Choice versus Given: Influence of Choice on Effectiveness of Retailers' Sweepstakes Promotion

Abstract

Purpose: This paper aims to investigate the influence of different methods of distributing sweepstakes (i.e., whether consumers choose to enter into the sweepstakes themselves or they are given the sweepstake ticket by default) on the effectiveness of the sweepstakes promotion (i.e., interest in the sweepstakes and intention to participate in the sweepstakes). Research design, data and methodology: The paper verifies this effect through three experimental studies: an online experiment using a sweepstakes promotion scenario at a department store, an online SNS sweepstakes promotion event, and a face-to-face card lottery game. Results: Participants belonging the group that chose sweepstakes tickets by themselves showed higher interest and intention to participate in the sweepstakes than those who were given the sweepstakes ticket by default. Furthermore, the group that chose the sweepstakes card thought it had a higher probability of winning than the group given the sweepstakes card. Conclusions: This paper shows a way to enhance the promotional effect of sweepstakes in the retail stores, without incurring additional costs, by approaching from sweepstakes design from the psychological perspective of the consumer. The study also sheds new light on the effect of sense of control manipulation using choice behavior in the promotional context.

1. Introduction

It is a fundamental human desire to want to have a sense of control in certain situations (Rucker et al., 2012). The sense of control is the belief that one can master, control, and shape the environment around him/her to his/her liking (Averil1, 1973; Lachman, 1986; Lachman & Weaver, 1998; Ross & Sastry, 1999). Companies are gradually empowering their customers and customers increasingly demand more control (Bues et al., 2017). Inesi et al. (2011) demonstrated that power and choice are substitutable. Choice is the expression of freedom and self-determination, which enhances sense of control (Averill, 1973; Langer, 1975; Zuckerman et al., 1978).

This study examines the effect of choice on effectiveness of sweepstakes promotion. Namely the difference between “I choose” and “I am given (by default)” with regards to the decision to enter into the sweepstakes, and how it affects interest in and intention to participate in the sweepstakes. In addition, the effect of sweepstakes distribution (i.e., by choice or by default) on the subjective possibility of winning is also investigated. A sweepstakes is a sales promotion which involves the offering of prizes to participants, where winners are selected by chance and no consideration is required (American Marketing Association, AMA Dictionary). In some countries (e.g., Korea), individuals can participate in sweepstakes when purchasing products. However, in some countries (including US), there are “no purchase necessary laws”, so marketers induce direct purchase through sweepstakes using alternate means of entry (e.g., locate codes inside product, and entering these codes online give access to entry). Sweepstakes are used not only to increase sales, but also for various purposes such as enhancing brand awareness, lead generation, and consumer engagement (Schulten & Rauch, 2015; Jung et al., 2020). They are often utilized at offline distribution stores such as supermarkets, discount stores and department store, as well as through online media. As a result of the rapid and global adoption of the Internet, online sweepstakes have become an indispensable tool in dialogue marketing (Schulten & Rauch, 2015).

To date, most sweepstakes have been conducted in the form where a seller gives an entry ticket to consumers, that is, consumers are given tickets by default. For example, in a store lottery promotion, a consumer receives an entry ticket and places the ticket in the entry box. The same goes for online and app sweepstakes events where the entry ticket (or entry number) is automatically given and a winner is selected through a computerized lottery. A Sweepstakes in the form of scratch-off tickets are also sent to the consumers in a DM (direct mail) or given by the salesperson in the same way. Sweepstakes ticket given in this way may reduce consumer interest, as the consumer was a passive actor in the program. In turn, sweepstakes that do not arouse interest may already have lost their meaning.

Sales promotions provide consumers with a variety of benefits in addition to financial savings (e.g., Dhar & Hoch 1996; Hoch et al., 1994; Inman et al., 1990; Ward & Hill, 1991). Chandon et al. (2000) divided promotional benefits into utilitarian benefits (savings, higher quality products, and improved shopping convenience) and hedonic benefits (opportunities for value expression, entertainment, and exploration). Sweepstakes is one of the promotions that provide hedonic benefits. Ward and Hill (1991) classified the reasons for participating in the sweepstakes into extrinsic values, such as prize money and winning probability, and intrinsic values, such as fun and excitement. Therefore, when consumers have fun and greater interest in the sweepstakes, it can lead to more participation in the sweepstakes and more purchase. The line between entertainment and marketing communication has become increasingly blended or even erased during recent years, particularly in the internet context (Zhang et al., 2010). In SNS, sweepstakes-based marketing has been successful in gaining consumers’ attention and in fostering their engagement (Jung et al., 2020; Schulten & Rauch, 2015). This is because the sweepstakes has a strong characteristic of inducing fun and interest compared to other types of promotions. SNS is a medium that functions through consumer participation (Cheung et al., 2015; Christy & Tracy, 2015). Heinonen (2011) reported that high consumer engagement is motivated by consumers’ need for entertainment, social connection, or information.

Prior research into sweepstakes has mainly focused on the prize structure and design, because they are conducted by lottery based on probability (e.g., Kalra & Shi, 2010; Schultz et al., 1998), while the behavioral approach (presentation form and context effect) has been overlooked. This study focuses on a method that can increase consumers' interest in sweepstakes - even under the same prize reward conditions - by approaching from the consumer's point of view rather than the sweepstakes itself. In other words, this research investigates how interest in sweepstakes promotion can be induced through manipulation of customer's choice behavior rather than by the compensation of prizes. This study sheds new light on how choice (i.e., “I choose”) affects interest, subjective winning probability, and the intention to participate in sweepstakes.

2. Conceptual Background

2.1. Choice and Sense of Control

Sense of control is a belief that the outcome of a specific task is related to one's behavior (Lachman, 1986; Lachman & Weaver, 1998; Ross, 1999; Rotter, 1966). A similar concept is self-efficacy, which is a belief in an individual's capability to successfully accomplish specific tasks (Bandura, 1977). Mourali et al. (2018) described that ‘sense of control’ and ‘self-efficacy’ are both self-focused, but the two concepts are different. That is, it is possible to believe that an outcome is related to one's behavior, but it is, nevertheless, not possible to believe that one has the ability to perform a particular behavior (Bandura, 1977; Haidt & Rodin, 1999). Self-efficacy has been studied mainly in relation to children's education and emotional regulation.

Sense of control has been researched along with topics such as social class (e.g., Kraus et al., 2009), education (e.g., Slagsvold & Sørensen, 2008), and mental health (e.g., Lachman & Weaver, 1998), in the context of social psychology. According to prior research, sense of control reduces fear, and lessens stress due to stimuli that arouse pain or anxiety (Lefcourt, 1973). When a sense of control is lost, the individual will attempt to restore it (Whitson & Galinsky, 2008).

Moreover, sense of control was studied in relation to choice behavior. Choice is an intrapersonal construct related to the ability to choose preferred behaviors. Choice behavior enhances a sense of control (Averill, 1973; Inesi et al., 2011; Lefcourt, 1973) and is an expression of freedom, autonomy, and independence (Markus & Schwartz, 2010; Taylor, 1989). Choice expresses an independent self and makes it appear as an active agent capable of shaping destiny and influencing the world (Markus & Schwartz, 2010). Individuals are more intrinsically motivated when they have choices; that is, when they have self-determination (Zuckerman et al., 1978). Inesi et al. (2011) mentioned that choice and power are interchangeable. In other words, the absence of one increases the desire for the other, and when the other is acquired, the need for control can be satisfied (Inesi et al., 2011). Since there is a threshold effect, when one is satisfied, the size of the satisfaction brought by the other decreases (Inesi et al., 2011). Meanwhile, power is the ability to control one's own and others' resources and performance (Galinsky et al., 2003; Galinsky et al., 2006). Power increases decision-making confidence, which is the belief that one is confident to make better decisions (Galinsky et al., 2008; Rucker et al., 2014). This result that “choice and power are interchangeable” hints at the effect of sense of control in choice behavior.

Research regarding sense of control and choice in the context of consumption behaviors are rarely conducted, but some studies have recently begun. For example, Botti and Mcgil (2010) showed that self-made decisions induce internal attribution and result in more satisfaction than externally given choices. Chang (2008) indicated that when service recovery options are presented so that the consumer can choose them in a service failure, the consumer's sense of control and satisfaction increase. Lee (2017) reported that lack of control leads to a tendency for reward consumption of purchasing products that are scarce. Only a very small number of studies have been conducted on the effect of sense of control on consumption behavior, and even fewer that have studied it in the context of sales promotion.

2.2. Sweepstakes and Lottery Promotions

Sweepstakes promotion is an important promotional tool for companies (e.g., Ward & Hill, 1991). According to the Promotional Marketing Association's 2007 Annual Promotions Industry Trends report, companies spent $1.83 billion on sweepstakes and contests, with 10.5% of marketers reporting that sweepstakes were the biggest communication expense (Kalra & Shi, 2010). Consumers are sometimes reluctant to use discount coupons due to concerns that using coupons may make them appear low-income to others. In this respect, promotions utilizing luck, such as lotteries, are sometimes more effective (Dhar & Hoch, 1996). Reid et al. (2015) commented that increasing the use of non-monetary promotions (like sweepstakes) instead of monetary promotions (like price discounts) can improve brand value and brand equity. Sweepstakes cost money like other promotions, but they help to avoid the possible impact of price discounts (i.e., discount expectation and purchase delay, lowering the reference prices). In addition, the costs of sweepstakes are very predictable (Jung et al., 2020).

Sweepstakes promotion, despite its importance, is only sporadically found in literature (Jung et al., 2020). Prior research on sweepstakes can be broadly divided into four types. First, studies regarding the motivation and reason for participating in sweepstakes were conducted. Kwang (1965) provided an explanation of the “indivisibility of spending” in that expensive items, such as automobiles and ships, cannot be divided indefinitely. In other words, reasonable consumers with limited incomes will participate in lottery promotion to acquire such expensive products. Another explanation is based on availability bias and representative heuristic. Companies generally communicate past prize winners, so consumers can have increased positive thoughts with regards to participating for prizes (Ward & Hill, 1991). Another point of view on the reason is that consumers are motivated by the excitement, social interaction, positive feedback about their competence, and escape from problems (Ward & Hill, 1991).

Second, studies were conducted in relation to the personal traits (i.e., demographic and socio-psychological characteristics) of consumers who are more responsive to sweepstakes. Regarding the psychological characteristics, influences of competitiveness, numeric interest, present-time orientation, locus of control, self-esteem, sensation-seeking, and impulsiveness were investigated (Fang & Mowen, 2009; Taylor & Kopp, 1991; Ward & Hill, 1991; Schulten & Rauch, 2015). Teichmann et al. (2005) suggested that customer disparity is greater online. Narayana and Raju (1985) compared the difference of the demographic and socioeconomic characteristics between consumers who prefer gift promotions and consumers who prefer sweepstakes.

Third, research studies on the effectiveness of sweepstakes were conducted. Huff et al. (1999) compared the gap between perceived consumer response and brand managers' use of monetary and non-monetary promotions. They suggested that brand managers were under-utilizing non-monetary promotions (samples, premiums, sweepstakes, contests, frequent user programs, etc.) and indicated that brand managers prefer monetary promotions due to competition and short-term pressure (Huff et al., 1999). Another study found that mail surveys using lottery prize giveaways have higher response rates and are more cost-effective than using cash (Balakrishnan et al., 1992).

Fourth, research has been conducted on sweepstakes design, which includes determining the total reward money, number of winners and probability of winning, reward allocation by ranking, reward type (cash or prize), reward period and frequency, determination of reward rank division, determination of advanced announcement of winning probability, determination of immediate rewards, etc. (Schultz et al., 1998). Kalar and Shi (2010) reported that to maximize consumers' perception of value, a sweepstakes format should consider the promotional objective and consumers’ attitude towards risk. According to them, one large prize (‘winner takes all’) is more effective when targeting current high-brand-valuation consumers, and many smaller prizes when attracting new customers. They also mentioned that multiple large prizes are more effective for risk-averse consumers and one grand prize (‘winner takes all’) for risk-neutral consumers (Kalar & Shi, 2010). Yan and Muthukrishnan (2014) showed that lottery promotion consisting of a large prize and consolation prize together lowered consumers’ valuations of the lottery and their intent to participate by shifting their focus from the value of the large prize to the probability of winning. Jung et al. (2020) proposed a sweepstakes design for social media and demonstrated that higher total prize money, more winners, short term, one grade, cash prizes increased consumer engagement. A study was also conducted that showed that sweepstakes were preferred when purchasing utilitarian products over hedonic products (Kim & Min, 2014).

3. Hypotheses Development

This study deals with factors that can increase interest in the sweepstakes from the consumer's perspective. In other words, this research investigates how to increase participation by changing consumers’ interest in the sweepstakes and perception of the possibility of winning, given equal prize design conditions. The effectiveness of lottery promotion does not increase simply by presenting a large number of prizes, but only when consumers' perception of winning possibility is high (Laporte, 2009).

Previous studies have demonstrated that availability and representativeness heuristics increase the perception of winning probability (e.g., Tversky & Kahneman, 1974; Ward & Hill, 1991). This study verifies that “perception/sense of control” by choice can increase perception of winning probability. Choice behavior increases a sense of control (Averill, 1973; Inesi et al., 2011; Lefcourt, 1973), and uncertainty decreases when a sense of control increases (Lefcourt, 1973). In probability games with high uncertainty, like sweepstakes, if the perception of uncertainty decreases, the perception of winning probability expectedly increases. Chandran and Morwitz (2005) dealt with the effect of a sense of control in participative pricing. Participative pricing mechanisms are where consumers participate in setting a final price for a product. They showed that not all consumers increased purchase intention in participatory pricing and individuals with high perceived control revealed higher intent to purchase than individuals with low perceived control (Chandon & Morwitz, 2005). Langer (1975), in his experiment with 53 office workers, asked half of the participants to choose a $1 lottery ticket while the other half were randomly given lottery tickets. Then, a few days later, the two groups were asked for how much they would give up their lottery tickets, before the lottery draw began. The result showed that participants who chose the tickets themselves asked for a higher amount than the participants who were given the tickets ($8 vs. $2). Langer describes this as “illusion of control.” This study predicts that individuals who choose the sweepstakes tickets by themselves will have a higher perception of the winning possibility than individuals who were given tickets. People are intrinsically motivated when they have a choice (Zuckerman et al., 1978). It is expected that individuals have a higher commitment into something when their own choices are made into it. If consumers choose the sweepstakes entry tickets by themselves, they will be more interested in what their choices will bring. That is, they will be more interested in the sweepstakes promotion. Therefore, this study derived the following hypotheses.

H1: Consumers who have a choice to select sweepstakes entry tickets will have a higher interest in the sweepstakes than consumers who are given the tickets.

H2: Consumers who have a choice to select sweepstakes entry tickets will have higher perception of possibility of winning a prize than consumers who are given the tickets.

Choice is an expression of autonomy (Markus & Schwartz, 2010; Taylor, 1989). Wohlfeil and Whelan (2005) commented that as a pull strategy in marketing communication, it is necessary to motivate consumers to voluntarily participate in marketing events for a successful event marketing strategy. This study expects choice behavior to increase the willingness to participate.

H3: Consumers who have a choice to select sweepstakes entry tickets will have a greater intent to participate in the sweepstakes than consumers who are given the tickets.

4. Empirical Studies

4.1. Study 1: Experiment Using Sweepstakes Scenario (Online)

4.1.1. Experimental Design and Data Collection

For the first experiment, a hypothetical sweepstakes promotion scenario was used. The experiment was conducted online through a marketing research company. One group of participants could choose their own sweepstakes entry tickets. The other group of participants were given sweepstakes entry tickets. Participants were randomly assigned to the two groups. The experimental stimulus used in the scenario were as follows. First, prizes were designed to be paid in cash. Jang and Mattila (2020) mentioned that cash rewards increase consumer engagement. Kalar and Shi (2010) were cited for determining the amount of prize money. The scenario had a sweepstakes prize of $1,000 when purchasing a product worth $100. Although there is “no purchase necessary laws” in US, sweepstakes are linked to purchase by using alternate means of entry indirectly. Also, as described in the introduction, some countries offer sweepstakes after purchase. So, in Experiment 1, participate in a sweepstakes after purchase design was used. The sweepstakes scenario which is not related to purchase was used in Experiment 2. The reason for $100 was to make the subjects consider participating in the promotion more carefully by not making the purchase price too low, and thus to measure the intent to participate more clearly and reliably. The probability of winning the prize was not directly presented, but indirectly with the phrase “$10,000 in total prize and 10 draws.” Multiple prize grades were not used to simplify the experiment. The brand name was only revealed as Brand A.

At the beginning of the survey, participants were revealed to the phrase “Imagine Brand A is running a sweepstakes promotion when you visit a department store. If you purchase Brand A, you can participate in the sweepstakes. Furthermore, if you win, you will receive a prize. The prize is $1,000 in cash” (Refer to Appendix 1). At this point, the choice manipulation group was presented with the phrase “You can choose the prize ticket yourself” and asked to choose one number between 1 and 100. It was explained that the lottery draws a number from 1 to 100 and gives you a prize if the same number is drawn. This number choice design was used by citing the method used in the lottery. The survey was programmed so that when the participants entered their number and clicked the ‘Next’ button, the number chosen earlier by them was displayed again on the next page. Then, the participants responded to questions such as ‘interest in the sweepstakes,’ ‘willingness to participate in the sweepstakes,’ and ‘perception to their chances of winning.’ To the non-choice (i.e., given) condition group, the phrase ‘the salesperson will give you an entry ticket’ was presented and a number between 1 and 100 is randomly presented on the next page (Refer to Appendix 2).

For sampling, an allocated sampling method by gender and age (20s, 30s, 40s, 50s) was used. A total of 200 samples were used for the analysis (100 men: mean age = 39.60 years old, 100 women: mean age = 39.15 years old).

4.1.2. Measurements

‘The degree of interest in the sweepstakes promotion’ was measured on a 7-point semantic differentiation scale (1: not at all interested, 4: moderately interested, 7: very interested) using the question “are you interested in this sweepstakes event?” ‘The intent to participate in the sweepstakes’ was measured on a 7-point semantic differentiation scale (1: not at all willing, 7: very much willing) using the question “are you willing to participate in this sweepstakes?” ‘The perception of chance to win’ was measured on a 7-point semantic differentiation scale (1: unlikely to win at all, 7: most likely to win) with the question “how likely do you think the chosen (given) lottery entry ticket will be won?” ‘The risk-seeking orientation’ was measured by referencing a previous study (Chu et al., 2014) whereby participants were presented with the following sentences: “there is a lottery with a 50% chance of winning $1000 or a 50% chance of getting nothing. How much are you willing to pay for it?” They were then being asked to write their desired purchase price using a ratio scale. Participants who do not wish to purchase were given instructions to enter 0 dollar.

4.1.3. Results

ANCOVA was used to test the hypothesis while the control variables included risk-seeking propensity, age, education level, and income.

The results of the analysis, as shown in Table 1, revealed that there was a significant difference between groups in the degree of interest in the sweepstakes (M_choice=4.460 vs. M_given=4.060, F(1,194)=4.585, p=0.034). Regarding the willingness to participate in the sweepstakes, there was a marginally significant difference between the groups (M_choice=4.435 vs. M_given=4.040, F(1,116)=3.776, p=0.053).

Table 1: Influence of “Choice” on Sweepstakes Effectiveness – Group Comparison (Choose vs. Given) : ANCOVA

A bootstrapping analysis using model 4 of the Hayes Process macro was performed to check the mediating effect that choice behavior raises interest in sweepstakes and that higher interest in sweepstakes leads to higher intention to participate in sweepstakes (Table 2). Group variables were made into dummy variables (choice group=1, non-choice group=0) and included as independent variables. As shown in Table 2, the analysis results showed that the choice behavior indirectly affects the intention to participate in the sweepstakes by mediating the interest in the sweepstakes (indirect effect=0.4138, 95% confidence interval=[0.0511, 0.7984]).

Table 2: Analysis of Mediating Effects between Choice, Interest, and Intention To Participate : Bootstrap Analysis using Hayes Process Macro Model 4

Note: Interest means ‘Interest in Sweepstakes’. Participate means ‘Intent to Participate in Sweepstakes’

Regarding the perception of winning probability, the mean of the choice group was higher, but there was no statistically significant difference between groups (M_choice = 3.110 vs. M_given=2.940, p=0.373). There was a significant correlation between interest in the sweepstakes, intent to participate in the sweepstakes, and perception of winning probability in the sweepstakes (‘interest in prizes’ & ‘perception of winning probability’: pearson r = 0.438, p < 0.001, ‘intent to participate’ & ‘perception of winning probability’: pearson r=0.515, p<0.001). The overall average of participants' perception of winning probability was below the mid-point (i.e., 4 point) in absolute terms (midpoint analysis: mean = 3.025 vs. midpoint 4.0; t (119) = -9.08, p<0.001).

To sum up, the group that chose the sweepstakes entry ticket number was more interested in the sweepstakes than the other group. The purpose of sweepstakes promotion is to attract consumer interest and participation. The results of this study found that choice improves the intent to participate in the sweepstakes through interest. Although choice was expected to increase the sense of control which in turn was expected to increase the perception of winning, there was no statistically significant difference between the groups in the probability of winning.

4.2. Study 2: Experiment Using SNS Sweepstakes Event (Online)

4.2.1. Experimental Design and Data Collection

The second experiment used a scenario referring to the real case of sweepstakes promotion on SNS. The object of the promotion was to increase social media followers. As described in the previous theory section, event-type sweepstakes to increase consumers’ engagement are often used on SNS. The prize was a product gift, not cash, and several prize grades were presented (Refer to Appendix 3). In this experiment, there were no choice actions like actually drawing or entering a number. Only the phrase ‘You can choose (are given, in the non-choice condition) an entry ticket’ was presented. The survey was conducted online through a marketing research company and the allocation sampling method for each gender and age range (20s, 30s, 40s, 50s) was used.

For the experiment, the event was introduced along with the explanation, “Company A is holding a sweepstakes event to increase social media followers. If you follow both of the company’s Twitter and Facebook, you can participate in the sweepstakes. To participate in the sweepstakes, click the ‘follow’ button on the event page, then enter your information.” Then, the following phrases were presented according to the choice vs. non-choice condition groups. For the choice group, the phrase ‘you can choose the sweepstakes entry ticket yourself’ was presented, while the non-choice group saw the phrase ‘sweepstakes entry ticket is automatically given.’ Afterwards, participants were directed to respond to the survey questionnaire.

The total number of participants was 457, consisting of 224 men (mean age=39.5 years) and 233 women (mean age=39.18 years old). Participants were randomly assigned to either the choice (225 persons) or non-choice (232 persons) manipulation conditions.

4.2.2. Measurements

‘Interest in this sweepstakes promotion,’ ‘intent to participate in this sweepstakes,’ ‘perception of winning possibility,’ and ‘risk-seeking tendency’ were measured in the same way as in the previous experiment.

4.2.3. Results

ANCOVA was conducted to test the hypothesis in the same way as in Experiment 1. As Table 3 shows, there was a significant difference between the groups in ‘interest in the sweepstakes’ (M_choice=4.500 vs. M_given=4.124, F(1,451) =5.901, p=0.016). There was a marginally significant difference between groups in ‘intent to participate in sweepstakes’ (M_choice=4.220 vs. M_given=3.915, F(1,451)= 3.499, p=0.062). Moreover, there was a significant correlation between ‘interest in sweepstakes,’ ‘intent to participate in sweepstakes,’ and ‘perception of winning probability’ (‘interest in sweepstakes’ & ‘perception of winning probability’: pearson r=0.466, p<0.001, ‘intent to participate’ & ‘perception of winning probability’: pearson r=0.497, p<0.001). However, there was no significant difference between groups in the perception of winning probability (M_choice=2.591 vs. M_given=2.431, p=0.229)

Table 3: Effect of Presenting the word ‘Choice’ on Interest in Sweepstakes : ANCOVA

Summing up, it was found that the mere presence of the word ‘choice’ made people more interested in sweepstakes. However, the perception of winning probability did not differ between groups. The average perceived probability of winning (mean=2.512) was lower in this study than that in study 1 (mean=3.025). Consequently, the word "choice" can increase interest in the lottery event, but is limited in increasing sense of control.

4.3. Experiment 3: Experiment Using a Card Game (Face-to-Face)

4.3.1. Experimental Design and Data Collection

In the third experiment, a lottery-type card game similar to the actual sweepstakes was performed face-to-face. This experiment was designed to make the choice behavior more explicit and to induce participants to make a choice with the belief that if they win, they will indeed receive a prize. As in the previous experiments, the 'perception of control' was compared between the group in which the card was chosen and the group in which the investigator gave the card to the participant.

The experiment was conducted with the students of a university in Korea, by posting an announcement to the university campus community and inviting those who wished to participate in the experiment to the lab. The study was set up so that participants could visit the lab individually. In a real sweepstakes event, consumers usually check the winning of the sweepstakes individually, so the experiment was conducted individually as well. This method was used to exclude the possibility that the autonomy of choice would be affected when the choice was performed alongside many people.

When the participants arrived at the lab, they were introduced to the experiment. The purpose of the experiment was described as a study related to probability in order to give the participants a reason for playing a card game and a sense of realism about winning. The participants were then given leaflets and asked to read them. The flyers for the choice group contained the following phrases: 'Choose one card. If the same card as the one you chose is drawn through a random draw from the same set of cards, you will receive a prize of $50.' After placing the cards face up, participants were asked to choose the one they wanted. For the non-choice group, a card was randomly given to the non-choice group by the researcher. The description of the lottery method was presented in the same way and the group assignment was random. Afterwards, a questionnaire was asked about the ‘the perception of winning possibility’ and ‘risk-seeking tendency.’

At the end of the experiment, a debriefing was conducted and $5 was paid to each of the participants as a participation reward. A total of 120 students participated in the experiment, of whom were 64 men (mean age=22.03) and 56 women (mean age=21.46).

4.3.2. Measurements

‘Perception of winning probability’ was measured on a 7-point semantic differentiation scale with the question 'How did you feel about your chances of winning with the chosen (given) card?’ (1: not likely to win at all, 4: moderate, 7: sure to win). ‘Interest in game’ was measured on a 7-point Likert scale with “I enjoy this game”. The risk-seeking propensity was measured in the same way as before.

4.3.3. Results

ANCOVA was conducted in the same way as previous studies to test the hypothesis. The risk-seeking propensity and age were included as control variables while education level and income were not included since the experiment was conducted on college students.

The result of the analysis revealed that there was a significant difference between the groups in the interest in the game, and the choice group showed a higher interest in the game than given group (M_choice=3.77 vs. M_given=3.41, F(1,116)=3.990, p=0.047). Moreover, as revealed in Table 4, the analysis result showed that there was a significant difference in the perception of the winning possibility between groups (Mchoice=3.42 vs. M_given=2.97, F(1,116)=4.938, p=0.028). In other words, it was found that the choice group felt a higher chance of winning than the non-choice (given) group. As in previous experiments, participants perceived the odds of winning as below average (mean=3.20 vs. midpoint 4.0; t (119)=-5.419, p< 0.001). However, it was higher than that of Experiment 1 (mean=3.025) and Experiment 2 (mean=2.512). From Experiment 3, it is inferred that the perception of winning probability is higher when the choice is more clear.

Table 4: Effect of Choice on Perception of Winning Probability: ANCOVA

5. General Discussions and Implications

Previous studies regarding sweepstakes focused on the design and extrinsic value of prizes. However, the effectiveness of sweepstakes promotion does not increase only by presenting a large number of prizes. The consumers’ participation in sweepstakes can be increased only when consumers feel interest in the particular sweepstakes and feel that they have a high probability of winning. Even under the same sweepstakes design, if consumers' perception of interest and winning possibility can be enhanced, sweepstakes promotion effectiveness can be improved without incurring additional costs. This study hypothesized that when a sweepstakes entry ticket is offered to a consumer, providing the consumer with a choice can change its effectiveness.

In the online sweepstakes experiment of this study (Experiment 1), the group that chose the sweepstakes entry ticket number showed higher interest than the group in which it was given. It was also found that choice manipulation was mediated by higher interest in sweepstakes, leading to higher intent to participate in the sweepstakes. In the SNS sweepstakes event experiment (Experiment 2), the group that saw they could choose the sweepstake entry ticket showed higher interest than the group who automatically applied. This demonstrates that prior exposure to ‘choice’ can be effective even before choices are made. In Experiments 1 (online sweepstakes) and 2 (SNS sweepstakes event), the perception of winning probability was not significant. However, considering that sweepstakes is a promotion to provide hedonic benefits of entertainment and interest, higher interest means that there is higher promotional effectiveness since higher sweepstakes interest can lead to higher promotion participation and product purchase. In the card game experiment, the perception of winning probability of the group that chose the card was higher than the group that was given the card. The group average of winning probability perception in three experiments was shown in the order of Experiment 2 (online, SNS sweepstakes, choice phrase only), Experiment 1 (online, sweepstakes entry number choice), and Experiment 3 (face-to-face, physical selection behavior). This result may be due to participants believing that they could actually win as it was a face-to-face experiment in Experiment 3. In the case of Experiment 1, it did not give a sense that the prize money will actually be awarded. In addition, in Experiment 3, the actual physical choice behavior was performed offline, so it is interpreted that the effect of choice manipulation was more powerful than in the online experiment. In other words, it can be inferred that the stronger the choice manipulation, the greater the effect.

The theoretical contributions and implications of the study are as follows. First, this study reveals the effect of marketing manipulation with regards to sweepstakes participation and its mechanism, which has not been addressed in previous studies. Second, this study sheds new light on the manipulation effect of choice and sense of control in the new context of ‘consumption’ behavior, contributing in expanding the scope of existing studies related to the sense of control.

The management implications for the field of this study are as follows. This paper demonstrates that marketers would be able to conduct more effective sweepstakes promotion if they understand the characteristics of customers' sense of control. Another contribution of this research is that the study proposed a method to increase the promotion effectiveness without incurring additional costs.

6. Limitations and Directions for Future Research

This study has some limitations. In the design of number which participants can choose from (Experiment 1), choice was only allowed within the range of 1 to 100. This method was used because there could be an issue that, if a wide range of choices were given, it may have presented difficulties to the participants in making a choice. But the choice range is narrow, there may be a problem of duplication. In particular, numbers that are familiar to people (e.g., 7, 1) may be more likely to be chosen. However, this duplication problem can be addressed in the following way. As with the on-site immediate sweepstakes lottery, it can be solved when the sweepstakes entry ticket is placed in the entry box after choosing a number and the lottery is immediately drawn from the entry box. In online sweepstakes, the same problem can be solved by managing log data so that all the sweepstakes entry numbers chosen by each participant are recognized as different entities and by drawing a lottery among them. This mechanism needs to be notified to participants in advance so that participants can understand the reason why they did not win even with the same number. Another solution is to divide and distribute the rewards if the same number comes out during the lottery. In this case, prior notice is also required. In the experiments of this study, the choice was made in a state where the participants recognized that they could choose. In order for the choice manipulation to have a clear effect in an actual sweepstakes promotional event, ‘prior awareness of the choice’ must occur beforehand.

Suggestions for future research are as follows. From the experiment, it is not easy to expect to have the same immersion as in real sweepstakes. Future research may delve deeper into the effect of choice manipulation using real sweepstakes events. Previous studies have shown the effect of individual differences in the sense of control on participatory pricing (Chandon & Mowitz, 2005). In a similar vein, future research may focus on the individual differences in choice manipulation and its effect in sweepstakes promotion. Moreover, the interaction between customer demographic characteristics and choice manipulation needs further study. In addition, future research may be conducted on dissatisfaction from not winning a sweepstakes differs before and after choice manipulation. Furthermore, sweepstakes promotions take place in various channels such as home shopping, SNS, internet shopping malls, etc. Future studies may be conducted on the influence of choice manipulation by different channels on the sweepstakes effectiveness under the same conditions. Also, comparative study on the perceived probability of winning between choosing a lottery number and when an automatic random number is assigned can be made regarding lottery ticket purchases. One of the reasons for purchasing a lottery ticket is for the anticipated pleasure before drawing. As with a sweepstakes promotion, if individuals choose numbers by themselves for a lottery ticket, they may have more interest and the perception of winning may increase. These will be interesting topics for future research.

Appendixes