to each subject. A chair and table were placed in front of the big screen, and the participants sat 6’ from the screen for optimal viewing. Before the beginning of the experiment, clear explanation regarding the purpose and procedure was given to the participant, and the researcher answered questions or concerns for participants and presented the Informed Consent Form. After the consent form was signed, the participant received a standard oral introduction to explain the process of the experiment. When the participants started to read the questionnaires, they read an introduction and hypothetical scenario in the first section of the questionnaire. Hypothetical scenarios can prime the participants to produce a higher feeling of involvement for the experiment.

This study used the SPSS statistical package to analyze survey data. A series of analyses of variance (ANOVA) were used to analyze the differences between variables. A 2 (contrast of light) x 2 (color temperature of light) factorial experiment design was utilized to test six dependent variables. Reliability analysis using Cronbach’s alpha was employed to test the internal consistency of scales that were used to measure the variables. Dependent variables consist of attention (6 items, α = 0.84), spatial complexity (3 items, α = 0.70), pleasure (5 items, α = 0.87), arousal (6 items, α = 0.77), store satisfaction (6 items, α = 0.96), and approach-avoidance intentions (7 items, α = 0.86). Since each of these alpha levels is above the acceptable threshold for reliability (George & Mallery, 2003), these variables were used for further analysis. A p-value of 0.05 was used to determine statistical significance for the study.

The 80 participants included 26 males (32.5%) and 54 females (67.5%). 11 participants’ were younger than 19 years old (13.8%) and 69 participants were 19-24 years old (86.3%). The range of shopping frequency is from “never” to “always”. No participant selected “never”. Five participants selected “seldom” (6.3%), 40 participants selected “sometimes” (50.0%), 27 participants selected “often” (33.8%), and 8 participants selected “always” (10.0%).

The contrast and color temperature of the lighting influenced the participants’ perception, attention, and spatial complexity states. Table 2 presents the mean and standard deviation scores for participants’ attention and spatial complexity states.

Table 3 displays the effects of light contrast and color temperature on participants’ attention and spatial complexity. The results show two main significant effects on participants’ attention. The first main effect is of light contrast, F(1, 76) = 10.272, P < .01, showing participants perceiving high-contrast lighting (M = 5.49, SD = 1.05) as more attention-drawing than low contrast lighting (M = 4.78, SD = 1.20). Hypothesis 1 is accepted in that subjects have higher attention states in high-contrast lighting conditions than in low-contrast lighting conditions.

The second main effect is of lighting color temperature, F(1, 76) = 5.868, P< .05 which shows that participants (M = 5.40, SD = 1.09) in the cool color temperature condition have a higher level of attention than in the warm color temperature condition (M = 4.86, SD = 1.21). However, no interaction between light contrast and color temperature was observed. The result of the measurement of spatial complexity shows that there are no main effects or interactions on participants’ feelings about spatial complexity. Hence, this result rejects hypothesis 2 that a subject will perceive higher levels of spatial complexity in high-contrast lighting conditions than in low-contrast lighting conditions when overall luminance is kept the same.

The contrast and color temperature of lighting influenced participants’ emotion, pleasure, and arousal states. Table 4 presents the mean and standard deviation scores for participants’ pleasure and arousal states.

Table 5 displays the effects of light contrast and light color temperature on participants’ pleasure and arousal states. The results show that there is a significant main effect of light contrast F (1, 76) = 7.938, P < .05 on pleasure in participants. This main effect shows that participants perceiving high light contrast (M = 5.18, SD = 1.12) feel more pleasure than those who perceive low light contrast (M = 4.55, SD = 1.13). However, no interaction between lighting contrast and lighting color temperature was obtained. Hence, this result rejects hypothesis 3 that subjects will perceive higher levels of pleasure in a high-contrast warm lighting condition.

For states of arousal, the result shows that there is a significant main effect and interaction effect. The main effect is of light color temperature, F(1, 76) = 5.689, P < .05. A significant two-way interaction of light contrast and light color temperature was obtained with a calculated F(1, 76) = 4.050, P < .05. Figure 2 shows the interaction effect of light contrast and color temperature on participants’ arousal states.

Figure 2 Interaction effect of light contrast and color temperature on arousal

Table 6 shows the results of a simple main effect for light contrast by light color temperature. Analysis of the simple main effect shows that high light contrast, F(1, 38) = 9.687, P < .01, and cool light color temperature, F(1, 38) = 3.035, P < .05, have significance. Hence, participants have higher levels of arousal in high-contrast light (M = 4.32, SD = 1.11) than in low-contrast light (M = 4.22, SD = .89). In addition, participants have a higher level of arousal in cool color temperature (M = 4.53, SD = .84) than in warm color temperature (M = 4.00, SD = 1.08). According to the results from the simple main effect, hypothesis 4 was confirmed: subjects will feel higher levels of arousal in a high-contrast, cool color condition.

The contrast and color temperature of lighting influenced participants’ store satisfaction and approach-avoidance intentions. Table 7 presents the mean and standard deviation scores for participants’ store satisfaction and approach-avoidance intentions.

Table 8 displays the effects of light contrast and light color temperature on participants’ store satisfaction and approach-avoidance intention states. The results show that there are no main effects or interactions on participants’ store satisfaction. Therefore, the results reject hypothesis 5 that subjects feel higher levels of store satisfaction in a high-contrast lighting condition. However, the p-value of lighting contrast reaches .06, so the statistical significance of store satisfaction may be obtained in a larger sample size.

For the states of approach-avoidance intentions, the results show that there is a significant main effect on approach-avoidance intentions in participants. This main effect is light contrast F(1, 76) = 3.433, P < .05, which shows that participants perceiving high-contrast lighting (M = 5.40, SD = .91) have higher approach-avoidance intentions than when perceiving low-contrast lighting (M = 4.99, SD = .80). This result supports hypothesis 6 that subjects have higher approach intention in a high-contrast lighting condition; however, no interaction between light contrast and light color temperature was obtained.