The effect of image-induced perceived sleepiness on impulse buying and preferences for the present versus the future

(1)

The Effect of Image-Induced Perceived Sleepiness on

Impulse Buying and Preferences for the Present

versus the Future

Master thesis author:

Anne-Rose Lynch (10179593)

University of Amsterdam

Faculty of Economics and Business

MSc. in Business Administration

Marketing Track

Academic year: 2016-2017

Supervisor: Dr. A. N. Weihrauch

Final Version

Amsterdam, June 23th 2017

(2)

This document is written by Anne-Rose Lynch who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

Abstract

Every person needs a certain amount of sleep to be able to function normally. A lack of sufficient sleep can cause a decrease in cognitive ability (Kerkhof & Van Dongen, 2010). A decrease in cognitive ability could in turn influence decision making and might lead to buying impulsiveness and preferences for present rewards over future rewards. Whereas the actual effect of a lack of sleep has been studied by many scholars, research into the possibilities of inducing feelings of sleepiness and the possible effect this has is quite new. This study aimed to examine the influence of image-induced perceived sleepiness on consumers’ buying impulsiveness and preferences for the present versus the future. Data was collected from 107 respondents via an online experiment. Perceived sleepiness was induced by repeatedly showing respondents images of people sleeping. Buying impulsiveness was assessed through 10 product items presented in both a hedonic (impulsive) and utilitarian (rational) option. Preferences for the present versus the future was assessed by 4 product related (earphones, television series, sunglasses, bag) delay discounting scenarios. Results showed that images of people sleeping resulted in increased wakefulness. Respondents in the control condition, who had a higher level of perceived sleepiness, showed a more impulsive buying tendency when it came to the product headset and a less impulsive buying tendency when it came to the product heater. Furthermore, there was no effect found of perceived sleepiness on preferences for the present versus the future. Implications of these findings are discussed.

(4)

LIST OF TABLES AND FIGURES

Tables

Table 1 Behavioral Buying Impulsiveness Scale 31

Table 2 Sample plan 36

Table 3 T-test results of Perceived Sleepiness Questions by Conditions 43

Table 4 T-test results and descriptive statistics for SSS by Conditions 43

Table 5 Univariate ANCOVA between Conditions and Stanford Sleepiness 44

Scale Table 6 T-test results and descriptive statistics for each Impulsivity Item per 46

Condition Table 7 Univariate ANCOVA between Conditions and Impulsivity (Muesli) 46

Table 8 Univariate ANCOVA between Conditions and Impulsivity (Heater) 47

Table 9 Univariate ANCOVA between Conditions and Impulsivity (Headset) 47

Table 10 Univariate ANOVA between Conditions and Delay Discounting 48 (Earphones)

(6)

Table 11 Univariate ANOVA between Conditions and Delay Discounting 49 (Television series)

Table 12 Univariate ANOVA between Conditions and Delay Discounting (Bag) 49

Table 13 Welch’s F between Conditions and Delay Discounting (Sunglasses) 49

Table 14 Univariate ANCOVA between Conditions and Delay Discounting 51 (Earphones)

Table 15 Univariate ANCOVA between Conditions and Delay Discounting 51 (Television series)

Table 16 Univariate ANCOVA between Conditions and Delay Discounting 51

(Sunglasses)

Table 17 Univariate ANCOVA between Conditions and Delay Discounting 52 (Bag)

Table 18 Repeated measures ANOVA– Differences of Delay Discounting 53

between Scenarios per Condition

Table 19 Mean and standard deviation for repeated measures ANOVA– 53

Differences of Delay Discounting between Scenarios per Condition

Table 20 Repeated measures ANCOVA– Differences of Delay Discounting 54 between Scenarios per Condition

Table 21 Univariate ANOVA between Conditions and Delay Discounting 55

(Earphones - scale)

Table 22 Univariate ANOVA between Conditions and Delay Discounting 55 (Television series - scale)

Table 23 Univariate ANOVA between Conditions and Delay Discounting 56 (Sunglasses - scale)

Table 24 Univariate ANOVA between Conditions and Delay Discounting 56 (Bag - scale)

(7)

Table 25 Univariate ANCOVA between Conditions and Delay Discounting 57 (Earphones - scale)

Table 26 Univariate ANCOVA between Conditions and Delay Discounting 58 (Television series scale)

Table 27 Univariate ANCOVA between Conditions and Delay Discounting 58 (Sunglasses scale)

Table 28 Univariate ANCOVA between Conditions and Delay Discounting 59 (Bag - scale)

Table 29 Repeated measures ANOVA– Differences of Delay Discounting 60 between Scenarios(scale) per Condition

Table 30 Mean and standard deviation for repeated measures ANOVA– 60 Differences of Delay Discounting between Scenarios (scale) per

Condition

Table 31 Repeated measures ANCOVA– Differences of Delay Discounting 61 between Scenarios per Condition

Figures

Figure 1 Conceptual framework 13

Figure 2 Conceptual framework with hypotheses 25

(8)

1. INTRODUCTION

Sleeping is something we all need. According to the National Sleep Foundation young adults and adults need between the 7 and 9 hours of sleep a day (Hirshkowitz et al., 2015). While actual sleep disorder can cause serious problems like deep depression, minor imbalances in sleep can also have its consequences. Sleep imbalances can cause errors in judgement, anxiety, trouble with processing of information, and loss of attention (Kamstra, Kramer, & Levi, 2000). Another consequence of insufficient sleep is impatience. This can be expressed in different ways, one of which is through impulsivity (McLesih & Oxoby, 2007)

When consumers are in a state of impulsivity they can purchase things without a lot of reasoning behind it. Within impulsive buying there is no pre-shopping intention but the purchase is a consequence of an urge to buy something (Beatty & Ferrell, 1998). Insufficient sleep can impact an individual’s vigilance and can alter people’s judgement and decision making (McKenna, Dickinson, Orff, & Drummond, 2007; Kerkhof & Van Dongen, 2010). Insufficient sleep might then indirectly increase the likelihood of an individual showing impulsive shopping behavior.

Moreover, another way of expressing impatience as a consequence of insufficient sleep, is through discounting behavior (McLesih & Oxoby, 2007). Discounting behavior involves a choice an individual has to make between a small reward now or a large reward later. While people would generally pick a larger reward over a smaller reward (Green & Myerson, 2004), the temptations that appear in an individual’s immediate surroundings can often conflict with the requirements of longer rage plans (Hirsh, Morisano, & Peterson, 2008). Delay discounting describes “the extent to which the value of a reward decreases as the delay to obtaining that reward increases” (Hirsh et al., 2008, p. 1646.). During a delay discounting task an individual must integrate large amounts of complex information (Shamosh & Gray, 2008). A lack of

(9)

sleep has a negative impact on information processing and might interfere with this process (Kerkhof & Van Dongen, 2010). As a consequence of this the smaller, sooner reward might be chosen more often than the larger, later reward. If individuals who perceive they are sleepy are more likely to choose the smaller, sooner reward over the larger, later reward, would they then also be more likely to prefer a regular priced product now over the same product with a discount later?

While sleepiness could thus possibly affect buying impulsiveness and preferences for the present versus the future, it is difficult for store managers to actually assess consumers’ real level of sleepiness. This study will therefore look at images of people sleeping, and see whether these can induce sleepiness, and therewith increase consumers’ buying impulsivity and preferences for the present over the future. Research has showed that emotions can be evoked in individuals by showing them images of people with similar emotions. Wild, Erb and Bartels (2001) found in their research that when participants were shown faces expressing sad and happy emotions this elicited the expressed feelings in the participants. Morewedge, Huh and Vosgerau (2010) on the other hand found that habituation can occur. Their research shows that individuals who continually imagined themselves consuming more food were habituated to it and less inclined to attain it. Little is known, however, about the possibilities of eliciting feelings of sleepiness through exposure of sleep images. This is even more relevant, as marketers such as IKEA and Granufink have used images of sleeping people in their advertisement, or like Snickers who has used communications about sleep (“you’re sleepy when you’re hungy”), without truly knowing the effect it might have on consumers.

Furthermore, while insufficient sleep could increase impatient behavior, little research has looked into the effects of image-induced perceived sleepiness. As mentioned earlier, research has shown that actual sleep deprivation can influence decision making and judgement (McKenna et al., 2007; Kerkhof & Van Dongen, 2010). Moreover, the relationship between

(10)

cognitive ability and delay discounting has also been studied very often (Hirsh et al., 2008). Little is known, on the other hand, about what the actual effect of image-induced perceived sleepiness is on decision making and impulsive buying behavior and what effect it has on preferences for the present versus the future.

In this research these gaps will be addressed by looking into the effect of the exposure of images of sleep on perceived sleepiness and whether this will evoke impulsive buying behavior and increase preferences for immediate rewards. Based on the previous, the following research question will be discussed:

To what extent does consumers’ image-induced perceived sleepiness affect impulsive purchase behavior and consumers’ preferences for the present versus the future?

From an academic perspective this research will provide new insights into the possibilities of eliciting feelings of sleepiness through the use of images of people sleeping and whether this can elicit impulsive buying behavior. Furthermore, while it is known that a lack of sleep has a negative impact on information processing and cognitive ability (Kerkhof & Van Dongen, 2010), little is known about the possible effect this has on preferences for the present versus the future. This research will give new insights into the delay discounting model and how individuals react to promotion types that focus on present or future rewards when they are in a state of perceived sleepiness.

Moreover, images of people sleeping and sleep communication is already used within marketing, but the actual effects it has are unknown. The outcome of this research can therefore guide managers in whether to choose sleep images to indirectly elicit impulsive buying behavior and if so whether this is ethically responsible. Sleep images could evoke

(11)

case that sleep images have no or even an opposite effect on perceived sleepiness and don’t elicit impulsive buying behavior. Furthermore, the results from this research can give managers insights into the effect of the use of present rewards versus future rewards as a promotional type on individuals who are in a state of perceived sleepiness.

In order to answer the research question an online experiment will be conducted. By this means the perceived sleepiness of the respondents can be manipulated by showing them images of people sleeping. Additionally, the effect image-induced perceived sleepiness has on buying impulsiveness and preferences for the present versus the future can be measured.

In the following sections the relevant literature on the previous stated topics will be discussed. In the first section the relevant theory with regards to the research question will be discussed and hypotheses will be formulated. In this section it will firstly be explained how sleep and images of people sleeping might be able to influence consumers behavior. Second, an overview will be given from what is known about consumer buying impulsiveness and how image-induced perceived sleepiness is related to this. Finally, there will be elaborated on delay discounting and how image-induced perceived sleepiness can influence this. In the second section the used method of this research will be explained. In the third section the results of the research will be presented. In the fourth and final part the main results and limitations will be discussed and suggestions for future research will be given.

(12)

2. LITERATURE REVIEW

As discussed in the previous section, the following problem statement will be discussed: To what extent does consumers’ image-induced perceived sleepiness affect impulsive purchase behavior, and preferences for the present versus the future? Firstly, the concept ‘perceived sleepiness’ will be discussed. Earlier research has looked into eliciting emotions in people similar to the images or videos they are presented with (Provine, 2005; Wild et al., 2001). Based on this, the following research question can be formulated:

RQ 1: Can images of sleep cause feelings of sleepiness?

Secondly, the concept of ‘impulsive purchase behavior’ will be discussed. What is impulsive purchase behavior and can feelings of sleepiness affect this. The following research question can be formulated.

RQ 2: Can image-induced feelings of sleepiness cause impulsive buying behavior?

Finally, the preferences for present and future rewards will be discussed and how this relates to image-induced perceived sleepiness. According to Kerkhof and Van Dongen (2010) a lack of sleep has a negative impact on information processing and cognitive ability. Is it possible that image-induced perceived sleepiness can change preferences towards the present instead of the future?The following research question can be formulated:

RQ 3: Can image-induced feelings of sleepiness increase preferences for the present over the future?

Based on the previous research questions the following conceptual framework is developed in Figure 1:

(13)

Figure 1 – Conceptual framework

In the following sections each research question will be discussed based on a review of the existing literature. Afterwards, a hypothesis will be formulated for each section.

2.1 HOW CAN SLEEP IMAGES AFFECT SLEEPINESS?

In our daily lives we have contact with a lot of different people; friends, neighbors, colleagues, and even strangers. These different people influence what we say, think, feel and do (Chartrand, Maddux, & Lakin, 2005). One way through which people can be influenced is via the mimicry and synchronization of facial, vocal, and postural expressions (Wild et al., 2001). Different studies have found that people are affected by facial expressions such as happy or sad emotions, anger, and expressions of romantic love (Hatfield, Bensman, Thornton, & Rapson, 2014). In the following section it will be discussed how people can be affected by facial expressions displayed by others and whether it is also possible to affect people’s perceived sleepiness in a similar way.

(14)

2.1.1 IMAGE-INDUCED PERCEIVED SLEEPINESS

As stated earlier, while an individual can feel a particular emotion, individuals can also become ‘infected’ with emotions displayed by others (Wild et al., 2001). Emotional contagion describes a process in which an individual can have the tendency to automatically mimic movements, expressions, and postures with those of another person and consequently feel similar emotions (Hatfield, Cacioppo, & Rapson, 1993). Primitive emotional contagion seems to be a basic part of human interaction. It allows us to understand and share the thoughts and feelings of the people we encounter and it is an important part of empathy. This emotional contagion is relatively automatic, uncontrollable and unintentional (Hatfield et al., 2014). The mimicking of other people’s facial expressions starts from the moment we are born (Field, Woodson, Greenberg, & Cohen, 1982). Previous research has found evidence that people tend to feel emotions consistent with the facial, vocal and postural expressions they mimic (Wild et al., 2001).

Moreover, people can read how other people are feeling and take on the emotional and affective state of another person (Chartrand et al., 2005). Wild et al. (2001) found in their research that when respondents were shown faces expressing sad and happy emotions this evoked the expressed feelings in the respondents. They also found that this emotion can be elicited with a single presentation.

Furthermore, Provine (2005) researched the contagious effect of yawning. He found that respondents yawned more than twice as much when they saw a video of a yawning man as when they saw a video of a smiling man. The effect was still present if the video was shown in black and white or was shown sideways or upside down. Moreover, when the respondents were shown a still image of a man yawning the effect was also present. If behavior and emotions can be evoked in individuals by showing them images of people exhibiting those

(15)

emotions or that behavior it might also be possible to evoke feelings of sleepiness by showing images of people sleeping.

However, according to Morewedge, et al. (2010) perception and mental imagery tend to similarly affect emotions, response tendencies and motor behavior. Thinking of a spider crawling on your leg can produce the same response as when an actual spider would crawl on your leg (Lang, 1977). Morewedge et al. (2010) researched whether thinking of consuming food could then also habituate one to it. Their research shows that individuals who repeatedly imagined themselves consuming more food where indeed habituated to it and less inclined to obtain it. If individuals are repeatedly shown images of people sleeping this could thus also lead to people feeling less sleepy. Moreover, if images of people sleeping could in fact elicit feelings of sleepiness, is it then also possible that these individuals act similar to people who are actually sleepy?

2.1.2 LAY BELIEFS

Potential reactions to image-induced sleepiness perceptions (such as becoming more impulsive, as stated in this thesis) could be driven or enhanced by the lay beliefs people hold about the behavior of sleepy people. Vohs, Glass, Maddox and Markman (2011) describe lay beliefs as ‘ how people think they would respond and how they do respond when faced with an event’(p. 168). Findings of the research conducted by Vohs et al. (2011) give an example of the effect of lay beliefs. They found that people hold lay beliefs that being deprived of sleep will lead to aggression (Vohs et al., 2011). Additionally, according to Morin, Stone, Trinkle, Mercer, and Remsberg (1993), people belief that one can hardly function during the day without a good night’s sleep. Furthermore, research has also found that lay beliefs influence individuals’ food and exercise choices (Burnette, 2010; Crum & Langer, 2007). McFerran and Mukhopadhyay (2013) also found that people’s food choices were affected by

(16)

the lay theories they have about obesity. People who believed that obesity is caused by a lack of exercise tended to consume more food than people who believed that obesity is caused by a poor diet. This shows that the lay beliefs an individual has can affect his or her decision making. If an individual has a lay belief that sleepiness leads to poor decision making, or that sleep leads to accurate decision making, he or she could act upon those beliefs.

If people are presented with images of people sleeping they can thus, on the one hand, feel sleepy through emotional contagion or become less tired through habituation. On the other hand, if people are presented with images of sleep and they believe they are sleepy, they might also act similar to people who are actually sleepy. Based on this, the following hypothesis can be formed.

HI : Exposure to sleep images will elicit feelings of sleepiness in consumers.

2.2 HOW MIGHT SLEEP AFFECT IMPULSIVITY?

When individuals are manipulated into perceiving they are sleepy, they could thus act according to their believes about how sleepy people act. It is very likely that individuals have heard of potential effects of a lack of sleep on decision making, and integrate these into their beliefs on how sleepy people behave. In the following section will be discussed what these actual effects are that sleep, or a lack of it, has on decision-making and how this relates to buying impulsiveness.

2.2.1 THE EFFECTS OF SLEEP ON DECISION MAKING

Sleep studies have looked into different forms of sleep deprivation, both long-term e.g. > 45 of wakefulness, and short-term e.g. < 7 hours of sleep within 24 hours. All forms of sleep

(17)

(Durmer & Dinges, 2005). Kerkhof and Van Dongen (2010) had similar findings and state that an important effect of sleep deprivation is the change it causes in normal mood and emotional functioning. Sleep is also an important factor in sustaining regular performance. Insufficient sleep can quickly impact an individual’s cognitive capacities, the quickest being alertness and vigilance. This while without some degree of alertness and attention, it will become nearly impossible to adequately engage in complex cognitive processing (Kerkhof & Van Dongen, 2010). According to Durmer and Dinges (2005) individuals who perform even brief cognitive tasks that measure memory, attention, and speed of cognitive processing have been found to be influenced even with little sleep deprivation. Additionally, according to Goel, Rao, Durmer and Dinges (2009) individuals will start showing a decrease in reaction time and worsening of performance accuracy on vigilance tests when people are continuously awake for more than about 16 hours. These declines in reaction time and vigilance will get even worse as time passes. Van Dongen, Maislin, Mullington and Dinges (2003) and Drake et al. (2001) also found that individuals’ behavior is effected more with the increase of sleep deprivation.

Furthermore, research conducted by McKenna et al. (2007) looked into the effect of sleep on people’s decision making and judgement. They found that sleep-deprived people are more likely to take risks when an outcome is framed as a potential gain. When the outcome is framed as a potential loss, sleep-deprived people are less likely to take risks than normally (McKenna et al., 2007).

Moreover, from prior research can be concluded that sleep deprivation and even a lack of sleep can lead to a decrease of an individual’s alertness and attention and can also change an individual’s decision making behavior. People might be aware of the above mentioned effects that a lack of sleep has and might think that sleep is important in order to make accurate decisions. If people are then manipulated into feeling sleepy, the knowledge about the

(18)

consequences that a lack of sleep has could indirectly impact their decision making through their lay beliefs.

Additionally, the decisions people make are influenced by impulsive and reflective mechanisms (Strack & Deutsch, 2004). Strack, Deutsch and Werth (2006) have developed a two-system model in which these mechanisms are explained. This model is called the reflective-impulsive model (RIM). In the following section this model and how it relates to image-induced perceived sleepiness are explained.

2.2.2 THE REFLECTIVE-IMPULSIVE MODEL

To explain how sleep might affect the reflective and impulsive mechanisms(RIM), the RIM needs to be explained first. The RIM describes the outcome of the combined reflective and impulsive mechanisms (Strack & Deutsch, 2004). In the impulsive system the information is processed automatically via a quick activation of links between content. These links are related to each other. Within the reflective system information is processed less automatically but based on rules and reasoning which makes the processing flexible but also slower than within the impulsive system.

Furthermore, according to Strack et al. (2006) the impulsive system entails a network of associations. This network is based on activation routes that are consistently seen within the environment and which altogether form a pattern of activation. When stimuli are presented in close temporal or spatial proximity the links within the routes are created or strengthened. When individuals are confronted with for example a product that they need or really want and is thus in close temporal and/or spatial proximity, this might trigger the impulsive system.

The reflective system complements the functions of the impulsive system (Lieberman, 2003). The RIM does not see a certain behavior as being either impulsive or reflective. It

(19)

almost all situations (Strack et al., 2006). The conditions under which a purchase is made define the relative impact of the impulsive and reflective mechanisms and are different for each individual. If the consumer is motivated and if the situation then allows them to thoroughly process the information and tap into their cognitive capacity, the eventual purchase will then most likely be based on reflective mechanisms (Strack et al., 2006). When an individual on the other hand is feeling sleepy, he or she is less capable to engage in cognitive processing (Kerkhof & Van Dongen, 2010). The reflective mechanism can in this case thus be influenced by the impulsive mechanism.

As discussed in the previous subchapter an actual lack of sleep has different effects on decision making. Sleep is an important factor in sustaining regular performance. Insufficient sleep can lead to a decrease in alertness and vigilance which will make complex cognitive processing very difficult (Kerkhof & Van Dongen, 2010). When linking this with the RIM, it can be suggested that sleep leads to less activation of the reflective mechanism. A reduction in alertness, vigilance and reaction time as a consequence of a lack of sleep could then make individuals more prone to using a greater amount of impulsive mechanisms for decision making. This could increase an individual’s impulsive behavior. In the following section the concept of impulsive buying is explained.

2.2.3 IMPULSE BUYING

A consumer’s impulsive decision to buy a product is, in comparison to a rational choice, not based on a well-considered evaluation of that product. Impulse buying can be seen as more irresistible buying behavior and less deliberate. Different researchers have defined impulse buying. According to Beatty and Ferrell (1998) impulse buying can be defined as “a sudden and immediate purchase with no pre-shopping intentions either to buy the specific product category or to fulfill a specific buying task. The behavior occurs after experiencing an urge to

(20)

buy and it tends to be spontaneous and without a lot of reflection (i.e., it is "impulsive"). It does not include the purchase of a simple reminder item, which is an item that is simply out -of-stock at home” (p. 170). Another definition of impulse buying is given by Kacen and Lee (2002): “Impulse buying behavior is a sudden, compelling, hedonically complex purchasing behavior in which the rapidity of the impulse purchase decision process precludes thoughtful, deliberate consideration of all information and choice alternatives" (p. 163) (Bayley & Nancarrow, 1998; Rook, 1987; Thompson, locander, & Pollio, 1990; Weinberg & Gottwald, 1982). Since this research doesn’t focus on whether a product is or isn’t possibly out-of-stock at home the latter definition will be used to define impulse buying within this research paper.

Besides sleep, the likelihood of impulse buying might also be related to a motivational orientation that infers a relationship between valence (attractiveness) and behavioral tendencies of approach and avoidance (Cacioppo, Priester, & Bernston, 1993; Neumann & Strack, 2000). Products that possess a positive valence elicit approaching behavior while products that possess a negative valence elicit avoiding behavior. Within the context of impulse buying positive valence could thus increase the likelihood of an impulsive purchase of the particular product. Mischel and Ebbesen (1970) and Mischel, Ebbesen, and Raskoff-Zeiss (1972) have even found that the actual or imaginal exposure and the proximity of the product strengthen the immediate response. Moreover, Hausman (2000) found that product type affects consumer impulse buying. When considering consuming a product, consumers focus on different information for hedonic products than for utilitarian products (Voss, Spangenberg, & Grohmann, 2003). When it comes to hedonic products, consumers focus on information related to inner sensory stimuli, when it comes to utilitarian products, consumers focus on information related to the decision itself (Okada, 2005). The choices for utilitarian products are more cognitively driven (Chen & Wang, 2016). If perceived sleepiness could

(21)

decrease cognitive ability than consumers might be more likely to choose a hedonic option over a utilitarian option.

As discussed in the previous sections, insufficient sleep can lead to a decrease in alertness and vigilance which will make complex cognitive processing very difficult (Kerkhof & Van Dongen, 2010). When linking this with the RIM, it can be suggested that a lack of sleep leads to less activation of the reflective mechanism. A reduction in alertness and vigilance as a consequence of a lack of sleep could then make individuals more prone to using a greater amount of impulsive mechanisms for decision making. An impulsive purchase is made without engaging in a lot of evaluation. People who buy impulsively are less likely to consider the consequences of them buying the product (Rook, 1987). A lack of sleep could thus increase an individual’s impulsive buying behavior. It is very likely that people are aware of these effects that a lack of sleep has on people and they might think that sleep is important in order to make accurate and non-impulsive decisions. If people are manipulated into feeling sleepy, the knowledge about the consequences that insufficient sleep has could indirectly impact their buying impulsiveness through their lay beliefs. Based on this, the following hypothesis can be formulated:

H2: If people are exposed (not exposed) to sleep images, then they will demonstrate

higher (lower) levels of buying impulsiveness.

2.3 HOW IS SLEEP RELATED TO PREFERENCES FOR THE

PRESENT VERSUS THE FUTURE?

In the previous section is discussed how one way of impatience, buying impulsiveness, could be affected by image-induced perceived sleepiness. When individuals are manipulated into

(22)

perceiving they are sleepy, they might act impulsively with regards to their buying behavior. In the following section another way of impatience through image-induced perceived sleepiness will be discussed. What is the relationship between imaged-induced perceived sleepiness and preferences for the present versus the future?

2.3.1 THE EFFECT OF IMAGE-INDUCED PERCEIVED SLEEPINESS ON PREFERENCES FOR PRESENT VERSUS FUTURE REWARDS

Consumers often have to make decisions that involve a trade-off between costs and benefits over time. Such a trade-off could occur when they have to choose between receiving a small reward now or a large reward later (Green & Myerson, 2004).

This choice between a small and a large reward that is eventually made, differs between individuals (Ainslie, 1975; Kirby, 1997) and is referred to as delay discounting. Delay discounting describes “the extent to which the value of a reward decreases as the delay to obtaining that reward increases” (Hirsh et al., p. 1646). In most research delay-discounting procedures involve a choice an individual has to make between rewards that are smaller but received immediate versus rewards that are larger but received at a later point in time (Reynolds& Schiffbauer, 2004).

Metcalfe and Mischel (1999) describe a mechanism of a hot and cool system that influences delay discounting. The balance between the hot and cool system determines whether an individual’s tendency to delay rewards will increase or decrease. The cool system represents the cognitive system and the hot system represents the emotional system. When the balance between the cool and the hot system is in favor of the cool system the tendency to delay rewards increases (Metcalfe & Mischel, 1999). Since a lack of sleep has a negative impact on cognitive processing (Kerkhof & Van Dongen, 2010), this could shift the balance

(23)

between the cool and the hot system in favor of the hot system and subsequently decreasing the tendency to delay rewards.

Additionally, similar results were found by Shamosh and Gray (2008) who showed that cognitive ability is related to delay discounting. During a delay discounting task an individual must integrate substantial amounts of complex information, e.g. calculating the costs and benefits of choosing one option over the other and the recall of previous choices. He or she needs to do this while also maintaining representations of reward value in his or her memory (Shamosh & Gray, 2008). A lack of sleep might decrease the ability to do those things (Durmer & Dinges,2005; Kerkhof &Van Dongen, 2010) and subsequently decrease the tendency to delay rewards (Shamosh & Gray, 2008).

Moreover, according to Harrison and Horne (2000) sleep deprivation decreases the accuracy of the timing of prior events. Sleep deprived respondents were still able to recognize the stimulus they saw but weren’t able to tell when they saw it. This effect on the timing process might affect the choices people make when having to make a decision between a small reward soon versus a larger reward later (Reynolds & Schiffbauer, 2004). Furthermore, according to Loewenstein (1996) sleep deprivation can decrease the importance of other motives that the decision maker might find important when viewed from a distance before or after a decision. When sleep deprived some rewards could then become overpoweringly attractive as they become closer in time. Research conducted by Reynolds and Schiffbauer (2004) showed that young adults had a higher preference for smaller, sooner rewards over larger, later rewards on a delay discounting task after being awake until 4 am than when they completed the task in the afternoon. However, it has also been shown that sleep deprivation does not affect delay discounting (Libedinsky et al., 2013). While delay discounting wasn’t affected by sleep deprivation, they did find a significant effect of sleep deprivation on effort discounting. Participants that were sleep deprived showed a devaluation of monetary rewards

(24)

in accordance with the prospective effort they need to commit to obtain the reward. This research, on the other hand, does not look at actual sleep deprivation but image-induced perceived sleepiness. If people belief sleepiness will affect their decision making, this could still impact their choices for present rewards over future rewards.

If discounting behavior depends on the relative strength of an individual’s cognitive control networks, it is reasonable to presume that a lack of sleep, which influences one’s cognitive abilities, should influence discounting behavior. If images of people sleeping would elicit feelings of perceived sleepiness in individuals and in turn, based on their lay beliefs, would act as sleepy people would act, image-induced perceived sleepiness might influence discounting behavior. Furthermore, if individuals who perceive they are sleepy are more likely to choose the smaller, sooner reward over the larger, later reward, then they might also be more likely to prefer a product now over the same product with a discount later. The following hypothesis can be formulated:

H3: If people are exposed (not exposed) to sleep images, then they are more (less) likely to

get a product now than the same product with a discount later in time.

The current study examines the possibility of image-induced perceived sleepiness to influence buying impulsiveness and preferences for the present versus the future. Based on the previous literature discussed, it is hypothesized that people who are presented with images of people sleeping will in turn feel sleepy themselves through emotional contagion and their lay beliefs. Furthermore, the behavior that occurs when one is experiencing a lack of sleep will also occur in the people who see the sleep images via their lay believes. Insufficient sleep can lead to a decrease in alertness and vigilance which will make complex cognitive processing very difficult (Kerkhof & Van Dongen, 2010). If people are aware of these

(25)

with the RIM, a reduction in alertness and vigilance as a consequence of a lack of sleep could then make individuals more prone to using a greater amount of impulsive mechanisms for decision making. This could increase an individual’s impulsive buying behavior. Lastly, image-induced perceived sleepiness could also influence delay discounting through the decrease in cognitive abilities. Theoretically, the strength of the preference for an immediate reward should be a stronger predictor of discounting when an individual’s cognitive ability decreases and it is thus more difficult to regulate behavior. The previously formulated hypotheses are added in the initial conceptual framework and presented in Figure 2.

Figure 2 – Conceptual framework with hypotheses

3. METHOD

In this research the effect of image-induced perceived sleepiness on impulsive buying behavior and preferences for the present versus the future is measured. This research is part of a bigger research that consists of another publication in which the effect of image-induced perceived sleepiness on impulsive buying behavior is also measured. That publication on the other hand is not looking at consumers’ preferences for the present versus the future but at the effect of buying impulsiveness on risk behavior.

(26)

3.1 DATA AND MEASURES

The data collection will take place in the Netherlands in April 2017. The data will be collected by distributing an online experiment via convenience sampling. The online survey will be distributed via Facebook, other social media, and email within the social group of the researchers. Through this method of sampling it is possible to collect data in a quick, convenient and economical way. Moreover, the focus was to mainly collect responses from students since they are within the environment of the researchers and the stimuli was focused on this target group. Since these channels are often visited by students this will increase the chances of collecting data from the target group. A disadvantage of this type of sampling is that it might not be as generalizable to other populations since the choice of respondents is not completely random.

Furthermore, there are several advantages to an online experiment versus a traditional lab experiment. An online experiment is less costly and time consuming since procedures can be automated (Reips, 2002). Furthermore, an online experiment can be accessed 24h a day and has less constraints in the setting than lab experiments (Reip, 2002; 2000b). Lastly, an online survey enables randomization of the order of the questions. This is relevant, as the online experimental study will also include another scale developed by another student on the effect of buying impulsivity on risk behavior (Evans & Mathur, 2005). This way it is possible to avoid order effects.

The online experiment is a one-factor between subjects design in which the independent variable has 2 levels (sleep images versus object images). To ensure that the respondents in the sleep images condition are successfully manipulated, a pretest is conducted. In the pretest respondents (N = 25) were asked to look at a video that was presented to them. After they watched the video they had to fill in the Stanford Sleepiness Scale (SSS). The SSS is

(27)

developed by Hoddes, Zarcone, Smythe, Philips, and Dement (1973) and is used to measure the sleepiness level after the sleep images are shown to the participants. The SSS is provided in Appendix 2. The validity of the scale is (α = .68), while the test-retest reliability of the scale is (α = .88).

Each respondent was randomly assigned to either the condition that showed the video with the sleep images (Appendix 1- Figure 1) or the condition that showed the video with the images of random objects (Appendix 1- Figure 2). According to the pretest there was no statistically significant difference in perceived sleepiness between the respondents who saw the sleep images (M = 4.14, SD = 1.61) and those who saw the images of random objects (M = 4.09, SD = 0.94), t(23) = -0.10, p = .921. This might be the case because the respondents where not asked to think more about what they saw. The result might also not be significant because most respondents filled the survey in at night between 7:00 p.m. and 9:00 p.m. which is the time when people are most active (Hopfe et al., 2001). To increase the possibility of a successful manipulation, in the actual experiment participants will be asked to briefly write down what they saw in the video. Writing down what respondents saw enables a process that triggers further thinking about sleep (Menary, 2007). Furthermore, a manipulation check will be introduced to see whether the manipulation was successful.

3.1.1 INDEPENDENT VARIABLE

In this study the independent variable is image-induced perceived sleepiness. In the literature perceived sleepiness has been measured objectively, however to the best knowledge of the author it has not been manipulated by showing images of people sleeping. The hypothesis that people can perceive being sleepy themselves is based on previous research on mimicry and emotional contagion. Emotional contagion describes a process in which an individual can have the tendency to automatically mimic movements, expressions, and postures, with those

(28)

of another person and consequently feel similar emotions (Hatfield et al., 1993). When an individual is exposed to images of people displaying a specific emotion, this can evoke similar emotions (Wild et al., 2001). Furthermore, Provine (2005) found that when respondents were shown an image of a man yawing, they yawned twice as much as when they saw an image of a man smiling. If emotions can be evoked in individuals by showing them images, it might also be possible to evoke feelings of sleepiness by showing respondents images of people sleeping. Morewedge et al. (2010) on the other hand found that individuals who repeatedly imagined themselves consuming more food were habituated to it and less inclined to obtain it. Images of people sleeping could thus also have an opposite effect.

Previous research used different methods to test their hypotheses. Wild et al. (2001) showed respondents images of people who were expressing either happy or sad emotions. They varied the expressive strength and the duration of the presentation. Provine (2005) exposed respondents to a five-minute video of a man yawning 30 times. He found similar results when the film was right side up, sideways or upside down. The findings also didn’t change when the video was in black and white or color, or when the animated stimulus was presented as a still image of the yawner in mid-yawn. Lastly, Morewedge et al. (2010) asked respondents to either imagine inserting 30 quarters into a laundry machine and then imagined eating 3 M&M’s, or imagine inserting 3 quarters into a laundry machine and then imagined eating 30 M&M’s.

To manipulate perceived sleepiness a combination of images and repetition will be used. Respondents will be shown 10 images of individuals who are sleeping. Each image will be shown for approximately 7 seconds with intervals of 3 seconds. The number of images shown and the amount in seconds for the intervals is chosen to reduce respondent fatigue during and for the subsequent tasks, as well as to ensure completion and response rates. Individuals (N =

(29)

10) were asked about the ideal amount of sleep pictures. After 10 pictures they expressed that they were losing focus and their thoughts were drifting elsewhere e.g. what to cook for dinner.

Furthermore, according to Hess and Fischer (2013) for mimicry to occur and thus emotional contagion, it is important that the individuals presented in the stimuli are similar to the respondents. In order to make respondents able to identify themselves with the images presented, individuals of different ages and ethnic origins are included in the stimuli. The control group will be shown 10 random images of objects or miscellaneous items with similar time periods and intervals as the sleep images. These random images will be picked with an online random object generator. It is made sure that the objects are in no way associated with sleep, comfort, or tiredness e.g. no socks, no pillows etc. Examples of objects that are neutral include: sharpie, tape, and table. The images for the manipulation group and the control group are provided in Appendix 1.

3.1.2 DEPENDENT VARIABLES

The two dependent variables that will be measured in this research are buying impulsiveness and preferences for the present versus the future. Impulse buying is defined by Kacen and Lee (2002) as “a sudden, compelling, hedonically complex purchasing behavior in which the rapidity of the impulse purchase decision process precludes thoughtful, deliberate consideration of all information and choice alternatives” (Bayley & Nancarrow, 1998; Rook, 1987; Thompson et al., 1990; Weinberg & Gottwald, 1982). This definition is also used in this research. Previous research has used different methods to measure behavioral impulsiveness. These methods include the delay discounting task (Richards, Zhang, Mitchell, & Wit, 1999), go/no task (Newman, Widom, & Nathan, 1985), stop signal task (Logan, Schachar, & Tannock, 1997), and the balloon analogue risk task (BART, Lejuez et al., 2002). These tasks mainly measure impulsive choice.

(30)

While previous research has used different types of scales to measure impulsive behavior. There is no scale that combines the measurement of impulsive buying behavior in combination with product-related choices. Therefore, a scale is developed that contains 10 products based on an assortment of household (candle, heater, shower head, light-bulb), hygienic (shower gel, detergent), electronic (headset, laptop, phone case), and food ( muesli bar) product categories. The products laptop and headset are taken from a study conducted by Lu, Liu, and Fang (2016). Each product is presented in a hedonic and in a utilitarian variant. Hedonic goods are products that provide more experiential consumption, fun, pleasure, and excitement, whereas utilitarian products are mostly instrumental and functional (Dhar & Wertenbroch, 2000). According to Dhar and Wertenbroch (2000) you can distinguish between hedonic and utilitarian products by examining whether the preference for an item is based on “should” or “want”. Based on the previous characteristics of hedonic and utilitarian goods it is assumed that the hedonic product option is an impulsive choice and the utilitarian product option is a rational choice. The participants are asked on a scale from 1 to 7 (1 = very unlikely, 2 = unlikely, 3 = somewhat unlikely, 4 = undecided, 5 = somewhat likely, 6 = likely, 7 = very likely) “How likely are you to purchase one product over the other”. In the survey the participants are instructed to pick one of the most left three boxes if they are likely to choose product 1, and one of the most right three boxes if they are likely to pick product 2. If they had similar preferences for product 1 and 2, they could pick the middle box. The scale is provided in Table 1 below.

(31)

Table 1

Behavioral Buying Impulsiveness Scale

Product 1 Very unlikely Unlikely Somewhat Unlikely Undecided Somewhat likely Likely Very likely Product 2 Chemical-free detergent 1 2 3 4 5 6 7 Detergent that has fragrance Water-saving shower head 1 2 3 4 5 6 7 Shower head with LED lights Anti-mosquito

candle 1 2 3 4 5 6 7 Scented candle

Muesli bar

with protein 1 2 3 4 5 6 7

Muesli bar with golden syrup Energy saving heater 1 2 3 4 5 6 7 Heater that looks like a fireplace Headset with high battery life but low

music indulgence

1 2 3 4 5 6 7

Headset with low battery life but high music indulgence

Anti-allergenic/no additives shower gel

1 2 3 4 5 6 7 Shower gel with

a great smell Phone case

with battery but no design

1 2 3 4 5 6 7

Phone case with graphic design

but no battery Laptop with

large hard drive space but

traditional design

1 2 3 4 5 6 7

Laptop with a sleek design but

smaller hard drive space Energy efficient light-bulb 1 2 3 4 5 6 7 Color-changing light bulb

The second dependent variable is preferences for the present versus the future. The hypothesis that people who perceive they are sleepy are more likely to prefer immediate rewards versus future rewards is based on delay discounting. Delay discounting describes “the extent to which the value of a reward decreases as the delay to obtaining that reward

(32)

increases” (Hirsh et al., 2008, p. 1686). No research has yet looked into the effect of image-induced perceived sleepiness on preferences for the present versus the future. Several studies have used similar tasks to measure this construct. Green, Fristoe and Myerson (1994) tested different statements in which they offered participants choices between for example $20 in 1 month or $50 in 1 year and a month and also whether they preferred $20 in 1 year or $50 in 2 years. With equal delays between the 2 options they found that preferences switched. As the delay to the smaller reward increased, the number of participants choosing the larger, later reward increased.

Furthermore, similar questions were used in a study conducted by Kirby, Petry and Bickel (1999). They developed a 27-item monetary-choice questionnaire from which a participant’s discount-rate parameter can be estimated. This questionnaire consists of statements in which different choices between present and future rewards are presented. They offered participants choices between relatively small monetary rewards ($11 - $80) available immediately versus larger monetary rewards ($25-$85) available later. The delays of the rewards differed between 1 week and 6 months. This questionnaire is used by many other researchers (Duckworth & Seligman, 2005; Hirsh et al., 2008; MacKillop et al., 2011). While the monetary-choice questionnaire looks at differences between smaller, sooner rewards versus larger, later rewards, there is no questionnaire that uses similar statements in a product-related context. Therefore a scale is developed which contains 4 products: earphones, sunglasses, television series, and bag. Participants are presented with 4 scenarios in which they are currently shopping in a store. They are told that the presented product in the store is something they are looking for. They are then presented with 2 questions. In the first question they can choose between buying the product immediately for the original price or buying it at a discount but receive it later. In the second question they are asked to indicate their

(33)

preference for one over the other on a 7-point Likert scale (1 = very unlikely, 7 = very likely). The different scenarios are provided in Appendix 3.

The prices and delays presented are based on the statements used in the monetary-choice questionnaire from Kirby et al. (1999). The immediate choice presents the preference for the present while the later, discounted product presents the preference for the future. While the monetary-choice questionnaire consists of 27 items, only 4 scenarios are used in this research. The length of the total survey is taken into consideration when deciding upon the number of scenarios to use. This research is part of a larger research and the scenarios on their own are longer than the original questions from the monetary-choice questionnaire. The respondents need to stay involved when reading the scenarios and when making their decisions.

Moreover, the products used in these scenarios are chosen because they are considered relevant for the student sample used in this research, making it more able for respondents to imagine themselves actually being in the situations presented by the different scenarios. Furthermore, the prices of these products could range from a low price to a high price making it possible to match them with different price range questions used in the monetary-choice questionnaire. Lastly, the price for each product is high enough to evoke involvement in participants. When picking the price ranges and the corresponding delays for the 4 products it was taken into consideration what the actual prices of the products are and what a reasonable time delay could be. For instance, the monetary-choice questionnaire also includes time delays of over a hundred days. It is very unlikely that consumers would be willing to wait a hundred days or more before they could receive one of these 4 products in exchange for a discount on those product.

(34)

3.1.3 CONTROL VARIABLES

In this study several control variables are being taken into consideration. The variables that are taken into account are: gender, age, nationality, time at which the survey is taken, and participants’ individual buying impulsiveness trait.

Gender: According to Dittmar, Beattie, and Friese (1995) impulsive items are affected by

social categories such as gender. Women value their possessions for emotional and relationship-oriented reasons, while men value their possessions for functional and instrumental reasons.

Age: Additionally, Wood (1998) conducted research on impulse buying. He found that

there is an inverse relationship between age and impulse buying. Between the ages of 18 and 39 there is a slight increase in impulse buying. After the age of 39 impulse buying declines again. Previous research suggests that the older consumers get, the more they learn how to control their impulsive buying tendencies (Kacen & Lee, 2002). This is because older individuals are able to better control their emotional expressions than younger adults (Lawton, Kleban, Rajogopal, & Dean, 1992)

Nationality: Furthermore, nationality has been found to impact impulsive buying

behavior (Kacen & Lee, 2002). For example, Western individuals focus on the self, and their individual needs and desires. They feel the tendency to buy impulsively to satisfy their hedonistic pleasures. Eastern individuals however, are more collectivists. They focus on interdependence and emotional control and moderation. They seem to be discouraged to buy impulsively because of the focus they put on the group needs and desires instead of their own (Kacen & Lee, 2002). Impulse buying is largely universal, but local market conditions and cultural forces influence how people act upon their impulses (Rook, 1987). The social acceptability of impulse buying could for example differ between countries (Rook & Fischer,

(35)

1995). Since there might be a difference between individualistic and collectivist countries or cultures it is important to control for this variable. This is to make sure that nationality does not become a confounding variable and affect the dependent variables.

Time of day: Moreover, the time of day at which the survey is filled in is taken into

account. According to Hopfe et al. (2001) people are most active in the morning after getting up around 7:00 a.m. and in the evening at 7:00 p.m. and 9:00 p.m. Differences in results in perceived sleepiness might occur between people who fill the survey in early in the morning or during the day.

Buying impulsiveness trait: Lastly, in this research the buying impulsiveness trait will be

controlled for. Since consumers who have a high impulsivity trait have a higher impulsive buying intention than those who have a low impulsivity trait (Dholakia, 2000), buying impulsiveness trait needs to be considered as a covariate. This will be measured with the Buying Impulsiveness Scale (BIS; Rook & Fischer, 1995). This scale is chosen since it assesses impulsiveness as a personality trait in relation to purchasing. The scale is provided in Appendix 4. The scale is highly reliable (α=.88), indicating high internal consistency and correlation between the items in the scale.

3.1.4 QUESTIONNAIRE

The online survey was distributed via Qualtrics and had a specific order which is as follows: firstly the respondents have to indicate on a 7-point Likert scale whether they agree with the statement that they feel a little sleepy right now and if they would like to go for a nap/get some sleep at the moment. After this the respondents will be randomly assigned to either the experimental condition (video with sleep images) or the control condition (video with images of objects). After they watched the video they are asked to please describe what they saw in the video in 50 to 75 characters. The respondents will then be asked to again indicate whether

(36)

they feel a little sleepy right now and if they would like to go for a nap/get some sleep at the moment. The questions are presented on a 7-point Likert scale. After this they will be presented with the Stanford Sleepiness Scale (SSS). Next, the respondents are asked to fill in the impulsiveness scale (Table 1) developed for this research. After completing the scale the respondents will be randomly assigned to either the risk– delay discounting scales or the delay discounting – risk scales (Appendix 5). The order in which the scales are presented is to ensure that the effects from one task do not carry over to the other task. Lastly, the respondents will be asked to fill in the Buying Impulsiveness Scale (BIS; Rook & Fischer, 1995) and give their demographic information, such as age, gender, and nationality. They will also be asked what time it is. The complete survey is provided in Appendix 6. The sample plan is provided in Table 2 below.

Table 2

Sample plan

Power

Manipulation IV: sleep images

Experimental group Control group

- Sleep images

- DV: Behavioral buying impulsiveness scale + perceived risk scale or delay discounting scale (randomized) - Demographics & controls

- Random images of objects

- DV: Behavioral buying impulsiveness scale + perceived risk scale or delay discounting scale (randomized) - Demographics & controls

(37)

3.2 ANALYSIS

To analyze the data a t-test, ANOVA/ANCOVA analysis and a repeated measures ANOVA/ANCOVA analysis will be used. These methods are chosen because this research investigates a causal relationship between the independent variable of image-induced perceived sleepiness to the dependent variable of buying impulsiveness and the effect of image-induced perceived sleepiness on the preferences for the present versus the future.

4. RESULTS

This research looks at the following research question: to what extent does consumers’ image-induced perceived sleepiness affect impulsive purchase behavior, and consumers’ preferences for the present versus the future? In the following section the results of the online experiment with regards to the main research question are discussed.

4.1 DESCRIPTIVE AND FREQUENCIES STATISTICS

As mentioned, the respondents were randomly assigned to either one of the two conditions. In total 112 respondents participated in this research. The response rate was 95.5% because 5 respondents had to be excluded from the research. 1 respondent was excluded because the duration of the survey of this person was much longer than others. The other 4 respondents were excluded because they saw both the video of the objects and the sleep images when

(38)

Qualtrics wasn’t able to load the video correctly. 49 respondents were exposed to the object images and 58 respondents were exposed to the sleep images.

The manipulation of image-induced perceived sleepiness was measured through two sets of sleep questions (before and after manipulation) and the Stanford Sleepiness Scale (SSS). The both sets of sleep questions and the SSS were normally distributed. SSS had a skewness of 0.60 (SE = 0.23) and kurtosis of 0.83 (SE = 0.46). The set of sleep questions before manipulation had a skewness of 0.02 (SE = .23) and a kurtosis of -0.93 (SE = 0.46), the set of sleep questions after manipulation had a skewness of 0.10 (SE = 0.23) and a kurtosis of -1.04 (SE = 0.46). Both the sleep (M = 3.12, SD = 1.17) and the control condition (M = 3.53,

SD = 1.08) had an average score ranging around 3 and 4 which corresponds to the “Awake,

but relaxed; responsive but not fully alert”(3) and the “somewhat foggy, let down”(4) items in the scale. The scores on the two sets of sleep questions showed minor differences for both the sleep condition (before: M = 3.90, SD = 1.58; after: M = 3.92, SD = 1.54) and the control condition (before: M = 4.00, SD = 1.79; after: M = 4.09, SD = 1.82).

Furthermore, the first dependent variable is buying impulsiveness and measured by 10 different items which were presented in there hedonic and utilitarian options. Of all 10 products, respondents in both conditions chose on average the light bulb as most utilitarian (sleep: M = 2.52, SD = 1.76; control: M = 2.84, SD = 2.06). The product with the highest hedonic score on average was for the sleep condition the candle (M = 4.40, SD = 1.70) and for the control condition the headset (M = 4.78, SD = 1.77).

Moreover, the second dependent variable was preferences for the present versus the future and was measured with 2 different questions. Of all 4 scenarios, both conditions had on average the highest preference for the future option when it came to the the item bag (sleep: M = 0.78, SD = 0.42; control: M = 0.76, SD = 0.43). Similar results were found on the second

(39)

question regarding the preferences indicated on a scale (sleep: M = 5.02, SD = 1.90; control:

M = 5.20, SD = 1.81).

Additionally, the age of the respondents differed between the 20 and 65 years old (M = 27.2, SD = 8.36). More than half of the respondents were between 22 and 25 years old. 33 male respondents participated in this study and 74 female participants (M = 0.69, SD = 0.46). The majority of the respondents was Western (N = 95; Eastern N = 12). While the times at which respondents filled in the survey ranged between 04:00 o’clock in the morning till 24:00 o’clock at night, most respondents filled in the survey between 12:00 and 14:00 o’clock (N = 50). Finally, the buying impulsiveness trait scale had a mean of 3.61 (SD = 1.06), the respondents were not highly impulsive or not impulsive at all.

4.1.1 CORRELATION MATRIX

An overview of the descriptive statistics, correlations, and scale reliabilities is presented in Appendix 7. First of all there are significant correlations between the 10 different impulsivity items, however this is not the case between all the impulsivity items. The items shower head (r = 0.20, p <.05), candle (r = 0.22, p <.05), muesli (r = 0.24, p <.05), heater (r = 0.32, p <.01), headset (r = 0.23, p <.05), shower gel (r = 0.50, p <.01), and light bulb (r = 0.25, p <.01) correlate significantly with detergent. The item heater also correlates significantly with showerhead (r = 0.28, p <.01), candle (r = 0.30, p <.01), and muesli (r = 0.35, p <.01), shower gel (r = 0.25, p <.01)and light bulb (r = 0.37, p <.01). All other significant correlations between the impulsivity items are medium to low. This indicates that some items are directly related to each other but not all of them are related to each other. The item heater and detergent do correlate with a lot of other items. This suggests that not all items are measuring the same construct. A reliability analysis will later on give more insight in the relationships.

The effect of image-induced perceived sleepiness on impulse buying and preferences for the present versus the future