This paper intends to test hypotheses through two experiments.
Experiment 1 tested how scarcity promotion and cause-related events lead to consumers' impulse purchases, that is, hypotheses H1-H4;
Based on repeatedly verifying the conclusion of experiment 1, experiment 2 focused on the interaction of cause-related events and scarcity promotion on consumers' impulse purchases, that is, hypothesis H5.
Study 1a
1. Pretest
Purpose. Ascertain the limited time and quantity that have the same stimulating effect on participants.
Design. Based on the experimental design of Aggarwal et al. [78], 10 graduate students who have watched the e-commerce live stream at a university in China were selected to participate in the pretest (male to female = 1:1). In the pretest, firstly, we asked participants the limited time that can bring the sense of scarcity when watching the live stream. Then we asked participants to answer the limited quantities that can bring the same sense of scarcity. In the end, more than half (8 / 10) of participants answered that the 5-minute limited-time discounts make them feel scarcity. Therefore, we regarded 5 minutes as a limited promotion time. In addition, in the context of limited-quantity promotion, consumers will feel the competition. The scarcity perception is affected by the number of online viewers. In view of this, we also controlled the number of viewers. According to the relevant data of Taobao rural stream, the viewers of different rural streamers range from 20 thousand to 100 thousand while most streams’ online viewers are about 50 thousand. Therefore, the material sets the number of online viewers to 50 thousand.
The specific situation of the pretest is "Suppose you are watching an agricultural live stream, with about 50 thousand people online. The products involve seasonal agricultural products such as grapes, litchis, onions et.al. All products are sold in groups, with the price of each group ranging from 30–50 yuan. Within the first five minutes of each product being sold, you can enjoy a 30% off for placing an order (automatic deduction for placing an order)". After reading the experimental situation, all participants were asked to answer the number of limited promotion that brings the same scarcity feeling as a 5-minute limited promotion. Then we calculated the mean and mode of all results. Some studies have shown that a 30% off is the critical level of consumer price perception (Berkowitz and Walton, 1980). Therefore, we set the price discount as 30%.
Results. All participants thought that the 5-minute could feel scarce. The average of the corresponding was 1088.88 and the mode was 1000 (6 / 10). In live streams, the limited quantity is usually an integer. Therefore, we use 1000 groups as the limited quantity of discounted agricultural products for subsequent formal experiments.
2. Method
Participants. In Study 1a, 118 subjects were recruited from Credamo, which is a professional online platform in China. After completing the experiment, each subject will receive 1 yuan as a reward. Subjects who haven’t watched the E-commerce live stream and failed to pass the screening were excluded. As a result, 98 valid samples (55.1% female, 32 in the limited-time promotion group, 32 in the limited-quantity promotion group, 34 in the non-promotion information group) were obtained. The demographic information of Study 1a is shown in Table 1.
Table 1
The demographic information of Study 1a
Items | | Frequency | Percentage(%) |
Gender | Male | 44 | 44.90 |
Female | 54 | 55.10 |
Age | 18–25 | 50 | 51.02 |
26–35 | 40 | 40.82 |
36–45 | 6 | 6.12 |
Over 45 | 2 | 2.04 |
Monthly disposable income | Under 1000 | 10 | 10.20 |
1000–2000 | 21 | 21.43 |
2001–3000 | 11 | 11.22 |
Over 3000 | 56 | 57.15 |
Design and Procedure. Study 1a adopts a single factor (limited-time promotion vs limited-quantity promotion vs no-promotion) inter-subject experimental design. The purpose is to verify H1-H3, that is, the positive impact of scarcity promotion on consumers' impulse purchases and the mediating role of arousal. Through the collection of information on agricultural products with blocked sales, the experimental materials selected the seasonal fruit - Kyoho Grapes. After investigating the market price, it was determined that the grapes were sold in groups, with 5 kg and 39 yuan per group. An impulse purchase is a sudden and unplanned purchase behavior, so the experimental situation is set as "You open Taobao to choose a T-shirt, and occasionally click into the benevolent live stream of Kyoho grapes with blocked sales...". In order to dispel consumers' concerns about the quality of grapes, the experimental materials also explained the blocked reasons with "Mature Kyoho grapes have sufficient glucose and compact fruit, but it faces the salable dilemma due to its single sales channel." The limited-time promotion in the experimental materials was set as "The discounted time of grapes in the live streaming room starts from being put on the shelf. In the first 5 minutes, you can enjoy 30% off for placing an order (automatic deduction for placing an order). But orders exceeding the limited-time need to be purchased at the original prices". The limited-quantity promotion was set as "The discounted grapes in the live streaming room are counted from being put on the shelf. The first 1000 pieces can enjoy 30% off (automatic deduction for placing an order). While from the 1001st piece it needs to be purchased at the original prices". The control group is without promotion information. In order to avoid the impact of real events on the experimental results, the experimental materials used "an ecological vineyard in a village" instead of grape origin. The number of live online viewers in all scenarios was 50 thousand.
Measure. The scale used in the experiment is appropriately modified based on the maturity scale to accord with the context of this study. The impulse purchase scale refers to the scale [79] prepared by Beatty & Ferrell (1998), including three items (Cronbach's = 0.939) such as "There is a great possibility to buy Kyoho grapes in the live stream, although I didn't want to buy them before". The arousal scale refers to the scale [80] compiled by Russell & Mehrabian (1977), which includes three items: "The discount information provided by the material makes me feel 'excited', 'stimulated' or 'Aroused'" (Cronbach'sα = 0.918). The perception of scarcity promotion is measured by "In this live stream, I feel that the time (quantity) that I can enjoy discount is limited", "In this live stream, I feel that if I buy within the promotion time (quantity), I can save more money". Considering the attractiveness of grapes described in the experiment to different subjects, personal price awareness, and monthly disposable income may disturb the result, we also measured them. Product attractiveness is measured by " Kyoho grapes in the material are attractive to me". Personal price awareness refers to the scale [81] of LICHTENST et al., which includes three items such as "I don't want to spend extra energy looking for a lower price" (Cronbach'sα = 0.887). On the other hand, we hold that personal self-construal will affect the results. Self-construal represents how individuals define themselves and perceive their personality [82]. It is divided into independent self-construal and interdependent self-construal. People with an independent self-construal pay more attention to personal goals rather than effective social relations. On the contrary, people with an interdependent self-construal pay more attention to the goals of others [83]. Self-construal was measured with reference to Choi & Totten's scale [84] (Cronbach's α independent = 0.742, Cronbach's α interdependent = 0.800). The following self-construal coefficients were constructed: (interdependent-independent) / (interdependent + independent) [85]. All scales adopt Likert 7-point scale (1= "strongly disagree", 7 ="strongly agree").
3. Results and discussion
Manipulation check. The manipulation check of scarcity promotion was successful(Compare with absolute standard 4)(M limited−time=6.09, t(31) = 16.63,p < 0.001; M limited−quantity=6.27, t(31) = 21.05, p < 0.001). Harman’s one-factor test showed that the percentage of the first principal component in the cumulative total variance is less than 40%, so there was no serious common method variance.
Hypothesis testing. Firstly, we estimated the proposed main effects of scarcity promotion. Independent samples t-test showed that there were significant differences in scarcity promotion groups. In limited-time promotion group, M limited−time promotion = 5.52, M n− promotion information = 3.67, t (64) = -4.89, P < 0.001; In Limited-quantity promotion group, M limited−quantity promotion = 5.81, M no−promotion information = 3.67, t (64) = -6.66, P < 0.001. This showed that scarcity promotion could promote consumers' impulse purchase, supporting H2a、H2b. Further, the independent samples t-test was conducted for the limited-time promotion group and the limited-quantity promotion group. The results indicate that there is no significant difference (M limited−time promotion = 5.52, M limited−quantity promotion = 5.81, t (62) = -0.93, P > 0.05).
Next, the mediating effect of arousal was estimated by Bootstrap [87]. Firstly, we tested the mediating effect of the limited-time promotion. The results showed that limited-time promotion played a positive role in consumers' arousal (β = 0.731, P < 0.001), supporting H1a. The limited-time promotion had a positive impact on impulse purchases. The total effect was 0.638 (t (65) = 2.57, P < 0.05), supporting H2a. The mediating effect of arousal was significant (95%, LLCI = 0.070, ULCI = 0.742). The indirect effect was 0.380. After controlling the mediate variable, the direct impact of limited-time promotion was not significant (95%CI= [-0.157, 0.673], P > 0.05). That is, arousal played a complete mediate role between limited-time promotion and impulse purchase, supporting H3. Similarly, the mediate effect test was carried out for the limited-quantity promotion. The results showed that limited-quantity promotion exerted a positive impact on consumers' arousal (β = 0.281, P < 0.001). The H1b was supported. It exerted a positive impact on impulse purchase too, with a total effect of 0.325, t (65) = 2.48, P < 0.05, supporting H2b. The mediating effect of arousal was significant (95%, LLCI = 0.009, ULCI = 0.266, indirect effect b = 0.111). After controlling arousal, the influence of the limited-quantity promotion on impulse purchase wasn’t significant (95% CI:-0.032,0.460, P > 0.05). It meant that arousal played a complete mediate role between limited-quantity promotion and impulse purchase. The H3 was supported. The specific regression coefficients are shown in Table 2 and Table 3.
Table 2
the mediating effect of arousal (limited-time promotion)
Variable | Impulse purchase |
Total effect | Mediating effect | Direct effect |
Covariate | | | |
Price awareness | 0.080 | -0.038 | 0.119* |
Self-construal coefficients | 0.822 | -0.004 | 0.826 |
Product attractiveness | 0.861*** | 0.361*** | 0.499*** |
Monthly disposable income | 0.041 | 0.032 | 0.008 |
Independent variable | | | |
Limited-time promotion | 0.638* | 0.380* | 0.258 |
Mediator | | | |
Arousal | | | 0.519*** |
R2 | 0.794 | 0.662 | 0.871 |
Note: *, * *, * * * in the table represent P < 0.05, 0.01 and 0.001 respectively |
Table 3
the mediating effect of arousal (limited-quantity promotion)
Variable | Impulse purchase |
Total effect | Mediating effect | Direct effect |
Covariate | | | |
Price awareness | 0.068 | 0.013 | 0.055 |
Self-construal coefficients | 0.715 | 0.111 | 0.603 |
Product attractiveness | 0.756*** | 0.223*** | 0.533*** |
Monthly disposable income | 0.115 | 0.077 | 0.038 |
Independent variable | | | |
Limited-time promotion | 0.325* | 0.111 | 0.214 |
Mediator | | | |
Arousal | | | 0.393*** |
R2 | 0.793 | 0.667 | 0.831 |
Note: *, * *, * * * in the table represent P < 0.05, 0.01 and 0.001 respectively |
4. Conclusion
Study 1a tests how scarcity promotion impact impulse purchases. The results show that scarcity promotion (limited-time promotion and limited-quantity promotion) can promote consumers' impulse purchase by stimulating arousal. Moreover, arousal plays a complete mediate role. The hypothesis H1-H3 is supported. Study 1a also examines there is no significant difference between limited-time promotion and limited-quantity promotion on impulse purchases. Next, study 1b will test the impact of cause-related events on impulse buying and its internal mechanism.
Study 1b
1. Method
Pretest. The purpose of the pretest is to determine the cause-related events used in the experiment. We investigated the blocked reasons for agricultural products and collected cause-related news. At last, we extracted "a special live stream of Kyoho grapes with blocked sales"、"charitable project for farmers"、" to solve the sale problem, a local enterprise sent its vehicles to purchase the first batch of mature grapes as employee benefits" as information of cause-related events.
Participants. In Study 1b, 75 subjects were recruited through Credamo to participate in the experiment. After completing the experiment, each subject will receive a reward of 1 yuan. Finally, we got 68 valid samples (63.2% female, cause-related events 31 and without cause-related events 37). The demographic information is shown in Table 4.
Table 4
the demographic information of Study 1b
Items | | Frequency | Percentage(%) |
Gender | Male | 25 | 36.80 |
Female | 43 | 63.20 |
Age | 18–25 | 62 | 91.20 |
26–35 | 6 | 8.80 |
36–45 | 0 | 0 |
Over 45 | 0 | 0 |
Monthly disposable income | Under 1000 | 11 | 16.18 |
1000–2000 | 28 | 41.18 |
2001–3000 | 7 | 10.29 |
Over 3000 | 22 | 32.35 |
Design and Procedure. Study 1b adopted a single factor (cause-related event: yes vs. no) inter-subject experimental design. The purpose is to test hypothesis H4. The materials provided information on cause-related events.
Measure. The impulse purchase intention, arousal, personal price awareness, and self-construal were the same as in Study 1a. The moral elevation refers to the scale [74] prepared by Jiang & Zheng. The scale is divided into three dimensions, including 11 items such as "all sectors of society make an effort to help farmers, it makes me moved"(Cronbach'sα = 0.969). The scale has high internal consistency.
2. Results and discussion
Manipulation check. The manipulation test of cause-related events showed that the manipulation was successful (Compare with absolute standard 4) (M cause event group = 5.48, t (30) = 7.366, P < 0.001).
Hypothesis testing. First, we estimated the main effect of cause-related events on impulse purchases. The independent sample t-test results showed that cause-related events can stimulate consumers' impulse purchases (M experimental group = 4.688, M control group = 3.721, t (66) = -2.801, P < 0.01). The first half of H4 was supported.
Next, the Bootstrap was used to examine the mediate effect of moral elevation [87]. The results showed that cause-related events had a positive impact on consumers' moral elevation(β = 0.788, P < 0.001). The mediating effect of moral elevation was significant (indirect effect b = 0.335, 95% CI= [0.106, 0.603]). After controlling the mediate variables, the direct impact of cause-related events on impulse purchase wasn’t significant (95% CI= [-0.309, 0.521], P > 0.05). It meant that moral elevation played a complete mediate role between cause-related events and impulse purchases. The second half of H4 was supported. The specific regression coefficients are shown in Table 5.
Table 5
the mediating effect of moral elevation
Variable | Impulse purchase |
Total effect | Mediating effect | Direct effect |
Covariate | | | |
Price awareness | 0.125 | 0.055 | 0.070 |
Self-construal coefficients | 0.359 | -0.333 | 0.692 |
Product attractiveness | 0.742*** | 0.180*** | 0.562*** |
Monthly disposable income | 0.019 | -0.066 | 0.085 |
Independent variable | | | |
Cause-related events | 0.441* | 0.335*** | 0.106 |
Mediator | | | |
Moral elevation | | | 0.425*** |
R2 | 0.704 | 0.505 | 0.765 |
Note: *, * *, * * * in the table represent P < 0.05, 0.01 and 0.001 respectively |
3. Conclusion
Study 1b examines the impact of cause-related events on impulse buying. The results show that cause-related events promote consumers' impulse buying by moral elevation. The moral elevation plays a complete mediate role. Hypothesis H4 was supported.
Study 1 examines the effects of scarcity promotion and cause-related events on consumers' impulse purchases. The results show that both cause-related events and scarcity promotion can stimulate impulse purchases. In marketing practice, businesses often integrate a variety of marketing methods to achieve better marketing effects. Studies have shown that strategies that combine intrinsic motivation with extrinsic motivation are not always effective [76]. Next, we will explore the interactive effects of scarcity promotion and cause-related events on consumers' impulse purchases.
Study 2
1. Method
Participants. After screening the samples, 124 valid samples were obtained (Female = 62.9%, 31 cause-related events * scarcity promotion, 31 cause-related events * no scarcity promotion, 30 no cause-related events * scarcity promotion, 32 no cause-related events * no scarcity promotion). After completing the experiment, each subject will receive a reward of 1 yuan. The demographic information of Study 2 is shown in Table 6.
Table 6
the demographic information of Study 2
Items | | Frequency | Percentage (%) |
Gender | Male | 46 | 37.1 |
Female | 78 | 62.9 |
Age | 18–25 | 45 | 36.3 |
26–35 | 66 | 53.2 |
36–45 | 8 | 6.5 |
Over 45 | 5 | 4.0 |
Monthly disposable income | Under 1000 | 5 | 4.0 |
1000–2000 | 26 | 21.0 |
2001–3000 | 24 | 19.4 |
Over 3000 | 69 | 55.6 |
Design and Procedure. Study 2 adopted a 2 (cause-related events: yes vs no) * 2 (scarcity promotion: yes vs no) between-subject experimental design. The purpose is to measure the interactive effect of cause-related events and scarcity promotion on consumers' impulse purchases. In order to improve the external validity of the experiment, Study 2 used purple garlic as the agricultural product material. In Study 2, "Alibaba platform subsidizes all discounts" was added to highlight the cause-related events. Limited-time and limited-quantity promotions were simultaneously presented, nevertheless, with different discounts in different periods. All subjects were randomly divided into four groups. The experimental procedure and measurement scale were the same as in Study 1.
2. Results and discussion
Manipulation check. The manipulation of cause-related events in Study 2 was successful (M cause-related events = 5.68, t (61) = 14.10, P < 0.001). The manipulation of scarcity promotion in Study 2 was successful (M scarcity promotion = .5.90, t (60) = 20.05, P < 0.001). The data passed the Harman common method bias test [86]. There was no serious common method bias problem.
Hypothesis testing. Firstly, we analyzed the main effect of cause-related events on subjects' impulse purchases. The scores of cause-related events were significantly higher than without cause-related events (M cause-related events = 5.350, M no cause-related events = 4.307, t (122) = -3.885, P < 0.001). Then the main effect of scarcity promotion was tested. The results showed that scarcity promotion played a positive role on consumers' impulse purchase (M scarcity promotion = 5.383, M no scarcity promotion = 4.291, t (122) = -4.089, P < 0.001).
Then, we adopted the covariance analysis. The price attractiveness, price awareness, and self-construal were regarded as covariates. The results showed that the main effect of cause-related events was significant (F (1,117) = 28.303, P < 0.0005). The main effect of scarcity promotion was also significant (F (1,117) = 10.906, P < 0.05). The H1 and H2 have been verified again. The interaction between cause-related events and scarcity promotion on impulse purchase was significant (F (1,117) = 13.870, P < 0.0005). In scarcity promotion group, M cause-related events = 5.344, M no cause-related events = 5.422, t (59) = -0.296, P > 0.05. In no scarcity promotion group, M cause-related events = 5.355, M no cause-related events = 3.260, t (61) = 5.531, P < 0.001. That is scarcity promotion moderates the impact of cause-related events on an impulse purchase. Without scarcity promotion, cause-related events have a greater positive impact on impulse purchases. The H5 was supported. The interaction effect is shown in Fig. 2.
3. Conclusion
An impulse purchase is an immediate purchase behavior aroused by external stimuli [13], which usually undergoes less cognitive processing. According to the motivational crowding theory, if the scarcity promotion and cause-related events are presented to consumers at the same time, the external motivation caused by promotion will cover up or even destroy the internal motivation caused by the cause-related events [88]. It requires consumers to expend more cognitive resources to balance the conflict between an individual's positive image cognition and bargain hunter image. The results of Study 2 show that scarcity promotion and cause-related events would interact to affect consumers' impulse purchases. When adding scarcity promotion, consumers are mainly driven by promotional incentives. At this time, cause-related events will no longer significantly affect consumers' impulse purchases. When without scarcity promotion, the cause-related events play a significant positive impact on consumers' impulse purchases.