‘To LED or Not to LED?’: Using Color Priming for Inﬂuencing Consumers’ Preferences of Light Bulbs

: Recently, we have seen energy-efﬁcient light-emitting diode (LED) light bulbs rapidly replace incandescent ones. However, results of new research are indicative of adverse health impacts of LED lighting, which is characterized by enriched blue light. Our study aims to reveal whether using color priming by attaching red/green trafﬁc-light icons on light bulbs inﬂuences consumers’ preferences of light bulbs. We conducted a ﬁeld study simulating the buying process, in which participants (N = 572) were presented with LED and carbon incandescent bulbs. We alternately displayed two pairs of bulbs: (1) in their original packaging and (2) in packages marked with trafﬁc light icons (red = LED). Our results conﬁrm that trafﬁc light icons signiﬁcantly ( p < 0.01) increase the odds of choosing the healthier carbon bulb. The results highlight the beneﬁts of attaching trafﬁc light icons to light bulb packaging, helping consumers to make more health-conscientious purchasing decisions. Nowadays, this study’s contribution is more signiﬁcant due to COVID-19 restrictions and stay-at-home policies, since people work or study remotely, which increases their exposure to household lighting. These results may incentivize policymakers to enforce adding trafﬁc light icons to light bulb packaging, thus encouraging LED light bulb manufacturers to reduce the blue light component in order to improve the health aspect of their bulbs.


Introduction
As environmentalists are worried about inefficient illumination, in many countries there has been a move towards efficient bulbs based on emissions of short-wavelength (SWL) lighting. The increased usage of light-emitting diode (LED) lighting has generated interest in the scientific world regarding its ramifications for our health. However, it appears the general public is not aware of the health risks connected to energy-saving bulbs emitting SWL illumination.
In the current study, we aim to reveal whether emphasizing the health aspect of light bulbs influences consumer preferences towards healthy illumination. Therefore, we used color priming as a means of providing health information to consumers regarding light bulbs.
Priming is defined as an unconscious remembering process when certain stimuli increase the availability of information categories [1]. Priming may affect information processing and decision-making [2].
A distinctive form of priming is by color, which is based on learned association developed from repeated pairings of colors with specific messages, or biologically based tendencies to respond differently to specific colors [3]. The significance of colors' effects in numerous psychological fields, including advertising and applied psychology, has been noted [4], specifically an association between the color red and avoidance motivation. Although some claim that responses to colors differ between genders, cultures, and other personal characteristics, careful use of colors can have a negative or positive influence on consumers [5].
From an early age, we are taught that the red light in the traffic light symbolizes danger, so we should stop and avoid crossing. In contrast, the green light means that it is safe to cross. A possible explanation is that red is evolutionally associated with status and dominance, which can be interpreted as a warning sign [6].
The results of a study revealed that in financial decisions, red color priming emphasizes value losses of the underlying asset [7]. Likewise, green has positive effects on several aspects of consumer behavior. The association of green with fertility or vegetation may provide an explanation for these positive effects [8].
As a result of these color effects, package design components as colors are often used in the food industry to distinguish between healthy and unhealthy products [9]. In a study estimating how different labels affected consumer choice behaviors and product perceptions, the results suggested that front-of-package nutrition labels featuring traffic light ratings have more effect on shifting food choice patterns compared to other methods [10].
So far, the use of labels that include a color scale has been implemented on light-bulb packages only regarding the energy-efficiency characteristics of the bulbs [11]. However, in the current study, our main objective is to reveal whether attaching health-related red/green traffic light icons (TLI) on light bulbs influences consumers' preferences towards choosing the healthier bulb and avoiding the energy-efficient bulb, which has been shown to be unhealthy [12]. Hence, the novelty of this work consists of implementing a method that was used in other areas on a new field with specific benefits to human health.
The next section includes the literature review as well as the hypothesis and research question development. We begin with discussing the background on LED lighting and its global usage. Then we present the adverse impacts of LED light bulbs that have been discovered in recent years. Afterwards we explain the psychological tool that we used to influence consumers towards choosing the healthier light bulb.
The contribution of this study is even more crucial nowadays due to the quarantines inflicted upon people since the COVID-19 breakout. In the past year, the COVID-19 pandemic has wrought havoc throughout the globe. Therefore, the aim to encourage consumers to buy healthier bulbs has become much more important due to stay-at-home policies. To the best of our knowledge, no study to date has examined the impact of health TLI on consumers' choice of light bulbs.
In the rest of this article, we discuss and hypothesize about the influence of health TLI attached to light bulb packages on consumers' preferences of light bulbs. Then we use the SPSS Statistics software to test the hypothesis and research questions, using data from a field study sample of Israeli consumers and discussing the findings. This form of tangible trial, which enabled the participants to touch the light bulbs and inspect their packages closely before making their buying decision, has not been implemented before as far as we know. Finally, we draw conclusions, and offer general and current implications.

Energy-Efficient Illumination: LED Lighting
The innovation of bright blue LEDs during the 1990s has led to an increased usage of LED-based lighting. In recent years, incandescent light bulbs have been increasingly replaced by LED bulbs as an energy-efficient (EE) light source for everyday use [13]. Moreover, LEDs will probably replace incandescent bulbs and fluorescent lamps throughout the globe [14]. Visual displays terminals (VDT), which constitute standard equipment both at home and at work, use LED backlighting as well. People are surrounded by VDTs on a daily basis, in both developed and developing countries. However, LED-based white Sustainability 2021, 13, 1401 3 of 12 light is highly enriched with a blue light component (460 nm), compared to incandescent bulbs [15].

Adverse Impacts of LED Lighting as Short-Wavelength Lighting
Regardless of the advantages of LED as anEE light source, the increased usage of this type of illumination also poses health threats that should be taken into consideration when choosing light bulbs for households. The risks to human health include problems related to sleep duration, body-mass gain, eye damage, and even breast cancer [16][17][18].
Researchers have warned about the health risks emanating from the extended use of the LED technology [12]. The results of these studies show that evening exposure to SWL light causes many physiological and psychological effects, including increased alertness and heart rate, and decreased ability to regulate body temperature [19,20].
A recent study revealed that LED lighting can damage retinal pigment epithelial (RPE) cells, including induced DNA damage [16]. Several lines of evidence have established the relationship between sleep and energy metabolism [14,21].
Furthermore, short sleep duration has been identified by epidemiological studies as a risk factor for the development of obesity [18,22]. Another study showed that blue LED light exposure in the evening causes drowsiness and energy expenditure suppression the next morning [14]. Their explanation for those next-morning effects was a possible phase shift of the circadian clock due to blue light exposure, since recent findings clearly demonstrate that the human circadian pacemaker is highly sensitive to SWL lighting [23,24].
All the above data showing health challenges from LED lighting highlight the need to provide consumers with information, using an effective tool to assist them in their light bulb consumption decisions.

Using Traffic Light Icons to Influence Consumer Preferences
The use of TLI as a labeling method has spread to a wide range of applications [25]. One objective of TL-labeling regards supplying sustainability information to consumers about products. A recent study evaluated over 40,000 online purchases on the Good Guide website [26], which provides an environmental color scale for a variety of products. The most robust effects were those of health ratings, where the green light on the healthier products yielded a significant response [25].
A different field that uses TLI intervention is food and beverages. Package design components as colors are used in the food industry to distinguish between healthier and regular products [9]. In order to improve the general population's health by means of improving food and beverage quality, nutritional rating systems were developed. The main goal was to develop rating systems that could be used at the point of sale and that are easy to understand, as well as be able to guide consumers of all segments of the population [27]. One of the front-of-package systems is a TLI labeling scheme in which the level of each nutrient (e.g., fat, sugar, salt, etc.) is symbolized by a color code ranging from red to green [28]. Finally, studies in this area indicated that symbolic colors used to distinguish nutrient levels facilitate consumers to choose healthier products [29].
The aim of our study was to test whether an implementation of a red/green TLI intervention system on light bulbs influences consumer choice towards the light bulb marked as healthier according to the results emerging from scientific study.
Our hypothesis, based on the literature reviewed regarding TLI, was as follows: Hypothesis 1. Health TLI (red versus green) attached on light bulb packages is likely to influence consumers' preferences towards choosing the healthier carbon incandescent light bulb (green) over the unhealthy LED bulb (red).
Several studies confirmed that vegetarians/vegans often articulate personal health improvement as well as the need to protect the environment amongst the reasons for their dietary choice. As a result, they tend to adopt healthy and environmentally friendly behavioral patterns [30][31][32]. Furthermore, a recent study indicated a relationship between dietary choice and gender with preferences for sustainable near-food products (cosmetic and cleaning products) [33].
Regarding gender, researchers established that females seek health-related information much more actively than males [34]. Therefore, in order to reveal whether these associations also apply to light bulbs, we added the following research questions: (1) Does attaching health TLI to each light bulb's individual package influence consumers' preferences towards choosing the healthier light bulb according to whether they are vegans or vegetarians or whether they eat meat? (2) Does attaching health TLI to each light bulb's individual package have a different influence on females compared to males regarding their preferences towards choosing the healthier light bulb?
The hypothesis and research questions are presented in the model below. Testing this model had practical implications about the mechanism that underlies consumers' choice of light bulbs that are labeled with TLI ( Figure 1). Several studies confirmed that vegetarians/vegans often articulate personal health improvement as well as the need to protect the environment amongst the reasons for their dietary choice. As a result, they tend to adopt healthy and environmentally friendly behavioral patterns [30][31][32]. Furthermore, a recent study indicated a relationship between dietary choice and gender with preferences for sustainable near-food products (cosmetic and cleaning products) [33].
Regarding gender, researchers established that females seek health-related information much more actively than males [34]. Therefore, in order to reveal whether these associations also apply to light bulbs, we added the following research questions: (1) Does attaching health TLI to each light bulb's individual package influence consumers' preferences towards choosing the healthier light bulb according to whether they are vegans or vegetarians or whether they eat meat? (2) Does attaching health TLI to each light bulb's individual package have a different influence on females compared to males regarding their preferences towards choosing the healthier light bulb?
The hypothesis and research questions are presented in the model below. Testing this model had practical implications about the mechanism that underlies consumers' choice of light bulbs that are labeled with TLI ( Figure 1).  We tested the hypothesis and these research questions in a field study, using red/greed health TLI on light bulb packages versus light bulbs displayed to consumers without TLI.

Sampling, Data Collection, and Procedures
During spring-summer 2018, we conducted a field study simulating the buying process, in which 572 adult participants (58% females; 14% vegans/vegetarians) were presented with two light bulbs: LED (light-emitting diode) and carbon incandescent light bulbs. The information on the manufacturer's original packages was presented to participants, including an energy-efficiency color scale of each light bulb (see Supplementary Materials Figure S1). The prices of the light bulbs were similar and were almost the same in both groups, so that even if the slight price difference could have influenced participants' choice of light bulbs, the same difference would have been detected in the control group.
In order to get a good representation of the population, the study took place at several locations in Israel, including a university campus, community centers, restaurants, DIY (do it yourself) stores, supermarkets, and a retirement home. Table 1 provides a description of the sample demographics. We placed a table, upon which we placed the pair of bulbs and hung a poster with pictures of the light bulbs, requesting people to participate in a research study regarding their light bulb preferences. The poster also announced the option of participation in a raffle. The study included a multi-ethnic population.

Measures
First, the participants had to write which of the two bulbs they chose to purchase, after examining each bulb as long as they wished.
Stage 1: There were two sets of light bulb pairs: (1) The bulbs were presented in their original packages with no additions, and (2) red/green health TLI were attached to the original packages of the bulbs (red was attached to the LED bulb's package, and green to the package of the carbon incandescent bulb) (see Figure S2).
Participants were presented with a pair of bulbs with or without TLI, making sure the total study population was randomized, blind, and balanced.
Stage 2: We asked each participant to fill in a questionnaire. The questionnaire that was intended for the TLI version included a short introduction, explaining that we attached health TLI in which red symbolized a higher health risk and green symbolized a lower health risk. The questionnaire comprised 10 questions, including questions regarding demographics and electricity bills. In addition, the respondents had to define themselves as vegans/vegetarians or as meat eaters, answer whether they preferred to get health information regarding the light bulbs, and which was more important to them when choosing a light bulb-whether the light bulb was healthy or energy-efficient (EE).
Stage 3: With the aim of encouraging participants to take part in the study, we utilized a lottery system where they had a 10% chance of winning one of the two light bulbs for free after participating in the study.
Dependent variable: Consumer preference for a light bulb-LED or carbon bulb. Independent variables: (1) Demographic variables; (2) Means of providing information to consumers, including TLI/original package; (3) Dietary choice; and (4) Preference of the light bulb: healthy/EE.
The study protocol was approved by the Ethics Committee of the university. All volunteers confirmed their compliance by written informed consent.

Results
For the preliminary statistical analysis, we examined the differences between the proportions of participants who preferred the carbon bulb within the sample's subgroups, presented in Table 2.  Lines 1-2 show that the proportion of females who chose the carbon bulb was significantly higher than the proportion of males, both for the participants who received the light bulbs with TLI (p < 0.01) and for those who were given the original packages (p < 0.02). Both genders were robustly influenced by the TLI. However, the proportion of males who Sustainability 2021, 13, 1401 7 of 12 chose the carbon bulb as a result of the TLI increased more mildly than the proportion of females. There was a significant difference between genders.
Lines 3-4 show that the proportion of vegans/vegetarians who chose the carbon bulb was higher than the proportion of meat eaters, both for the participants who received the light bulbs with TLI (p < 0.05) and for those given the original packages (ns). Participants from both dietary choice groups were robustly influenced by the TLI. However, the proportion of meat eaters who chose the carbon bulb as a result of the TLI increased more moderately than the proportion of vegans/vegetarians.
In terms of participants' points of view, lines 5-6 show that only 19.2% of the participants who were given the original packages, who claimed that it was more important that a light bulb be EE than healthy, chose the carbon bulb, whereas 29.8% of the participants who claimed the opposite chose the carbon bulb. The TLI significantly influenced participants with both points of view. However, the increase was much steeper for the "healthy" view than for the "EE" view. The difference was statistically significant. Table 3 introduces the model summary in three levels. We calculated the odds ratios (OR) as well as the 95% CIs for choosing the carbon incandescent light bulb. According to the OR results of the regression, TLI increased the odds of choosing the carbon bulb significantly (p < 0.01). Similarly, participants who stated that it was more important that a light bulb be healthy than be EE increased the probability of choosing the carbon bulb significantly (p < 0.01). Finally, the odds of males of selecting the carbon bulb were significantly lower than those of females (p < 0.01). Step one demonstrates that the accumulative explained variance of all variables in the model was 28.1%.

Results of Hierarchical Logistic Regression
Step three, females who were presented with the light bulbs including the TLI, who stated that it was more important that a light bulb be healthy than that be EE, explained 26.6% of the variance. The total variance for step two was 26.9%.
Vegan or vegetarian consumers, having a membership in an environmental organization, and having an interest in receiving health information regarding the light bulb explained merely 0.3% of the variance. Therefore, participants' income level, education level, age, living arrangements, and electricity bill explained only 1.2% of the variance. Finally, we excluded the "Importance" variable (representing the "healthy" point of view) from the regression. The TLI remained statistically significant (p < 0.01), having a profound effect on the light bulb choice, with an odds ratios of 4.617 for choosing the carbon bulb. As for the gender variable, although the odds ratio of females for choosing the carbon bulb changed, it remained statistically significant. The results including the "Importance" variable and after excluding this variable are presented in Table 4. Abbreviations: OR-odds ratio. (Note: the interaction between TLI and the Female result is statistically nonsignificant). Based on a sample of n = 572.

Discussion
The use of LEDs as an EE light source is spreading across continents, driven by a massive marketing campaign run by LED manufacturers and retailers. However, there is growing evidence in the scientific community regarding the negative impacts of LED bulbs, characterized by enriched blue light. These include causing damage to human health, including increased prevalence of prostate and breast cancers [12], as well as other health ramifications.
Several studies have demonstrated linkage between colors and consumer preferences [5,35]. Although it was suggested that there are segments of the population that do not fully comprehend the significance of red/green in TLI [36], colors can be used as a marketing cue and are shown to have profound effects on consumers based on associations and biological instincts in most cases [37].
In the current field study, we examined whether attaching health red/green TLI on each light bulb's individual package would influence consumers' preferences towards choosing the healthier light bulb as follows: Green represented healthy-attached to the carbon bulb, and red represented unhealthy-placed on the LED bulb.
Moreover, we tested the differences between meat eaters and vegans or vegetarians, and between genders, regarding their response to the same health TLI on the bulbs.
Our study results revealed that TLI significantly increased the odds of choosing the carbon bulb, which we presented as the healthier one based on scientific studies. Therefore, it may be concluded that the participants experienced a cognitive bias, in which their decision to choose the healthy option over the EE light bulb was based on the presentation of the carbon bulb as positive (green) compared to the presentation of the LED bulb as negative (red). These results are consistent with those reported in a study that tested whether the effectiveness of loss-framed information would be enhanced by the presence of the color red-which was expected to prime the threat via its association with blood and danger [38]. Their study demonstrated that exposure to red amplified the effects of a loss-framed message promoting a new vaccine. Additionally, our results are in accordance with those of studies regarding TLI on food and beverages, showing that TLI reduced sales of the unhealthy items, and increased sales of the healthy items [39].
The comparison between consumers from different dietary choice groups showed that the TLI strongly affected participants from both dietary choice groups (meat eaters versus vegans/vegetarians) towards choosing the carbon bulb. However, the increase among the vegans/vegetarians towards choosing the healthier light bulb was profoundly higher than among the meat eaters. These results are in line with studies regarding food choice motives within these groups. A recent study that classified dimensions for food choice motives revealed that health was ranked second after taste [40], consistent with the positive effect of health TLI on both groups. Nonetheless, since reducing animal-based food products is considered an environmentally friendly and healthy food behavior [41], it may explain why the vegan/vegetarian participants were more significantly affected by the TLI cues added to the light bulbs than the meat eaters.
Regarding the differences between genders, our results revealed that not only were the odds of females choosing the carbon bulb significantly greater than those of males (p < 0.01), but also that although the TLI influenced both genders robustly, it had a greater influence on females than it did on males towards choosing the carbon bulb. This result corresponds with a study regarding the use of TL food labeling in Ecuador, which revealed that adult females used the TL-label most frequently while selecting products, whereas adult males occasionally used the TL-label [42]. In a study regarding gender differences, females showed more interest in health-related information and in the health effects of the products they bought for themselves and their children than males [34]. Therefore, this can explain why adding health TLI affected females' bulb choice more than the males'.
In our study we also compared participants who stated that it was more important that a light bulb be healthy than that it be EE to those who stated the opposite. The results showed that despite the significant influence of the TLI on participants with both points of view, the impact was much more robust on participants with the "healthy" view than on those with the "EE" view.
The results of the binary logistic regression emphasized the robustness of the TLI effect, since the TLI variable remained significant even after incorporating the "Importance" variable (which represented the "healthy" point of view).
During our field study, there were participants who willingly supplied information regarding their considerations for choosing a light bulb in its original package without TLI. Among them were (1) "I choose the bulb that looks prettier," (2) "I read somewhere that LED is unhealthy, but I am not sure if it is true," and (3) "I prefer LED because it is well known that it is much more EE than other light bulbs, and therefore will reduce my electricity bill profoundly." Both the results concerning participants who differed in terms of their point of view and these remarks shed light on the fact that adding health TLI as an information source helped consumers in their decision-making process, which can explain the significant sweeping shift towards the healthier carbon bulb after adding the TLI. Moreover, the impact of the health TLI on consumers' choice of light bulbs occurred in spite of the energy-efficiency color scale on the original packages of the light bulbs, which highlights the profound effect of the seemingly simple health TLI.
However, it was noted that there is a need for government supervision on eco-labeling schemes in order to ensure that these schemes are implemented properly [43]. Similarly, for the purpose of preventing the "greenwash" phenomena, it is necessary that the TLI system be monitored by the government.
Finally, our study's major limitation that lays the foundation for future research is the choice of light bulbs that were presented to participants. Other light bulbs might bring about a different outcome. In addition, since vegans and vegetarians constitute about 15% of the country's population, according to the Israeli Central Bureau of Statistics, a larger sample including more vegans and vegetarians may also contribute to the research.

Conclusions and Implications
In conclusion, our results revealed that using TLI to alert consumers regarding potential health risks when using LED bulbs, and supplying them with information required to make educated purchasing decisions, can stir consumers towards choosing the healthier bulb option. Moreover, consumers with higher health consciousness such as females, vegans/vegetarians, or those who claimed that it is more important that a light bulb be healthy, are even more affected by TLI. The main advantage of the cognitive bias that consumers experience due to the TLI lies in the ability to nudge consumers towards policymakers' desired light bulb, whereas the downside is that people are, in a way, being "manipulated." Nevertheless, these days, while people around the world are suffering from a raging pandemic due to COVID-19, there are reports regarding new variants of the virus that causes COVID-19, which have been documented in the United Kingdom and are spreading globally. Due to the COVID-19 pandemic, radical changes have been imposed on individuals' daily routines, varying from stay-at-home policies and social distancing to quarantines. As a result, many people are working or learning from their homes, exposed to household lighting, which highlights the significance of this study. Therefore, the study results accentuate the benefits of providing health-related information to consumers in the form of TLI added to light bulb packaging, helping them to make more health-conscientious purchasing decisions.
In addition, the study results may incentivize policymakers to enforce adding healthrelated TLI to light bulb packaging as a means of promoting healthier consumer preferences of light bulbs, which may encourage light bulb manufacturers to reduce the blue light component. Lastly, we recommend that firms that succeed in manufacturing LED light bulbs that are EE, as well as less enriched with the SWL component, use TLI as part of their marketing strategy in order to emphasize the health advantage of their light bulbs. The implementation of health TLI should be accompanied by a media campaign heralding the new TLI scheme.
Supplementary Materials: The following are available online at https://www.mdpi.com/2071-105 0/13/3/1401/s1, Figure S1. The energy-efficiency color scale on the original packages. Figure S2. A presentation of the light bulbs with TLI. Funding: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Institutional Review Board Statement:
The study protocol was approved by the Ethics Committee of the University of Haifa, approval number: 192/18.

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy and ethical reasons.

Conflicts of Interest:
The authors declare no conflict of interest.