Symbol Matters: A Sequential Mediation Model in Examining the Impact of Product Design with Buddhist Symbols on Charitable Donation Intentions

: Though religion-informed charitable behavior has gained much academic attention recently, there is an ambiguous relationship between individual donation intention and Buddhist determinants. Drawing on the theories of reasoned action and product semantics, this study tries to examine the effects of product design with Buddhist symbols on individual charitable donation behavior and the underlying mechanism. Some studies have argued for the positive effects of religious symbols on people’s behavioral reactions; however, limited prior research has addressed the impact of a touchable product, such as a donation box, on people’s charitable donation intentions from the perspective of Buddhism. The current study statistically shows that donation boxes with Buddhist symbols could signiﬁcantly improve people’s charitable donation intention through two sequential mediators, namely, their perceived religiosity and product attitude. The study provides both theoretical contributions and practical implications for future research.


Introduction
Charitable donations are of importance to help the needy population and to improve well-being worldwide (Macías Ruano et al. 2020;Tsai 2021). Moreover, individual donations, except for government funding, are the largest funding source for many charitable organizations facing a shortage of funds (Lee and Chang 2009). However, according to Nonprofit Source's recent data, the annual growth of charitable donations has been slow in the last decade (Nonprofits Source 2018). Though many governments have launched economic approaches, such as tax-exempt programs, to encourage donations to charity, the prospects for individual donation intention remain unoptimistic worldwide (Lo 2012;Sargeant 1999). In this context, understanding the individual donors and what motivates them to contribute to charities is something that has been of increasing interest to nonprofit organizations and institutions.
Charity and religion are inherently bound together. Major religions, like Christianity and Hinduism, are characterized by teachings and doctrines that give importance to helping others (Arli and Lasmono 2015;Atkins et al. 2019). For this reason, prior studies suggest that people with such religious beliefs tend to have a higher intention to help the needy population (DeAngelis et al. 2016;Muralidharan et al. 2018;Ranganathan and Henley 2008). However, Buddhism, as one of the dominant religious belief systems across the world, is seldom empirically examined in terms of religion-informed charity, especially in the context of making donations. Recent data indicates that 9% to 10% of the world's total population is engaged with the practice of Buddhism (Center Pew Research 2012). Thus, the need for an empirical investigation of Buddhism-informed donation behavior has never been greater.
Religious symbols have been regarded as an effective informational determinant to predict charitable behavior based on the theory of reasoned action (Muralidharan et al. 2018). Relevant research reports that the positive impact of religious symbols presented on advertising is mediated by perceived religiosity to provoke the intention to perform actual behaviors (Henley et al. 2009;Naseri and Tamam 2012). Other media, like a touchable product such as a donation box affecting the behavioral performance, gains little attention. Previous scholars (Martin and Randal 2008) construe that the appearance of a donation box positively influences donors' actual behavior. Therefore, it is reasonable to anticipate that a donation box with proper religious symbols contributes to increasing the attitudinal intention to perform donation.
Drawing on the theories of reasoned action (Fishbein and Ajzen 2011) and product semantics (Krippendorff 1984), this study examines how Buddhist symbols presented on a product impact donation behavior, and suggest the causal route to positive donation intention. The results of this study provide these academic insights and also have practical implications, given the evidence that Buddhist symbols play an effective role in predicting individual donation intention.

Individual Charitable Donation Behavior and Buddhism
Charity includes giving a wide range of resources to help the underprivileged (Kashif et al. 2015). It has been suggested that individual charitable donations work as a direct way to support needy people where government funding has been limited, especially for those suffering from unexpected disasters and accidents (Lee and Chang 2009). Recently, there is a growing interest in investigating religion-informed charitable behaviors. Religious belief, as an emotional and spiritual factor, predicts charitable donations significantly (Yoon 2018). Major religions, such as Christianity, Hinduism, and Buddhism, advocate the teachings of helping others (Ranganathan and Henley 2008). Regarding this, the issue of how and why religious factors influence individual charitable donation has been examined and studied to improve current well-being worldwide. For example, the factor of religiosity has been proposed to positively predict individual donation intentions within the context of western religions, such as Christianity (Hopkins et al. 2014;Ranganathan and Henley 2008). However, little attention has been paid to the determinants of Buddhist giving.
These charitable behaviors have been rooted in the Buddhist doctrine and practice, such as "morality" (Snellgrove 1956) and "mercy" (Chamberlayne 1962). While Buddhists do not believe in the immortality of the soul, given the doctrine of anatta (no-self), on the popular level, they might believe in the immortality of the soul and reincarnation, which is shaped and evaluated by all the sins and virtues they performed as actual behaviors and present thoughts (Snellgrove 1956). The goddess of mercy, Guan Yin, in Chinese Buddhism has been a manifestation of the Buddhist deity of "compassion," Avalokitesvara. This suggests a supreme symbol of selfless love to advocate individual charitable behavior to cultivate virtues inside the self (Blofeld 2009). Though Buddhist religiosity has been examined in improving people's positive attitude towards worldwide well-being and prosocial behaviors, such as decreasing prejudice (Clobert et al. 2015) and taming materialism (Pace 2013), the impact of Buddhism-informed charitable donation behavior remains vague in prior empirical studies.

Buddhist Symbols and Religiosity
Religious symbols have been frequently exploited to influence people's attitudes and consequent behaviors (Baring 2018;Dotson and Hyatt 2002;Henley et al. 2009;Long 2019;Lumpkins 2010;Park and Baker 2007;Taylor et al. 2010;Wang and Jang 2018). The reason is that a religious symbol has been conceived to represent fundamentally conceptual characteristics of a specific religion through its iconicity (Tillich 1958), which provokes individual cognitive perceptions of symbolic spirituality (Hallman 2006). For instance, a recent empirical study shows that the symbol of goddess Durga in Hinduism has worked as a visual cue to increase the reporting intention of violence (Muralidharan et al. 2018).
Buddhist symbolism is of importance to construct the profound meaning system of Buddhism (Coomaraswamy 1935;Beer 2003). For this reason, Buddhist symbols have been utilized and manifested in many Buddhist artworks and sites to propagate the divine power and presence of Buddha (Miller 2015). For example, it has been suggested that the symbol of lotus has symbolized the purified self enlightened by Buddhist teachings and wisdom (Blau and Blau 2003;Lin 2007;Ward 1952). It is frequently represented in Buddhist sculptures as lotus pedestals to symbolize the fundamentally inner thought and meaning of much Buddhist philosophy (Ward 1952). Through the artistically visual representation, the cognitive comprehension of symbolic spirituality tends to be provoked when encountering Buddhist symbols. Nevertheless, the empirical effects of Buddhist symbols seldom have been examined before. Therefore, it is difficult for researchers to identify its significance and conclude evidence to nourish further studies.
Further literature has proposed that the moderator of religiosity elicited by religious symbols tends to positively predict attitudinal behaviors (Henley et al. 2009). The definition of religiosity refers to the extent to which an individual is devoted to religion (Hill and Hood 1999). Scholars have reported that religiosity tends to predict behavior through the attitudes shaped by a specific religious belief (Weaver and Agle 2002). A religious symbol with symbolic spirituality of a specific religion can underpin the perceived religiosity to some extent. Relevant studies have been conducted to investigate individual behavior predicted by the sequential mediators of religious symbols and religiosity dominantly in the context of Islam (Naseri and Tamam 2012), Christianity (Henley et al. 2009), and Hinduism (Muralidharan et al. 2018). However, the effects of sequential mediators of Buddhist symbols and religiosity remain ambiguous.

Theory of Reasoned Action and Product Semantics
Scholars (Fishbein and Ajzen 2011;Sheppard et al. 2002) have developed the model of reasoned action to frame and explain the mechanism of individual motivational behavior. It proposes that individual engagement in action is positively predicted by attitudinal intentions to perform the behavior. Attitude, defined as an individual's internal evaluation of an object, affects intention to act (Mitchell and Olson 1981). Many studies have examined the theory of reasoned action in relevant social issues in the past decade (Han et al. 2010;Chang 1998), but application to charitable behaviors has received limited attention. It has been suggested that individual actual charitable donation behaviors are influenced by the donation intention informed and framed by relevant message inputs (Cao 2016). As aforementioned, use of Buddhist symbols with spiritual and emotional meanings tends to predict religiosity; consequently, it is worth studying how attitudes influenced by religious inputs affect individual behavior in the discourse on charitable donation.
Previous literature on the religious symbol, as informational input, has mainly focused on the medium of advertising; other forms or mediums seldom have been investigated empirically. In recent research into charitable donations, the appearance of a donation box has been regarded as a crucial factor influencing donation behavior (Martin and Randal 2008). Moreover, product design with religious symbols has been discussed as an effective way to carry symbolic meanings to influence individual attitudes and elicit an emotional response (Jiang 2010;Yao and Ming 2015). Krippendorff and Butter (Krippendorff 1984) have introduced the term "product semantics" to theorize that product appearance serves as medium and form to express meanings through symbolic quality, and finally, provoke users' emotional and attitudinal response. Products are not only affording physical functions, but also communicate meanings to improve individual attitudes towards them. In other words, after exposure to a product with symbolic meanings, an individual might firstly have a cognitive perception of symbolic quality and then form an attitude regarding the product. For this reason, a religious symbol transformed as a visual sign on product appearance can elicit people's cognitive perceptions related to specific religious beliefs, then predict attitudinal reactions, and consequently influence behavioral intentions.

Hypotheses and Research Framework
Based on the observations mentioned above, we propose that a product, such as the donation box, with Buddhist symbols, for example with the lotus-shaped design, might significantly increase people's donation intention, and that this process is sequentially mediated by perceived religiosity and product attitude. As aforementioned in Section 2.2, the symbol of lotus plays a crucial role in indicating the fundamental teachings in Buddhism, which is widely accepted across different regions (Blau and Blau 2003;Lin 2007;Ward 1952). Concerning this relationship, we developed the following hypotheses: Hypothesis 1 (H1). Donation box with Buddhist symbol (vs. without Buddhist symbol) could increase the perceived religiosity.
Hypothesis 2 (H2). Donation box with Buddhist symbol (vs. without Buddhist symbol) could increase product attitude.

Hypothesis 3 (H3). Donation box with Buddhist symbol (vs. without Buddhist symbol) could increase donation intention.
Hypothesis 4 (H4). Donation box with Buddhist symbol has an indirect and positive impact on donation intention, mediated through the perceived religiosity.
Hypothesis 5 (H5). Donation box with Buddhist symbol has an indirect and positive impact on donation intention, mediated through the product attitude.
Hypothesis 6 (H6). Donation box with Buddhist symbol has an indirect and positive impact on donation intention, sequentially mediated through the perceived religiosity and then product attitude.
To examine and validate the related hypotheses, we conducted an empirical investigation to address the significant effects of Buddhist symbols in religion-related product design, improving the perceived religiosity, product attitude, and the consequent behavior reaction. Figure 1 shows the research framework with six hypotheses (H1-H6) and the related relationship among those constructs.
affording physical functions, but also communicate meanings to improve individual attitudes towards them. In other words, after exposure to a product with symbolic meanings, an individual might firstly have a cognitive perception of symbolic quality and then form an attitude regarding the product. For this reason, a religious symbol transformed as a visual sign on product appearance can elicit people's cognitive perceptions related to specific religious beliefs, then predict attitudinal reactions, and consequently influence behavioral intentions.

Hypotheses and Research Framework
Based on the observations mentioned above, we propose that a product, such as the donation box, with Buddhist symbols, for example with the lotus-shaped design, might significantly increase people's donation intention, and that this process is sequentially mediated by perceived religiosity and product attitude. As aforementioned in Section 2.2, the symbol of lotus plays a crucial role in indicating the fundamental teachings in Buddhism, which is widely accepted across different regions (Blau and Blau 2003;Lin 2007;Ward 1952). Concerning this relationship, we developed the following hypotheses:

Hypothesis 1 (H1): Donation box with Buddhist symbol (vs. without Buddhist symbol) could increase the perceived religiosity.
Hypothesis 2 (H2): Donation box with Buddhist symbol (vs. without Buddhist symbol) could increase product attitude.

Hypothesis 5 (H5): Donation box with Buddhist symbol has an indirect and positive impact on donation intention, mediated through the product attitude.
Hypothesis 6 (H6): Donation box with Buddhist symbol has an indirect and positive impact on donation intention, sequentially mediated through the perceived religiosity and then product attitude.
To examine and validate the related hypotheses, we conducted an empirical investigation to address the significant effects of Buddhist symbols in religion-related product design, improving the perceived religiosity, product attitude, and the consequent behavior reaction. Figure 1 shows the research framework with six hypotheses (H1-H6) and the related relationship among those constructs.

Measurement Items
Regarding the six hypotheses, we used a nine-point Likert scale to measure different variables. To be more specific, all the scales' items or questions were retrieved from prior relevant research. For example, the perceived religiosity was measured by a three item Likert scale from previous works (Van Cappellen and Saroglou 2012; Saroglou and Muñoz-García 2008) (sample item, "Buddhism is important to me", 1 = Strongly Disagree, 9= Strongly Agree). Product attitude was measured by a three item Likert scale adapted from a previous paper (Howard and Gengler 2001): sample item, "It is a good donation box", 1 = Strongly Disagree, 9 = Strongly Agree. Charitable donation intention was measured by five items (Ranganathan and Henley 2008): "The likelihood for me to donate money in this box is high" (1 = Strongly Disagree, 9 = Strongly Agree).

Stimuli and Pilot Study
The current research followed a between-subjects experimental design that contained two different scenarios (with Buddhist symbols and without Buddhist symbols). One professional product designer used C4D to model two types of stimuli: the donation box with a Buddhist symbol design, namely, a lotus symbol arranged on the top (for stimuli, see Figure 2a) and the donation box without Buddhist symbol design (Figure 2b,c). As stated in Section 2.2, the lotus symbol in this study referenced the shape of the lotus pedestal in both Indian and Chinese Buddhist statues in history. Buddhism originated in ancient India (Harvey 2013) and China has been reported as the major country practicing Buddhism, with the largest population of Buddhists (BuddhaNet 2015). Consequently, the representation of the lotus symbol in this study can maintain the original and recognizable figure quality. Figure 3a,b presents the original images of the lotus pedestals.

Measurement Items
Regarding the six hypotheses, we used a nine-point Likert scale to measure different variables. To be more specific, all the scales' items or questions were retrieved from prior relevant research. For example, the perceived religiosity was measured by a three item Likert scale from previous works (Van Cappellen and Saroglou 2012; Saroglou and Muñoz-García 2008) (sample item, "Buddhism is important to me", 1 = Strongly Disagree, 9= Strongly Agree). Product attitude was measured by a three item Likert scale adapted from a previous paper (Howard and Gengler 2001): sample item, "It is a good donation box", 1 = Strongly Disagree, 9= Strongly Agree. Charitable donation intention was measured by five items (Ranganathan and Henley 2008): "The likelihood for me to donate money in this box is high" (1 = Strongly Disagree, 9= Strongly Agree).

Stimuli and Pilot Study
The current research followed a between-subjects experimental design that contained two different scenarios (with Buddhist symbols and without Buddhist symbols). One professional product designer used C4D to model two types of stimuli: the donation box with a Buddhist symbol design, namely, a lotus symbol arranged on the top (for stimuli, see Figure 2a) and the donation box without Buddhist symbol design (Figure 2b,c). As stated in Section 2.2, the lotus symbol in this study referenced the shape of the lotus pedestal in both Indian and Chinese Buddhist statues in history. Buddhism originated in ancient India (Harvey 2013) and China has been reported as the major country practicing Buddhism, with the largest population of Buddhists (BuddhaNet 2015). Consequently, the representation of the lotus symbol in this study can maintain the original and recognizable figure quality. Figure 3a,b presents the original images of the lotus pedestals.  The reason why the designer made two different donation boxes without Buddhist symbols (Figure 2b,c) lies in: donation box (b) is the most commonly used box in the Buddhist temples (Marketplace 2011), and its shape is different from the six-sided cylinder of donation box (a). Thus, we could not confirm whether the Buddhist symbol, rather than the shape of the six-sided cylinder, influences people's donation intention.
In order to avoid the potential confounding factors stemming from existing perceptions of donation boxes, the designer carefully made two donation boxes with the same height, positioning, money-donation direction, and color, as shown in Figure 2b,c. After that, a pilot study was run to rule out the effect of the shape of the six-sided cylinder (donation box (b) vs. (c)) on people's donation intention. Sixty college students with Buddhist beliefs (mean age = 20.5; 56.6% female) were recruited in the experiment to confirm whether the cylinder shapes of these two donation boxes influence donation intention. They were firstly randomly assigned to two groups (each group contained 30 students) and they were asked to observe the specific donation box (Figure 2 box (b) vs. (c)) and report their intention to donate. T-test result confirmed that there was no significant difference on donation intention between two scenarios (Figure 2 box (b) vs. (c); t(58) = 0.982; p = 0.32), statistically ruling out the influence of the shape of the six-sided cylinder on donation intention. Thus, in the following study, we could use the donation box (a) as the experimental stimulus and the donation box (b) as the control (typical) stimulus.

Experiment Procedure
To examine the relation between donation boxes with Buddhist symbols and their corresponding effects, a sample group was recruited from Amazon Mechanical Turk, AMT, in this experiment. AMT is a valid data collection source since a great many cognitional, perceptual and social experiments have been conducted through AMT, and it has indicated sufficient quality and reliability, compared with experimental data collected in physical laboratories (Saritoprak et al. 2018). Instead of convenience sampling, AMT could recruit a more diverse and random sample in which the selection criterion for participants was to hold Buddhist beliefs.
A total of 140 participants with the selection criterin (holding Buddhist belief) were recruited for this study (mean age = 36.01 years; 84 men and 56 women). They first The reason why the designer made two different donation boxes without Buddhist symbols (Figure 2b,c) lies in: donation box (b) is the most commonly used box in the Buddhist temples (Marketplace 2011), and its shape is different from the six-sided cylinder of donation box (a). Thus, we could not confirm whether the Buddhist symbol, rather than the shape of the six-sided cylinder, influences people's donation intention.
In order to avoid the potential confounding factors stemming from existing perceptions of donation boxes, the designer carefully made two donation boxes with the same height, positioning, money-donation direction, and color, as shown in Figure 2b,c. After that, a pilot study was run to rule out the effect of the shape of the six-sided cylinder (donation box (b) vs. (c)) on people's donation intention. Sixty college students with Buddhist beliefs (mean age = 20.5; 56.6% female) were recruited in the experiment to confirm whether the cylinder shapes of these two donation boxes influence donation intention. They were firstly randomly assigned to two groups (each group contained 30 students) and they were asked to observe the specific donation box (Figure 2 box (b) vs. (c)) and report their intention to donate. T-test result confirmed that there was no significant difference on donation intention between two scenarios (Figure 2 box (b) vs. (c); t(58) = 0.982; p = 0.32), statistically ruling out the influence of the shape of the six-sided cylinder on donation intention. Thus, in the following study, we could use the donation box (a) as the experimental stimulus and the donation box (b) as the control (typical) stimulus.

Experiment Procedure
To examine the relation between donation boxes with Buddhist symbols and their corresponding effects, a sample group was recruited from Amazon Mechanical Turk, AMT, in this experiment. AMT is a valid data collection source since a great many cognitional, perceptual and social experiments have been conducted through AMT, and it has indicated sufficient quality and reliability, compared with experimental data collected in physical laboratories (Saritoprak et al. 2018). Instead of convenience sampling, AMT could recruit a more diverse and random sample in which the selection criterion for participants was to hold Buddhist beliefs.
A total of 140 participants with the selection criterin (holding Buddhist belief) were recruited for this study (mean age = 36.01 years; 84 men and 56 women). They first consented to take part in the study, and were then randomly assigned to two scenarios (donation box (a) and (b)). To specify, each scenario contained 70 participants. First of all, they were informed that they are going to evaluate a charitable donation box for a Buddhist temple. Then, they were required to observe the donation box, then respond to several questions on the given donation box, and at last complete the manipulation statement at the end of the questionnaire (the extent of their agreement with the statement, "This donation box is decorated with a Buddhist symbol, the lotus." 1 = strongly disagree; 9 = strongly agree).

SPSS 22
.0 was introduced to analyze all the data. No missing or incomplete data was found in the current data set. Regarding the manipulation check, we could find a significant difference from the t-test. As predicted by the manipulation, people who were exposed to the donation box with the Buddhist symbol scenario (Mean = 6.07) rated higher than those exposed to the donation box without a Buddhist symbol scenario (Mean = 4.24; t(138) = −8.46, p < 0.01), suggesting that the manipulation check was successful.
Then, we conducted a sequential mediation analysis by the PROCESS macro for SPSS (Hayes 2015). By utilizing a bias-corrected confidence interval with bootstrapping (10,000 resamples), PROCESS could analyze both the corresponding direct and indirect effects. To be more specific, the PROCESS could prove the significant mediation effect when the confidence interval (CI) of the indirect effect does not contain zero. Figure 4 shows the result of the mediation analysis.
Religions 2022, 13, x FOR PEER REVIEW 7 of 12 consented to take part in the study, and were then randomly assigned to two scenarios (donation box (a) and (b)). To specify, each scenario contained 70 participants. First of all, they were informed that they are going to evaluate a charitable donation box for a Buddhist temple. Then, they were required to observe the donation box, then respond to several questions on the given donation box, and at last complete the manipulation statement at the end of the questionnaire (the extent of their agreement with the statement, "This donation box is decorated with a Buddhist symbol, the lotus." 1 = strongly disagree; 9 = strongly agree).

SPSS 22
.0 was introduced to analyze all the data. No missing or incomplete data was found in the current data set. Regarding the manipulation check, we could find a significant difference from the t-test. As predicted by the manipulation, people who were exposed to the donation box with the Buddhist symbol scenario (Mean = 6.07) rated higher than those exposed to the donation box without a Buddhist symbol scenario (Mean = 4.24; t(138) = −8.46, p < 0.01), suggesting that the manipulation check was successful.
Then, we conducted a sequential mediation analysis by the PROCESS macro for SPSS (Hayes 2015). By utilizing a bias-corrected confidence interval with bootstrapping (10,000 resamples), PROCESS could analyze both the corresponding direct and indirect effects. To be more specific, the PROCESS could prove the significant mediation effect when the confidence interval (CI) of the indirect effect does not contain zero. Figure 4 shows the result of the mediation analysis. According to the statistical results, 41.90% of the donation intention was explained in the current model. First of all, the results indicated that the donation box with the Buddhist symbol displayed a significant total effect on donation intention; effect = 0.800, p < 0.001. Then, we tried to explore whether the perceived religiosity and product attitude could, uniquely and sequentially, mediate the effect of donation boxes with the Buddhist symbol on donation intention. Table 1 illustrates all the mediation analyses of the current research model (both unique and sequential effect). In accordance with our main hypothesis, the results suggested a significant indirect effect (total indirect effect); effect = 1.088, 95% CI [0.64, 1.56]. After controlling the two mediators, however, the results showed an insignificant direct effect of donation box with the Buddhist symbol on donation intention; effect = −0.288, p = 0.277. According to the statistical results, 41.90% of the donation intention was explained in the current model. First of all, the results indicated that the donation box with the Buddhist symbol displayed a significant total effect on donation intention; effect = 0.800, p < 0.001. Then, we tried to explore whether the perceived religiosity and product attitude could, uniquely and sequentially, mediate the effect of donation boxes with the Buddhist symbol on donation intention. Table 1 illustrates all the mediation analyses of the current research model (both unique and sequential effect). In accordance with our main hypothesis, the results suggested a significant indirect effect (total indirect effect); effect = 1.088, 95% CI [0.64, 1.56]. After controlling the two mediators, however, the results showed an insignificant direct effect of donation box with the Buddhist symbol on donation intention; effect = −0.288, p = 0.277. Moreover, we explicitly evaluated each indirect effect on our research framework. First of all, the results indicated that perceived religiosity did not uniquely (independent of product attitude) mediated the effect of donation box Buddhist symbol on donation intention; effect = 0.049, 95% CI [−0.066, 0.183]. Supporting H1, the donation box with the Buddhist symbol would improve the perceived religiosity; effect = 0.485, p < 0.001. To be more specific, t-test results show that the donation box with the Buddhist symbol (Mean = 6.77, SD = 1.31) would raise perceived religiosity higher than the donation box without Buddhist symbol (Mean = 6.29, SD = 1.52; t(138) = −2.02, p < 0.05). However, not supporting H4, the perceived religiosity consecutively had an insignificant impact on donation intention; effect = 0.102, p = 0.258. Accordingly, the perceived religiosity did not play a unique mediating role in explaining the effect of donation boxes with Buddhist symbols on donation intention.
Independent of the perceived religiosity, product attitude showed a significant mediating effect of the donation box with Buddhist symbols on people's donation intention; effect = 0.906, 95% CI [0.533, 1.342]. Supporting H2 and H5, the donation box with Buddhist symbolic decorations would promote people's product attitude; effect = 1.508, p < 0.001. Particularly, people would have a more positive product attitude towards a donation box with a Buddhist symbol than a donation box without a Buddhist symbol (Mean = 7.36 vs. 5.63, SD = 1.13 vs. 2.19; t(138) = 34.32, p < 0.05). Product attitude, in turn, fed uniquely into donation intention; effect = 0.601, p < 0.001. Accordingly, product attitude played a simple mediating role in explaining the effect of donation boxes with Buddhist symbols on donation intention.
Finally, we examined whether perceived religiosity and product attitude could play a sequential mediating role in the relationship between the donation box with Buddhist symbols and donation intention. Supporting H6, results suggested that the donation box with Buddhist symbols significantly contributes to the perceived religiosity; effect = 0.485, p < 0.001. Consecutively, the perceived religiosity positively increased the level of product attitude; effect = 0.452, p < 0.001. This finally predicted donation intention; effect = 0.601, p < 0.001. More specifically, people would have a higher donation intention towards the donation box with Buddhist symbols, rather than the donation box without Buddhist symbols (Mean = 6.40 vs. 5.60, SD = 1.41 vs. 2.08; t(138) = 7.06, p < 0.05). Thus, H3 was supported. The sequential mediation was statistically significant; effect = 0.132, 95% CI (0.004, 0.336). It confirmed that the donation box with Buddhist symbols has an indirect and positive impact on donation intention, sequentially mediated initially through the perceived religiosity and then through product attitude.

Discussion, Limitation and Future Research
The current research tries to examine the effect of product design, notably the donation box with Buddhist symbols, on people's donation behavior and the underlying mechanism. Concerning the slow annual growth of charitable donations in the last decade, it is necessary to address and improve charitable donation intentions towards the needy population. Although some studies have argued for the positive effect of religious symbols in advertising on people's behavioral reactions (Muralidharan et al. 2018), limited prior research has addressed the impact of touchable product design with Buddhist symbols on people's charitable donation intentions. This study both contributes to the charitable donation literature and research on Buddhist symbols from an empirical perspective since it extends the scope of previous researches which are mainly discussed in the context of advertising and have seldom discussed the Buddhist determinants of charitable donation intention from a product semantics perspective.
In order to address this research gap, the current study statistically shows that product design, such as a donation box with Buddhist symbols, could significantly improve people's charitable donation intention compared with one without Buddhist symbols. Moreover, the underlying mechanism of this process is mediated through two sequential factors, namely the perceived religiosity and product attitude. The results of this research are summarized as follows.
First of all, after taking into account product attitude, the perceived religiosity did not uniquely mediate the effect of donation boxes with Buddhist symbols on donation intention. This fact might have a possible explanation: both perceptions of Buddhist symbols and religiosity belong to the cognitive perception level while charitable donation intention belongs to the behavioral intention level (Ajzen 1991). Consequently, the perception of Buddhist symbols is naturally closer to the perceived religiosity compared to donation intention, resulting in the insignificant indirect path between the perceived religiosity and charitable donation intention. It is consistent with previous research that religious symbols and perceived religiosity are strongly associated with each other since the religious symbols work as an iconicity of a specific religion, which elicits a cognitive perception of the symbolic spirituality (Hallman 2006;Tillich 1958). However, unlike symbols and perceived religiosity, the charitable donation intention is the result of cognitive perceptions. The association between the perceived religiosity and donation intention is somehow weaker than the association between the perception of symbols and the perceived religiosity.
Second, product attitude uniquely mediated the effect of donation boxes with Buddhist symbols on charitable donation intention. In this case, Buddhist symbols presented on product appearance could significantly improve product attitude, thus raising charitable donation intention. It is consistent with the prior finding that product design with religious symbols could play a valid role in eliciting emotional responses (Jiang 2010;Yao and Ming 2015). Furthermore, it further extends Martin and Randal's (2008) research on the relationship between the design of the donation box and donation intention, from the perspective of religious symbols.
Finally, the impact of a donation box with Buddhist symbols on charitable donation intention is significantly mediated by the perceived religiosity and then by product attitude. This finding, for the first time, provides preliminary evidence to show the mechanism of how product design with religious symbols influences individual donation intention by a sequential mediating model in terms of Buddhism. Based on the theory of reasoned action (Fishbein and Ajzen 2011;Sheppard et al. 2002), an individual would firstly hold cognitive perceptions, then gradually form and shape cognitive attitude, and finally, have an intention to perform a specific behavior (Ajzen 1991). Since the perception of Buddhist symbols presented on the product, the perceived religiosity, and the product attitude all belong to cognition level (both cognitive perception and cognitive attitude), this explains that the perceived religiosity significantly contributes to product attitude. Moreover, the donation intention is motivated by cognitive attitudes towards Buddhist belief and product design, which is consistent with prior research on Buddhism-informed prosocial behavior (Clobert et al. 2015;Pace 2013).
This research has some practical implications. Regarding the current situation in charitable donations, Buddhist temples, religious institutes or non-profit organizations could incorporate appropriate religion-related symbols into a product or installation design to increase people's charitable donation intentions and other charitable behavior intentions.
There are some limitations to the present research. Firstly, there might be other antecedents of charitable donation intention to be investigated. For example, attitude towards charitable organizations is also crucial in determining donation intention (Ranganathan and Henley 2008). Future studies would incorporate other factors to give a holistic picture of charitable donation intention. Secondly, the current study uses a universal Buddhist symbol, "lotus", in product design. However, there are many other Buddhist symbols, such as the "Buddhist Flower Gesture" (Ghori and Chung 2007). It would be rather interesting to explore the difference in donation intentions under the effects of different Buddhist symbols.
Author Contributions: Data curation, Y.S.; Formal analysis, Z.Q. and Y.S.; Investigation, Z.Q. and Y.S.; Methodology, Y.S.; Writing-original draft, Z.Q. and Y.S. All authors have read and agreed to the published version of the manuscript. Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.