Priming the daredevils : the role of personality traits in the effect of visual subliminal priming on product preference

(1)

PRIMING THE DAREDEVILS:

The role of personality traits in the effect of visual

subliminal priming on product preference

Master’s Thesis

Msc. in Business Administration – Marketing Track

Amsterdam Business School

Student

: Ravenska Atwinda Difa

Student ID

: 10872078

Supervisor

: Tina Dudenhöffer, Msc.

Second reader

: dhr. dr. Alfred Zerres

Date of submission: 24 June 2016

Version

: Final draft

(2)

2

Statement of Originality

This document is written by Student Ravenska Difa who declares to take full responsibility for the content 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 content.

(3)

3

Abstract

This research explores the effect of subliminal priming towards individuals’ product preference, and whether this effect varies based on the respondents’ personality traits. The basic notion of the research question lies on the theories of subliminal priming and personality traits. Past studies in subliminal priming showed that when an individual is exposed to a subliminal prime prior to making a decision, they will have a tendency to choose the primed offering. More recent studies further explored this effect and suggested that individuals who scored high in certain personality traits are more sensitive to subliminal priming which are in line with their personal characteristics.

This study measures one specific personality trait namely ‘sensation seeking’ and assesses whether individuals who scores high in sensation seeking are more reactive towards sensation seeking relevant primes (i.e. individuals primed with words such as “Adventure” or “Seek Adventure”) when exposed to one prior to making leisure activities choices (i.e. ‘Hike-and-Track’, ‘Weekend in a Country Side’, ‘Skydiving’, and ‘Spa Treatment’). Results showed that on average people who were previously primed have higher tendencies to choose high sensation seeking actvities (i.e. Skydiving) compared to those who were not primed. In addition, although no moderating effect was found, the study revealed that there were differences in the priming effects between high sensation seeking individuals and low sensation seeking individuals. The study adds additional finding which found that the use of imperative sentences in priming resulted in lower priming effect compared to the use of static words across all participants.

With the above findings, the present study revalidates the importance of taking individual personal characteristics and the means of priming content into account in subliminal priming.

(4)

4

Acknowledgements

This master’s thesis would not have been succesfully completed without the support, trust and encouragement from my family, friends and colleagues. Without them, the thesis process would not go as smooth and inspiring as it has been. Thank you for the endless reminder that reassure me to always deliver the maximum result.

I would specifically thank my supervisor Tina Dudenhöffer for the mentoring, constant constructive feedback, and in-depth discussions which I greatly had benefited from. I truly owe the success of this thesis to her for giving the much needed supervision any students can ever have.

(5)

5

scored high in sensation seeking……….25 Figure 7. Expected moderating effect of sensation seeking level on subliminal priming effect…36 Figure 8. Two possible independent effects of two predictor variables……….36 Figure 9. Results of the present study – independent effects of individuals’ level of sensation

seeking and priming content on leisure activities choices………42

List of Abbreviations

BSSS…… Brief Sensation Seeking SS………..Sensation Seeking SSS………Sensation Seeking Scale

(8)

8

Introduction

Have you ever craved for something out of the blue or went to the supermarket and ended up buying things that were not in your list? Or, do you sometimes feel like buying something out of your ‘gut feeling’? How if those instances were not caused by an ‘instict’ nor coincidence, but instead, a set up? At one point in life, people must have heard the controversy of hidden messages placed in the media that eventually affects peoples behavior. As crazy and unethical it may sound, several occasions confirmed the application of such persuasion attempt. For example, several well known brands such as Coca Cola, McDonalds, and Disney were accused of putting hidden sexual messages in their marketing campaigns to attract more consumers. Additionally, Fedex’ logo showed a shape of an arrow in between the letter ‘E’and ‘X’ which subliminally communicates the company’s fast and efficient service (Butler, 2014).

Subliminal priming – also referred to as subliminal persuasion is defined as the exposure of a stimulus or an association of words or objects prior to carrying out an action below the threshold of perception (Hawkins, 1970). The concept became particularly well-known when James Vicary, an American researcher, claimed to have increased the sales of Coke and Popcorn by 57 percent in 1957 by secretly flashing words of persuasion on the screen of a movie theatre in front of thousands of people (Karremans, Stroebe, & Claus, 2006). The words ‘Eat Popcorn’ and ‘Drink Coca Cola’ were played in a single frame for 1/3000 of a second – which was supposedly too short to be caught by naked eye, but long enough to be picked up by the audience’s subconscious, and consequently persuading audiences to buy Popcorn and Coca Cola (Moore, 1982). Although the study of Vicary was later known as a hoax as many researchers failed to revalidate the same results (Strahan, Spencer, Zanna, 2002), starting early 2000 academia was finally able to find empirical evidence which showed that subliminal priming can be effective under certain conditions.

(9)

9 Henceforth, the interest around the application of subliminal priming – especially in a marketing context, continued to emerge.

In marketing, subliminal priming refers to “a technique of exposing consumers to product pictures, brand names, or other marketing stimuli without the consumers having conscious awareness” (Trappey, 1996, p. 571). Strahan et al. (2002) demonstrated that subliminal priming might be effective in influencing consumption behaviour and enhancing persuasiveness of an ad as long as the prime, e.g. text, picture, or sound, could help in pursuing a goal (also referred to as ‘goal-relevant’) and there is a preliminary motive of the recipient of the prime to pursue the goal. In Strahan et al.’s study (2002), it was found that participants who were primed with the word ‘thirst’ drink more and were more persuaded by an ad for thirst-quenching beverage when they were thirsty. Karremans et al. (2006) elevated the theory to the next level by showing that subliminal priming not only influences behaviour and ad persuasiveness, but also affects consumers’ brand choices. In his study, participants who were thirsty showed an increase in their intentions to drink ‘Lipton Ice Tea’ after being subliminally primed with the brand name.

The above past studies on subliminal priming focused primarily on conditions where subliminal advertising has proven to be affective (Bermeitinger et al., 2009; Karremans et al., 2006; Veltkamp, Custers & Aarts, 2011). A limited amount of studies looked at the respondents and asked, ‘For whom does subliminal advertising work (not work)?’ (Bustin, Jones, Hansenne & Quoidbach, 2015; Yoshino, Kimura, Yoshida, Takahashi & Nomura, 2005). If situational factors are believed to influence the effectiveness of subliminal priming, one might argue that this efficacy will also be influenced by individual differences. One study that captures this concept is by Bustin et al. (2015) which demonstrates that individual personality traits (also known as dispositions or characters) moderate the effect of subliminal priming in general. In the study, Bustin et al. (2015)

(10)

10 argued that individuals’ differences in the level of personality traits such as extraversion (social and outgoing), novelty (being different) or sensation seeking, affect one’s goals and needs, which subsequently are linked to consumer choice and different sensitivity towards advertisement. The study presented that people who scored high in certain personality traits are more sensitive to subliminal priming which is in line with the characteristic of such personality traits. It was shown that participants who ranked high in sensation seeking showed a higher intention to drink “Red Bull” – a drink perceived to be resonant to the risk-taker characteristic of sensation seekers, after being subliminally primed with the brand name. On the contrary, there was no effect of subliminal priming found on participants who scored low in sensation seeking. The result demonstrated that there is a significant moderating effect of personality traits with regards to people’s sensitivity to subliminal priming.

Building onto the concept, this study aims to revalidate the theory that personality traits moderate the effect of subliminal priming by testing another potential outcome variable namely product preference. In addition, the present study tested a different situational setting by looking into the upper level of Maslow’s hierarchy of needs which is individuals’ self-actualization e.g. recreational needs, needs for new experiences or leisure activities (Reisinger, 2009). Moreover, instead of using a brand name, participants were subliminally primed with the word associated with sensation seeking. Once primed, participants were asked to choose between different offerings of leisure activities with different perceived sensation levels. Based on the previous research, it is expected that people who scored high in sensation seeking will be more likely to be affected by sensation seeking associated primes than people who scored low in sensation seeking. Past studies have also looked into the use of static word vs. imperative sentence in persuasion tactics, and how each brings positive and reversed effects based on perceived sources

(11)

11 of persuasion (Laran, Dalton & Andrade, 2011). The study by Laran et al. (2011) found that the usage of imperative sentences in persuasion resulted in consumers’ resistance towards the persuasion attempt. As a consequence, consumers acted in contrary manner. A similar concept was also applied in subliminal priming e.g. the use of a static word ‘Cola’ vs. imperative word ‘Drink cola’ in the study by Hawkins (1970) – although the study did not result in significant different results. For these reasons, another potential predictor variable namely ‘primer content’ (i.e. the type of words used in the prime) was added to the present research, dividing the priming type into two: static word vs. imperative word. As per study by Laran et al. (2011), it is expected that the use of static words in priming will result in positive priming effects while the use of imperative word will result in reversed priming effects.

In this study, an online survey experiment containing certain instructions and a set of questions was conducted. Participants were asked to imagine as if they have won a leisure package and needed to choose one out of four different leisure packages. These leisure packages consists of two high sensation seeking activities and two lower sensation seeking activities. Before participants were able to make their choice, a ‘commercial’ video was shown to the participants – which ultimately contained (or did not contain) sensation seeking primes. There are three treatments in the experiment i.e. video which does not include prime, video with imperative sentence prime ‘Seek Adventure’, or video with static word prime ‘Adventure’. Once participants have been (or not been) primed, participants stated their choice by clicking one of the leisure activities. Based on previous research surrounding subliminal priming and the moderating effect of personality traits, it is expected that participants who are primed with sensation seeking concept (both ‘Seek Adventure’ and ‘Adventure’) are more likely to choose sensation seeking leisure activities. This effect is also expected to be different among participants who scored high versus

(12)

12 low in sensation seeking level. Finally, it is theorized that the use of static word in a prime will result in a more positive effect compared to the use of imperative sentence due to the reversed priming effect.

Prior to the results and discussion sections, an overview and the basic concept of subliminal priming, personality traits, sensation seeking and priming content are presented. Furthermore, as a means of contribution, the limitation, implication and suggestions for future research are discussed at the end of the study.

(13)

13

1 Literature Review

1.1 Subliminal priming

1.1.1 History

Schvaneveldt and Meyer (1973) described priming as an effect in memory whereby the exposure to one stimulus influences the response to another stimulus. Priming originated from the word ‘prime’ which is defined in Oxford dictionary as “make something ready for a use or action, or make someone ready for a situation” typically by supplying them with relevant information (Oxford Dictionaries, 2016). By nature, priming can be done consciously or unconsciously – depending on the way the prime is presented. Hawkins (1970) defines the unconscious processing of a prime as the registration of a stimulus below the threshold of conscious perception, also known as subliminal.

The concept of subliminal priming originated from the field of psychology, often explained side by side with supraliminal priming.Supraliminal priming is a concept where people are aware

of the presence of the cue, but not on how it can influence them. In subliminal priming, people are not even aware of the presence of the cue itself (Strahan et al., 2002). This phenomenon where the consumer is exposed to and acts upon certain information without acknowledging the source is known as source amnesia (Schacter, Harbluk, & McLachlan, 1984). Although the two types of priming are different in nature, the influence effect of the prime is the same. The determinant of the effectiveness of the prime is whether people realize the influence power of the prime and whether they have the ability to control the influence (Strahan et al., 2002).

Although the concept of subliminal priming stated to be known since the late 19th Century, the focus on the commercial application of subliminal priming only started to emerge following Vicary’s study in 1957 (Broyles, 2006). Since then, academic and commercial papers surrounding

(14)

14 subliminal priming and subliminal advertising started to develop – not only in the area of marketing, but also psychology, social politics, and, even, military (Bermeitinger et al., 2009; Karremans et al., 2006; Veltkamp et al., 2011).

The application of subliminal advertising invites numerous reactions from consumers around the world, though mainly negative, as it is considered to be unethical. Consequently, subliminal advertising has been legally banned in several countries including the United States, the United Kingdom, and Australia (Karremans et al., 2006).

1.1.2 Types of subliminal priming

Different studies described different types of subliminal stimulation techniques. In the study by Moore (1982), the author described three identifiable means of subliminal stimulation:

1. Brief presentation of visual stimuli, also known as tachistoscopic – projecting or flashing a (series of) brief image(s) into a screen at rapid speed to test visual perception (Kelly, 1979; Moore, 1982). The objective of this technique is to flash a visual stimulus so quickly that it is unnoticeable, nevertheless, registered subconsciously and consequently affect subsequent behaviour.

2. Sub-audible messages – also known as auditory masking which involves “the use of accelerated speech in low volume auditory messages” (Moore, 1982, p.39). This technique of audio stimuli is perceived to be weak and lacking empirical evidence. 3. Subliminal seduction – embedded sexual imagery or words in pictorial advertisements,

which “supposedly influences consumers to buy and find more satisfaction with the advertised brands” (Gable, Wilkens, Harris & Feinberg, 1987, p. 27). Although many companies were allegedly using sexual subliminal stimuli in their advertisements, including Heineken, Skittles, Burger King and Coca Cola (as per Annex 3), the

(15)

15 empirical study by Gable et al. (1987) established that subliminally embedded sexual images or words in printed advertising do not significantly influence consumer preference towards the illustrations.

1.1.3 Priming, consumer behaviour, and product choices

One of the studies that was able to demonstrate the effect of subliminal priming on consumer behaviour is of Strahan et al. (2002), which discovered that subliminal priming enhances persuasiveness of an ad and affects the actual consumption behaviour when certain conditions are met. In the study, thirsty participants who were subliminally primed with the concept of thirst (i.e. thirst related words) had the tendency to drink more compared to those who were not primed, and showed higher preferences towards the thirst-quenching beverage. A more recent study by Karremans et al. (2006) demonstrated that this priming effect also influences brand preferences. Participants were primed using a brand name and asked to make choices between two brands. The study shows that thirsty participants who were primed with the brand name ‘Lipton Ice’ (a thirst quenching drink) have a higher probability of choosing ‘Lipton Ice’, compared to non-thirsty participants.

Both studies discuss the role of goal relevancy and individuals’ motivation in the effectiveness of subliminal priming. For subliminal priming to work, the prime needs to be relevant for the pursuance of the goal, and the responsents must have the motivation to pursue the goal. For example, when a woman is exposed to a prime related to a shaver, she will not be influenced to buy it simply because there is no motivation to buy one in the first place. Strahan et al. (2002) demonstrated the same idea when exposing a beverage prime to non-thirsty individuals. Subliminal primes in this case act as a cognitive cue which activate needs that are already present. Hence, subliminal priming seems to provide ‘access’ to individuals’ existing needs for the primed object

(16)

16 and encourage the pursuance of the goal when it is applicable (i.e. when there is motivation to do so) (Strahan et al., 2002).

Now this is the key of the two aforementioned studies: participants clearly have (or at least were manipulated to have) the motivation to pursue the goal. In both Strahan et al.’s (2002) and Karremans et al.’s (2006) studies, the participants were thirsty or made to be thirsty by giving them specific thirst inducing foods (i.e. salty snacks). The need that was presented was a drink which is a deficiency or a basic need. Therefore, there is a clear and rather strong motivation for people to satisfy this need. The longer this need is unmet, the stronger the need will become (Huitt, 2004). What happens if the motivation is unrelated to specific needs, is rather latent or not as strong? Will the same subliminal priming effect work on a different product or service offerings other than food and beverage, for example leisure needs? Strahan et al. (2002) proved that this would still work by demonstrating that priming a sad face enhances the persuasiveness of an ad for a mood-restoring product. The need which was tested is the need to interact with other people, which is also a part of Maslow’s higher level of needs (Huitt, 2004; Reisinger, 2009). Other recent research showed that the subliminal priming effect can also work with no presence of the real motivation itself. By presenting the motivation in the people’s mind as ‘desirable’ can be enough to activate the willingness to pursue the goal, the concept known as subliminal conditioning (Veltkamp, 2011). With the above assumptions, it is expected that the subliminal priming effect will also work on individuals’ preferences of products or services that fulfil higher level of needs, such as self actualization needs e.g. leisure activities choices.

Studies discussed above summarized that subliminal priming not only enhances persuasion, but also affects consumer behaviour, preferences, and brand choices. Though it is important for the prime to be relevant to individuals’ needs at that specific point in time, the need

(17)

17 does not have to be ‘real’ and can be something that is ‘conditioned’. It does not matter whether the need pre-exists naturally or is created, as long as it is present and provides enough motive for an individual to pursue them, it is sufficient to make priming works.

With this notion, it is expected that individuals who are subliminally primed with words associated with a specific concept will have higher tendency to choose products or services which is in line with the concept compared to individuals who are not primed. When the aforementioned concept is applied into individuals’ choice of higher level needs i.e. leisure activities choices, it is assumed that the same subliminal effect will take place.

Goal-relevant (1, 2) (subliminal) prime

Ad’s persuasiveness (1)

Motive to pursue the goal (1, 2)

Consumption behavior (2)

(e.g. drink consumption)

Brand Choice (2)

Figure 1. The effect of subliminal priming towards ad persuasiveness, consumer behaviour and

brand choice (Karremans et al., 2006; Strahan et al., 2002)

1.2 Factors influencing the effectiveness of subliminal priming

Although the effect of subliminal priming on ad persuasiveness, consumer behaviour and brand choice is confirmed under certain condition or situational setting, it goes without saying that this effect is neither independent, nor isolated from other influencing factors. These factors can be

(18)

18 related to the nature of the prime itself, such as the type of prime varying between picture or word, the content of the prime, for example the use of brand name, associative word, or persuasion sentence, or the frequency of the prime (e.g. one time vs. multiple exposures) – to name a view. Influencing factors can also come from the priming target as well as the individual’s personal characteristics. Research conducted in psychoanalytics demonstrated that individuals‘ personal characteristics such as level in anxiety, social desirability, and personal preferences might moderate the subliminal priming effect (Bustin et al., 2015). Considering the importance of these influencing factors, the present study explores the potential influence of two possible predicting factors: personality traits and priming content.

1.2.1 Personality traits and subliminal priming

Despite being a seemingly obvious influential factor, limited research has focused on the impact of dispositional factors on the effect of subliminal priming. In the field of psychology, there are quite a number of studies surrounding the relationship between persuasion and personality traits. This so called message-person congruence is based on the understanding that varieties of personality traits result in different types of value or motivations, which eventually reflected in the individuals’ diverse sensitivity towards specific type of stimuli. Stimulus which are in line with individuals’ motivational orientation are therefore expected to be evaluated more positively (Hirsh, Kang, & Bodenhausen, 2012).

As one of the studies which specifically addresses the relevancy of personality traits in the effect of subliminal priming, Bustin et al. (2012) demonstrated that personal dispositions moderate the effect of subliminal priming towards consumer preferences. Participants who scored high in sensation seeking showed higher intention to drink “Red Bull” – a drink associated with high sensation seeking activities – after previously being primed with the brand name, compared to no

(19)

19 effect of priming when applied to participants who scored low in sensation seeking. The study proposed the following potential explanatory reasons: Firstly, depending on personality traits, people might have higher sensitivity to specific types of subliminal persuasion, but not to others. This is in line with the concept of message-person congruence. Secondly, although the prime which is presented is identical, depending on distinct personality, some people might have different perceptions of the subliminal stimulus than others. Lastly, personality traits are typically associated with inclination towards certain values or goal pursuance. Thus, it is expected that subliminal priming which is aligned with individuals’ dispositional tendency or personality is much more effective (Bustin et al., 2012; Yoshino et al., 2005).

This concept is used as the basis of the main research question in the present study. In order to test whether different personality traits result in different effects of subliminal priming towards product preference, we focused on one personality trait, which is sensation seeking, following the study by Bustin et al.(2015).

Goal-relevant (1, 2, 3) (subliminal) prime

Ad’s persuasiveness (1)

Motive to pursue the goal (1, 2, 3)

Consumption behavior (2)

(e.g. drink consumption)

Brand Choice (2)

Consumption intentions (3)

Prediction to

Personality traits (3)

(i.e. sensation seeking)

(20)

20

1.2.2 Sensation seeking

Sensation seeking was first proposed as an addition to personality traits in the 1970s by Zuckerman et al. Sensation seeking is “a trait defined by the seeking of varied, novel, complex, and intense sensations and experiences, and the willingness to take physical, social, legal, and financial risks for the sake of such experience” (Zuckerman, 1994, p. 27). Individuals who are high in sensation seeking are characterized by their interest to engage in risky activities, craving for adventure and new experience, and their need for novelty (Zuckerman, 1994).

In 1971, Zuckerman presented four factors that represent the dimensions of sensation seeking, which are thrill and adventure seeking, experience seeking, disinhibition (i.e. the needs to loosening of one social inhibition such as alcohol consumption and sexual sensation), and boredom susceptibility (i.e. the need for change and varieties). These four factors or subscales are then used as a basis to develop a means to measure individuals’ level of sensation seeking – known as Sensation Seeking Scale (“SSS”). The SSS went through quite number of revisions until the latest version in 1978 known as SSS-V (i.e. SSS form five) consisting of 40-items of forced-choice format (Zuckerman, 2007). In 2002, Hoyle, Stephenson, Palmgreen, Lorch and Donohew (2002) developed a simpler version of the SSS-V named Brief Sensation Seeking Scale (“BSSS”) which includes 8-item loadings with two items representing each of the four subscales of sensation seeking. BSSS has been tested for reliability and validity in comparison with SSS-V and proven to elicit the same result. Since then, BSSS has been widely used in studies surrounding sensation seeking (Hoyle et al., 2002).

Referring to the previously explained effect of subliminal priming towards leisure activities choices, it is hypothesized that :

(21)

21 H1: Individuals who are subliminally primed with words associated with sensation seeking will have higher tendency to choose high sensation leisure activities compared to individuals who are not primed.

Motive to pursue the goal

(controlled)

Goal-relevant subliminal prime

• Imperative i.e. “Seek Adventure” • Static i.e. “Adventure”

Preference for high SS leisure activities

• Skydiving • Hike-and-track

H1 (+)

Figure 3. Hypothesis 1 – Sensation seeking related subliminal prime positively influences the

preference for high sensation seeking activities.

Coming back to the relationship between personality traits and the effect of subliminal priming, it is expected that individuals who scored high in sensation seeking will be more reactive when primed with sensation seeking related words or concepts. Therefore, the second hypothesis was formulated:

H2a: The effect of subliminal priming towards preferences for products ranked high in sensation seeking is more pronounced for individuals who scored high in sensation seeking when exposed to sensation seeking-related primes.

H2b: The effect of subliminal priming towards preferences for products ranked high in sensation seeking is less pronounced for individuals scored low in sensation seeking when exposed to sensation seeking-related primes.

(22)

22

(controlled)

Goal-relevant subliminal prime

• Prime

• No Prime _{Preference for high SS leisure activities}

H1 (+)

Level of Sensation Seeking

• High (above sample mean)

H2ab (+)

Figure 4. Hypothesis 2a and b – Individuals‘ level of sensation seeking influences the effect of

subliminal priming towards product preference.

1.2.3 Priming content and subliminal priming

Once one potential predictor variable has been discussed, the second potential predictor variable from the priming nature point of view will be tested, namely priming content. According to the previous explanation of subliminal priming – subliminal priming will only work when people do not realize the influencing power of the prime and do not have the ability to control the influence (Strahan et al., 2002). A similar concept was discussed in the field of consumer psychology called automatic correction against persuasion (Laran et al., 2011). This attribute, known as a type of behavioural backlash, is a reversed persuasion effect when consumers unconsciously resists or build some sort of “fence” towards a persuasion attempt when they perceive marketing tactics as a source of persuasion. Consequently, “rather than behaving in a manner implied by the tactic, consumers may automatically behave in contrary manner” (Laran et al., 2011 p. 999) – hence the reversed behavioural effect. This phenomenon is referred to as reverse

priming effect (Laran et al., 2011).

Different priming effects of the exposure to a brand name versus a slogan were tested in the study. Brands – just like a person’s name, are perceived by consumers as something that is ‘necessary to have’ and is seen as non-persuasive – hence allowing a normal priming effect.

(23)

23 Slogans, on the other hand, are perceived as persuasive due to their nature to sell brands and their constant reminder of brand attributes (e.g. Walmart’s slogan “Save Money. Live better”). As a result, slogans generate reverse priming effect on the consumers. In the context of the present study, the findings are particularly interesting since it also proves that the reverse priming effect works in the case of subliminal exposure.

The above theory is underlying the third hypothesis; whether different priming content will result in different subliminal priming effects. Instead of drawing a comparison between slogans and brand names, the present study compares the use of imperative sentence and a static associative word in a prime. By itself, imperative sentence is a form of direct persuasion. Based on the theory of automatic correction against persuasion, the use of imperative sentences in a prime is expected to result in a reverse priming effect as with slogans due to its persuasive nature. In contrary, a static associative word will result in a normal priming effect. Hawkins (1970) had touched upon the use of imperative sentence in subliminal priming, by comparing the use of the static word ‘Coke’ in a prime to the imperative sentence ‘Drink Coke’. Although the study did not find a specific reverse priming effect, it was proven that the use of an imperative sentence did not increase the influence of the prime. In the present study, two content types of subliminal priming are compared; the imperative sentence ‘Seek Adventure’ and the associative static word ‘Adventure’. Based on the above points, it is hypothesized that:

H3a: The usage of imperative sentence ‘Seek Adventure’ in subliminal priming will result in reverse priming effect (i.e. less or absence of priming effect), while the use of static associative word ‘Adventure’ will result in the expected positive priming effect.

(24)

24

(controlled)

Presence of goal-relevant subliminal prime

• Prime

• No prime Preference for high

leisure activities

H1 (+/-) Priming content

• Imperative i.e. “Seek Adventure”

H3a (-)

Figure 5. Hypothesis 3a – The usage of imperative sentence in subliminal priming results in

reverse priming effect.

One other interesting argument that might support the third hypothesis is the possibility that there is actually a relationship between sensation seekers and the tendency to avoid persuasion tactics. Zuckerman (1971) mentioned that sensation seekers are generally more likely to “violate the rule”, “rebel”, do not like being controlled and are likely to engage in illegal activities. It is fair to assume that because sensation seekers do not like to be told what to do, they might have higher “automatic correction against persuasion” and thus, the reverse priming effect will be much more significant.

Based on the aforementioned reasons, it is hypothesized that:

H3b: The reverse priming effect from the use of imperative sentence is especially pronounced for participants scored high in sensation seeking.

(25)

25

(controlled)

Goal-relevant subliminal prime

• Prime

• No Prime Preference for high SS

leisure activities

H1 (+)

Level of Sensation Seeking

• High (above sample mean)

H2ab (+)

Priming content

• Imperative sentence “Seek Adventure”

H3ab (-)

Figure 6. Hypothesis 3b – The reverse priming effect is especially pronounced for participants

(26)

26

2 Data and Method

2.1 Participants

The initial sample size of the experiments consisted of 312 participants, 154 were recruited from Amazon’s Mechanical Turk (MT) – which was also used in the previous study done by Bustin et al. (2015) and discussed to be comparable in terms of strength with other data collection method. From this initial sample, 13 participants failed to watch the video included in the survey (completion time of survey < 0.50 seconds) and hence were excluded from the analysis. This created a final sample of 299 participants (52% female, 48% male, age ranges from 18 to 64), with the majority age group of 25-34 years (53%). From this final sample, all variables were checked for missing data using frequency test which resulted in 0% of missing data1_.

2.2 Procedure

The study was conducted using experiment Vignette study which was presented in the form of an online survey consisting of instructions and set of questions. In the beginning of the survey, participants were told that they will participate in a study about the relationship between people’s personality and the choice of leisure activities. The survey was divided into three main parts. In Part I, participants were asked about their general background information i.e. gender and age group. In Part II, participants were asked to fill in a preliminary survey which determines participants’ level of sensation seeking using Brief Sensation Seeking Scale (“BSSS”) following the study by Bustin et al. (2015). In Part III, participants were instructed to imagine themselves having won a package of leisure activity and need to choose one among different packages presented in the online survey site which especially designed for the experiment.

(27)

27 Before the participants were able to make their choice, they were instructed to watch a short commercial video of a bank which was neither to be paused, nor replayed. The video was designed to randomly subliminally prime one-third of the participants with a static word ‘Adventure’ (A), one-third with imperative words ‘Seek Adventure’ (SA), and the rest with no prime (NP). The prime was presented using the tachistoscopic technique by flashing a brief word in the middle of the video at rapid speed (one time for approx. 20ms). In order to ensure that participants paid close attention during the ‘commercial break’, they were informed that they could win a bonus prize if they were able to guess the bank’s slogan (shown at the end of the commercial). Once participants had been primed (or not) in the experimental and control condition, they were asked to state their leisure activities choice by clicking one of the activities packages. The packages include four types of leisure activities with different level of perceived sensation (from high to low): Skydiving, Hike-and-Track, Weekend in a Countryside, and a Spa Treatment. Finally, in order to ensure the subliminal nature of the prime, a validity check was conducted at the end of the experiment asking participants whether they had seen something unusual in the video in an open-end response format.

2.3 Measures

2.3.1 Participants’ level of sensation seeking

Participants’ level of sensation seeking was determined using a well-validated Brief Sensation Seeking Scale (Hoyle, 2002) consisting of 8 items rated on a 5-point scale (1 = strongly

disagree to 5 = strongly agree). Please refer to Annex 1 for the list of items contributing to the

scale. A preliminary test on the reliability and normality of distribution of BSSS was conducted on the 8 items of the scale. BSSS was shown to have a relatively high reliability with Cronbach’s Alpha=.787 and good correlations (score >.30) among all items, with no counter-indicative items

(28)

28 (confirmed by positive Cronbach’s Alpha for each item). A test for normality resulted in substantial skewness across items (skewness between -.5 and 1.5, kurtosis -1.5 and 1.5). The large number of sample (n=299) reduces the risk of underestimate of variance and substantive results in the analysis – and thus should not be worried or corrected (Field, 2014).

For the hypothesis testing, the author was interested in the difference between people who scored high in Sensation Seeking (“SS”) and people who scored low in SS. In order to divide the participants into two groups, a new variable was created by computing the mean of the level of sensation seeking as per BSSS. The means and the standard deviations of the studied quantitative variable is presented in table 2. Based on the aforementioned mean, the sensation seeking continuous variables were recoded into two nominal variable groups i.e. people that scored above average level of SS (High SSµ>2.7) and people who scored lower than average of SS (Low SS µ<2.7). From the total of 299 participants, 155 scored high in sensation seeking (54 percent female with the majority age group of 25-34 years) while 144 scored low in sensation seeking. (50 percent female with majority age group 25-34 years).

Table 1. Mean, Standard deviation and correlations of study (quantitative) variables

Variables Mean SD Cronbach’s Alpha

Sensation Seeking 2.797 0.801 (0.78)

2.3.2 Participants’ choice of leisure activities

In the survey, participants were able to select one out of four leisure activities which were presented i.e. 1) Skydiving, 2) Hike-and-Track, 3) Weekend in a Countryside, and 4) a Spa Treatment. The choice of activities were derived from the dimensions of sensation seeking developed by Zuckerman (1971) with two out of four representing the high sensation seeking

(29)

29 choices i.e. ‘Skydiving’ and ‘Hike-and-Track’, and the other two representing the low sensation seeking choices i.e. ‘Weekend in a Countryside’ and ‘Spa Treatment’. The reason to provide two choices of activities for each sensation seeking levels was to minimize other decisional or disability factors such as personal preferences or fear of height.

Skydiving signifies more extreme sports and hard adventures compared to hiking which characterizes soft adventure and recreational sport (Williams & Soutar, 2005 ; Zuckerman, 2007) (Rauter & Doupona Topič, 2011). One study showed that skydivers scored higher in sensation seeking compared to other participants including hikers and river rafters (Hymbaugh & Garrett, 1974).

To revalidate the perceived level of the activities, a pretest with a simple ranking task was conducted to 50 participants asking them to rank the activities’ levels of sensation seeking based on perceived danger (i.e. ‘please rank the four activities from 1 being the most dangerous and 4 being the least dangerous’) and perceived uniqueness of experience (i.e. ‘please rank the four activities from 1 being the most unique experience and 4 being the least unique experience’). The pretest confirmed that ‘Skydiving’ perceived to be most dangerous and unique among other activities, and therefore scored the highest in sensation seeking level. ‘Spa Treatment’ was perceived to be the least dangerous and unique among other activities, hence, it is assumed to have the lowest sensation seeking level. ‘Hike-and-Track’ ranked higher compared to ‘Weekend in a Countryside’ in terms of perceived danger, but ranked lower when it comes to perceived uniqueness. Thus, both activities are assumed to have moderate sensation seeking level.

(30)

30

2.4 Pre-test and manipulation

2.4.1 Subliminal check

Prior to the study, a preliminary test on the subliminal nature of the prime was conducted to 37 participants using 0.30 ms length of prime time. Six participants confirmed that they have seen a word flashing in the middle of the video with three able to recall the exact word (‘Adventure’ or ‘Seek Adventure’). Following the preliminary study, the prime duration was reduced to 0.20 ms. At the end of the main study, eight individuals out of 299 participants confirmed to have seen a word or screen flashing in the video, whereby four were able to read the word ‘Seek Adventure’ or ‘Adventure’.

2.4.2 Accessibility, applicability and motivation

Based on the notion that the accessibility and the applicability of the prime as well as the individuals’ motivation are all necessary for priming to work (Strahan et al., 2002), certain conditions were set in the present study. In order to ensure that participants have certain motivations to pursue the goal of choosing one of the leisure activities, participants were instructed in the beginning of the study to imagine themselves as having won a leisure package and had to choose one out of four options. By doing so, the present study took into account the role of goal-relevancy (i.e. choosing a prize) and motive to pursue the goal (i.e. willingness to go on a leisure activity) subsequent to the previous study by Strahan et al. (2002). As for the accessibility and the applicability of the choices, the four leisure packages were stated to be of equal value and include the same standard accommodation and transport (i.e. two-night-stay and three-course meal). The packages were mentioned to be valid for a period of ten months and are located in the area within 500 km from the participants’ residence. In this way, we try to make the choices as plausible and flexible as possible.

(31)

31

2.5 Results

2.5.1 Main effect of subliminal priming towards product choices

The preliminary part of the study which is also presented in the first hypothesis is to test whether an exposure to subliminal priming influences participants’ leisure activities choices.

H1: Individuals who are subliminally primed with words associated with sensation seeking will have higher tendency to choose high sensation leisure activities compared to individuals who are not primed.

To test the first hypothesis, a likelihood ratio statitic analysis was conducted using a multinomial logistic regression. Likelihood ratio statistic on the use of both primes ‘Seek Adventure’ and ‘Adventure’ (both recoded into a new variable called ‘Prime’) was compared to no exposure to prime (variable named ‘No Prime’), showing a significant association between the

presence of prime and participants’ choice of leisure activities, with

ᵪ

2 (1) = 8.13, p (.043) < .005. Further calculation of the odds ratio showed that for participants who were primed by either ‘Seek Adventure’ and ‘Adventure’ the odds of choosing high leisure activities were 0.694 higher than if they were not primed. This result supported the first hypothesis, whereby the participants who are subliminally primed with words associated with sensation seeking have a higher tendency to choose products or services that are in line with the concept of sensation seeking compared to participants who are not primed.

In order to have a more specific analysis on the priming effect, a frequency table analysis was conducted which look into each of the leisure activities choices. Based on the pretest which determined the sensation level of different activities, it is expected that people who were primed

(32)

32 with sensation seeking concept will have a higher tendency to choose high sensation leisure activities.

Table 2. Contingency Table – Primed individuals vs. Non Primed individuals

Rank Leisure activities Count %

Primed individuals 1 Weekend in countryside 79 39%

2 Hike-and-Track 47 23%

3 Spa 43 21%

4 Skydiving 32 16%

Total 201 100%

Rank Leisure activities Count %

1 Hike-and-Track 37 38%

Non Primed Individuals 2 Weekend in a countryside 29 30%

3 Spa 22 22%

4 Skydiving 10 10%

Total 98 100%

a. Total number of participants 299

b. Assumptions for log-linear analysis are met with no expected counts less than 1 and no more than 20% of the total less than 5.

Result confirmed the first hypothesis showing that there is a significant – and correspondingly the highest increase of percentage for people choosing ‘Skydiving’ in the primed group compared to non-primed group (56%). The result also confirmed the hypothesis in the way that the priming effect is not seen in lower sensation seeking leisure activities i.e. ‘Spa’ as it is shown a decrease of five percent of people choosing ‘Spa’ within the primed group compared to the non-primed group.

However, results on the moderate level sensation seeking activities showed some small unexpected results. ‘Weekend in a Countryside’ indeed showed a significant increase of 33 percent of people choosing the activity in the primed condition compared to the non-primed condition. Conversely, in the Hike-and-Track activity choice, there is a distinct decrease of 38 percent of people in the primed condition compared to the non-primed condition. This result is contradicting

(33)

33 with the pretest result where people ranked Hike-and-Track as the second highest leisure activities. This results suggest that the priming effect is significantly apparent in the highest and the lowest of sensation seeking level for the activities, but does not seem to be supported in the case of moderate level of sensation seeking activities.

2.5.2 Different effects of subliminal priming based on individuals’ level of sensation seeking

The second hypothesis which tests if there are different effects of subliminal priming between individuals who scored high and low in sensation seeking are presented by the following two sub-hypothesis:

H2a: The effect of subliminal priming towards preferences for products ranked high in sensation seeking is more pronounced for individuals who scored high in sensation seeking when exposed to sensation seeking-related prime.

H2b: The effect of subliminal priming towards preferences for products ranked high in sensation seeking is less pronounced for individuals scored low in sensation seeking when exposed to sensation seeking-related prime.

Before testing the hypothesis, a log-linear analysis on the general effect of participants’ sensation seeking level on product preference was conducted to see whether it is plausible for sensation seeking level to moderates the subliminal priming effect in the first place. The predictor variable used in the analysis was individual average score in sensation seeking with outcome variable leisure activities choices. The analysis revealed a highly significant association between

participants’ level of SS and the choice of leisure activities

ᵪ

2 (1) = 41.596, p (.000) < .001. According to the odds ratio, participants who scored high in sensation seeking the odds of choosing

(34)

34 high leisure activities were 0.225 higher than if they are low in sensation seeking (from hereafter referred to as ‘SS’ in the Result section). This confirmed the assumption that by itself participants’ level in sensation seeking affects their product preferences.

When the plausibility of the assumption is confirmed, the second hypothesis was tested using cross-tabs analysis, presenting two groups of participants’ average level of sensation seeking (High SS – M>2.797, Low SS – M<2.797) and the presence of prime (‘Primed’ and ‘Not Primed’) as two predictor variables, and outcome variable leisure activities choices (presented in Table 3).

Table 3. Contingency Table

Leisure Activities (LA) Spa Weekend in a countryside Hike-and-Track SkyDiving Total

High SS Primed Count 31 52 18 7 201

% within LeisChoice 69% 77% 56% 64% 100%

Not primed

Count 14 15 14 4 98

% within LeisChoice 31% 22% 44% 36% 100%

Low SS Primed Count 12 27 29 25 93

% within LeisChoice 60% 65.9% 55.8% 80.6% 100% Not primed Count 8 14 23 6 51 % within LeisChoice 40% 34.1% 44.2% 19.4% 144 Total Count 65 108 84 42 299 % within LeisChoice 100% 100% 100% 100% 100%

An analysis on the parameter estimates showed that the main effect of the present of Priming is the most important effect of the model (z=4.832 with 95% confidence interval), followed by the interaction between Priming x Leisure choice (z=-2.072) and SS category x Leisure choice (z=-1.945). The likelihood ratio of this model was significant

ᵪ

2 (7)=39.165, p<0.001. On the other hand, the interaction between the two independent variables Priming x SS

(35)

35 category was not significant for the model (p>0.05). This is in line with the K-way effect analysis which shows that removing the two-way interaction will affect the model (p<0.01) but not significant in case of removing the three-way interaction (p>0.05).

In summary, as also shown in the contingency table, majority of participants with High SS chose ‘Weekend in a Countryside’ (43%) with 77 percent of them being primed with sensation seeking concept. Moreover, 20.6 percent of the people scored high in SS chose ‘Hike-and-Track’ with 56 percent of them being primed, and 7.1 percent chose to go ‘Skydiving’ with 64 percent being primed. Participants with low SS were less likely to choose ‘Weekend in countryside (28.5%) and the majority chose Hike-and-Track (36%) with 65.9 percent and 55.8 percent of them having been primed with goal-relevant prime, respectively.

Across three leisure activities (Spa, Hike-and-Track, and Weekend in a Countryside), participants who scored low in SS had lower percentage of being effectively primed with goal relevant primes. Only in SkyDiving (21.5%), participants who scored low in SS shown a higher score in terms of being effectively primed with goal relevance prime with 80.6% of the participants are primed.

Based on the above analysis, the result confirmed the second hypothesis, whereby the leisure choice of high SS individuals seems to be more influenced by the present of goal relevance prime compared to low SS individuals. The two sub-hypothesis were confirmed; The effect of subliminal priming towards product preference is more pronounced for individuals who scored high SS when exposed to SS-related prime, and less pronounced for individuals scored low in SS. The study also highlights an interesting point where in the case for High SS individuals the influence of priming seems to work across all leisure activities, the presence of the prime only

(36)

36 seems to influence the low SS individuals’ choice on highest scored sensational activity i.e. Skydiving.

Although it is confirmed that both sensation seeking and subliminal priming affects individuals’ leisure activities choices, it is important to note that the above results do not say anything about the direction of the relationship – whether it is a positive or negative nor the differences between the type of activities. More importantly, it does not reveal the nature of the interaction, whether it is a moderating effect or merely two independent effect of two predictor variables (Figures 7 and 8). Henceforth, a cross-tabs analysis and moderating analysis were performed.

(controlled)

Presence of goal-relevant subliminal prime

• Imperative sentence “Seek Adventure” • Associative word “Adventure” • No prime

Preference for leisure activities

• Skydiving (High) • Hike-and-track (High) • Weekend in a countryside • Spa treatment

Personality traits i.e. level of Sensation Seeking

• High (above sample mean) • Low (below sample mean)

H2a,b

Figure 7. Expected moderating effect of sensation seeking level on subliminal priming effect

(controlled)

Personality traits i.e. level of Sensation Seeking

H2a,b

(37)

37 Cross-tab analysis was conducted comparing the priming effect between the high and low SS participants within the priming condition (Table 4).

Table 4. Contingency Table – Primed individuals High SS vs. Low SS Primed

individuals

Preference for activities Count %

1 Weekend in countryside 52 48%

High SS 2 Spa 31 29%

3 Hike and track 18 17%

4 Skydiving 7 6%

Total 108 100%

1 Hike and track 29 31%

Low SS 2 Weekend in a countryside 27 29%

3 Skydiving 25 27%

4 Spa 12 13%

Total 93 100%

The cross-tab analysis revealed an interesting result, showing a significant higher priming effect for high SS participants in lower SS activities. Weekend in a countryside and Spa have 66 percent and 122 percent higher priming effects among the high SS participants, respectively, when compared to low SS participants. In addition, lower priming effects were observed in higher sensation seeking activities (-76%) Skydiving and Hike-and-Track (-47%) among high SS participants when compared to low SS participants.

Based on the log-likelihood measure, the chi-square test showed a difference in unexplained variance from 377.39 to 321.23 = 56.16 – this change is significant (p <.05) which means that the model explains a significant amount of the original variability – meaning that both predictor variables i.e. SS level and priming have significant main effect to leisure activities

(38)

38 choices (SS level

ᵪ

2 (1) = 47.68, p (.000) < .005, priming content

ᵪ

2 (1) = 13.171, p (.040) < .005). However, the analysis showed that there is no effect of the interaction between the two

predictor variables – priming content and SS level – on leisure activities,

ᵪ

2 (1) = 11.104, p (.52)

>.005.

Table 5. Moderation analysis using hierarchical regression

R R2 _R2_change _B _SE β _t Step 1 .358 .128 .128** SS level .435 .066 .359** 6.568 Prime Presence .033 .112 .016 .293 Step 2 .360 .129 .001 SS Level .376 .121 .310* 3.1 Prime Presence .029 .113 .014 .259

Inter SS Level Prime Presence

.084 .145 .058 .579

*P <0.05 , ** p <0.01

Moreover, hierarchical multiple regression was performed to investigate the moderating effect of SS level to the effect of priming towards choices of leisure activities. The analysis was done by transforming categorical variables (i.e. presence of the prime) to dummy variables and centering the potential moderator (i.e. SS level). In the first step of hierarchical multiple regression, two independent variables were entered, namely SS Level and Prime presence (i.e. whether individual was primed or not). This model was statistically significant F (2, 296) = 21.79; p <.001 and explained 12.8% of variance choices of leisure activities. However, after the entry of the interaction between SS Level and Prime Presence at step 2, the total variance explained by the

(39)

39 model is not significant. In the second model, only SS Level as independent variable was statistically significant (β=.310, p < 0.05).

These results showed that prime presence has no significant effect after controlling for SS level, moderator and interaction. On the other hand, when a regression analysis was done with only each independent variables separetely i.e. Prime Presence with leisure activities, or SS Level with leisure activities, without taking into account other factors and zoom in on independent effects, the effect showed to be significant.

In summary, although these results supported the second hypothesis which confirmed that there is indeed different effects of subliminal priming between high SS and low SS participants, there is no moderating effect or any interaction between the two predictor variables (Figure 7b is confirmed).

2.5.3 Possible reverse priming effect resulting from the use of imperative sentence

Lastly, the third hypothesis is tested to determine whether or not the use of imperative sentence will result in reverse priming effect, especially for participants who scored high in sensation seeking. The following sub-hypotheses are tested:

H3a: The usage of imperative sentence ‘Seek Adventure’ in subliminal priming will result in reverse priming effect (i.e. less or absence of priming effect), while the use of static associative word ‘Adventure’ will result in the expected positive priming effect.

H3b: The reverse priming effect from the use of imperative sentence is especially pronounced for participants scored high in sensation seeking.

(40)

40 The results showed that there was a significant association between priming content and

individual’s choice of leisure activities

ᵪ

2 (1) = 9.833, p (.024) < .005. Based on the odds ratio, the individuals who were primed by ‘Adventure’ the odds of choosing high leisure activities were 1.586 higher than if they were primed with ‘Seek Adventure’. Another test was conducted by creating an additional variable “Use of imperative sentence”, in which a log-linear analysis

revalidated the significance of priming content as predictive variable (

ᵪ

2 (1) = 14.463, p (.020) < .005), showing that individuals who were primed by ‘Adventure’, the odds of choosing high leisure activities were 1.68 higher than if they were primed with ‘Seek Adventure’. The result confirmed hypothesis 3a, whereby in general the use of imperative sentence in priming lead to a lower priming effect to the choices of higher leisure activities.

Furthermore, a cross-tabs analysis was conducted to zoom in on the differences of the effect among different leisure activities.

Table 6. Contingency Table – Primed individuals Seek Adventure vs. Adventure Primed

individuals

Preference for activities Count %

1 Weekend in country side 49 48%

Seek Adventure 2 Hike and track 21 20%

3 Spa 19 18%

4 Skydiving 14 14%

Total 103 100%

1 Weekend in a country side 30 31%

Adventure 2 Hike and track 26 27%

3 Spa 24 24%

4 Skydiving 18 18%

Total 98 100%

(41)

41 When comparing the differences in percentage, a reverse priming effect is seen among three leisure activities namely ‘Skydiving’ and ‘Spa’ – which are the highest and lowest SS leisure activities, correspondingly. However, for moderate SS leisure activities (i.e. ‘Weekend in a Countryside’ and ‘Hike-and-Track’), the results showed no pattern of reverse priming effect. In fact, for ‘Weekend in a Countryside’, prime ‘Seek Adventure’ seems to provide a positive priming effect compared to ‘Adventure’. The result provides an additional point to the current confirmed hypothesis 3a, whereby the use of imperative sentence in priming lead to a lower priming effect to the choices of highest and lowest SS leisure activities, while no consistent effect seen among moderate SS activities.

Further analysis was conducted to compare whether there is a difference in the role of imperative sentence between high SS participants versus low SS participants by separating the sample from the initial data set. However, the chi-square analysis on the two separate data sets i.e. High SS (M > 2.797) vs Low SS (M<2.797) revealed no significant effects between priming content with leisure activities choices. This finding followed the moderation analysis which was conducted previously in relation to SS level - priming content has no significant effect when controlling for SS Level. On the other hand, when analysis is done on the complete sample, without zooming into different level of sensation seeking, it is resulted in significant effect.

(42)

42

(controlled)

Personality traits i.e. level of Sensation Seeking

H2a,b

Priming content

• Imperative i.e. “Seek Adventure” H3ab

Figure 9. Results of the present study – independent effects of individuals’ level of sensation

(43)

43

3 DISCUSSION AND CONCLUSION

Following the experiment, results suggest that both subliminal priming and individuals’ level of personality traits – more specifically the level of sensation seeking – affect individuals’ leisure activities choices. In addition, it is confirmed that the subliminal priming effect varies based on individuals’ level of personality traits.

3.1 Hypothesis 1

The test of hypothesis 1 found a significant association between the presence of the prime and individuals’ choice of leisure activities. In general, participants who were primed with word or sentence related to the concept of sensation seeking have higher probability to choose high sensation seeking activities compared to participants who were not primed. The strongest effect was seen in the highest sensation seeking activity (i.e. Skydiving) and almost no priming effect was seen in lower sensation seeking activity (i.e. Spa). The result revealed additional finding which was not discovered in previous studies in which the priming effect on individuals’ choice of leisure activities does not seem to present in the case of moderate level sensation seeking activities (Hike-and-Track and Weekend in a Countryside). This could be due to the reason that moderate level of sensation seeking activities are susceptible to individuals’ perception. This was proven in the pretest which was conducted to determine the level of sensation seeking amongst the four activities. While Skydiving and Spa remain at their constant ranks (the first and the last, respectively), Hike-and-Track and Weekend in a Countryside were ranked interchangeably (second and third place) when assessed through different measures (i.e. perceived danger, perceived uniqueness, word association test). Nevertheless, this finding succesfully confirms the result of previous studies on the effect of subliminal priming on consumer decisions e.g. product and service choices.

Priming the daredevils : the role of personality traits in the effect of visual subliminal priming on product preference