I Was Thinking to Help but Then I Changed My Mind! The Influence of Injunctive and Descriptive Norms on the Donation Intention-behavior Link

ABSTRACT Public donations represent a substantial part of a charitable organization’s income. Thus, it is important to understand what influences donation behavior. Researchers mainly focus on intentions, rather than behavior. Although intentions often lead to behavior, there might be a gap between them. This experimental study investigated whether and how aligned and unaligned injunctive (what the majority of people approve of) and descriptive (what the majority of people do) social norms moderate the intention-behavior link. Additionally, I checked whether personal involvement moderates the influence of social norms. Four hundred and twenty-eight respondents were randomly assigned to one of the five conditions where the injunctive and descriptive norms were manipulated. Results revealed a significant and positive relationship between intentions and behavior (p = .000); this relationship is moderated by both aligned (supportive injunctive and descriptive) (p = .01) and unaligned (unsupportive injunctive and supportive descriptive) norms (p = .037). Personal involvement does not affect the moderating power of social norms. Findings are important for better understanding the intentions-behavior-norms dynamic. This can be used to create effective marketing strategies to make sure both favorable and unfavorable intentions translate into monetary donations.


Introduction
Kevin is a 25-year old student and works part-time at a gym reception. One day, when scrolling through his Facebook feed, he sees a post asking for donations to an orphanage in Uganda. He is so moved by the pictures that he decides to donate 50 pounds from his upcoming salary. Additionally, he tags his Facebook friends in the post. Most of them agree with Kevin´s good intentions and comment on how important it is to help others, but ultimately, his friends decide not to donate for various reasons. After learning how his friends behave, Kevin starts to reconsider his own donation intentions. Will Kevin act as initially planned?
To understand how different stimuli activate behavior, researchers use the Theory of Planned Behavior (TPB; Ajzen, 1985), which follows the logic that intentions inform behavior. However, Ajzen (1985) suggests that intentions are not always actualized, finding that "some are abandoned altogether while others are revisited to fit changing circumstances" (p. 11). One of these "changing circumstances" can be a revelation of social norms (Carrington et al., 2010;Fishbein & Ajzen, 2009), which are considered either injunctive (what the majority of people approve of) or descriptive (what the majority of people do). In general, injunctive norms can motivate behavior as noncompliance often elicits social disapproval. On the other hand, descriptive norms can influence people as they suggest an effective and appropriate course of action (Cialdini et al., 1990). The independent influence of injunctive and descriptive norms has been empirically demonstrated (Lönnqvist et al., 2009;Nipedal et al., 2010). In reality, though, both of the norms may be known and might even contradict each other (just like in Kevin´s story; Smith et al., 2012).
It is yet to be determined whether individuals stick to their intentions and behave accordingly in the presence of aligned or unaligned social norms.
The literature on the donation intention-behavior dynamics and the influence of social norms on donations is limited (Agerström et al., 2016;Herath, 2013). To my knowledge, none of the studies have yet addressed how the interaction of norms can affect this link. Little research on the simultaneous effect of injunctive and descriptive norms mostly addresses environmental or health-related issues (Rimal & Real, 2003;Smith & Louis, 2008;Smith et al., 2012). These studies focus on intentions rather than actual behavior and provide conflicting findings (e.g., Rimal & Real, 2003;Smith et al., 2012). Smith and Louis (2008) encourage scholars to examine the interaction between descriptive and injunctive norms and their effect across a range of domains. How the two sets of norms interact is especially important to charitable organizations because public donations represent a substantial part of income for the majority of them Ling, 2012) and whether the public donates is largely dependent on social norms. The topic becomes even more critical when scandals -such as the one involving the WE Charity in Canada -arise and taint the entire sector, drastically decreasing public donations (Murphy, 2020).
Addressing the gaps in our knowledge, in this experimental study, I propose that (a) donation intentions lead to donation behavior; (b) social norms moderate the intention-behavior link; (c) moderating the power of norms depends on whether injunctive and descriptive norms align with each other or not; and (d) personal involvement moderates the influence of social norms.
From a theoretical standpoint, this study contributes to bridging the donation intentions-behavior gap and understanding that aligned and unaligned norms can be used in this process. From a practical perspective, the study provides insight into how charitable organizations can employ social norms in their marketing strategies to harvest donations even from those individuals who had initially planned otherwise.
The remainder of this article is divided into the following sections: (2) literature review on how social norms influence donation intentions and behavior, (3) description of present research and hypothesis development, (4) results suggesting that intentions lead to behavior and both aligned and unaligned social norms can moderate this relationship, (5) discussion of the findings, and (6) conclusions with implications of the current study, limitations, and future research suggestions.

Intentions and behavior
The Theory of Reasoned Action (TRA) and its extension, the Theory of Planned Behavior (TPB) by Ajzen (1985) are widely used to study how stimuli activate intentions and behaviors . The theory follows the progression that (1) attitudes, subjective norms, and perceived behavioral control determine intentions, and (2) intentions inform behavior.
Intentions are self-instruction to perform particular actions (Sheeran & Webb, 2016). They are seen as the proxy to behavior-the higher intentions, the higher probability of behavior (Smith & McSweeney, 2007). Numerous correlational studies and meta-analyses (e.g., Armitage & Conner, 2001;Sheeran, 2002) reported high intention-behavior correlations Hassan et al., 2016). Moreover, according to these models, the effects of the antecedents of the TRA (attitude and subjective norms) and the TPB (attitude, subjective norms, and perceived behavioral control) on behavior are fully mediated by intentions (Hassan et al., 2016). Intentions also predict behavior better than other concepts, such as attitudes, norms, self-efficacy, and risk or severity perceptions (e.g., Sheeran et al., 2017Sheeran et al., , 2014 as well as personality factors (e.g., Chiaburu et al., 2011;Sheeran & Webb, 2016).
The TPB has been used in a variety of domains (e.g., Kashif & De Run, 2015;Van der Linden, 2011), but relatively fewer scholars have applied the model to charity (Mittelman & Rojas-Méndez, 2018). Most of this limited amount of studies investigated donation intentions rather than actual behavior. However, the ones that measured both intentions and behavior report a high correlation. For example, Smith and McSweeney (2007) conducted a longitudinal study, where they applied the TPB and measured donation intentions and actual donation behavior four weeks apart. They found that intentions were the only significant predictor of donation behavior. Moreover,  employed the TPB to study the monetary donation intentions and behavior of respondents in Kuala Lumpur. These researchers also found a positive and significant correlation between intentions and actual donation behavior. Based on the literature, I hypothesize: H1: Donation intentions will positively influence donation behavior.

Social norms and normative influence on donations
Even though intentions often predict behavior (Sheeran & Webb, 2016), this relationship does not always exist as intentions may be abandoned or revised to fit the changing circumstances (Ajzen, 1985). Researchers refer to this discrepancy as an intention-behavior gap (Parsatemijani, 2020). The first approach to studying the intention-behavior gap focuses on mediators and looks at the post-intention processes. The second approach addresses the variability of the relationship and suggests the possible moderators (Sniehotta et al., 2005). Employing the second approach in this study, I propose that social norms moderate the intention-behavior link.
As group-based standards, social norms inform what type of attitudes and behaviors are common, valued, and accepted within a specific society (McDonald & Crandall, 2015;Women´s Philanthropy Institute, 2018).
The Social Norms Approach (SNA; Perkins & Berkowitz, 1986) categorizes norms into injunctive and descriptive, each of which has independent and distinct influences (Manning, 2009;Rimal & Real, 2003). Injunctive norms impose social pressure to act in a certain manner (Cialdini et al., 1990). For instance, Raihani et al. (2014) demonstrated that in a Dictator Game, players share more money when they are signaled what they should do. On the other hand, descriptive norms suggest effective or adaptive behavior in a specific situation (Cialdini et al., 1990). Empirical evidence about the effect that descriptive norms have on donation intentions and behavior is inconclusive. For example, Martin and Randal (2008) explained that donation boxes displaying higher bills had larger average donation amounts. In addition, Croson et al. (2009), and McAuliffe et al. (2017 found that people donated larger sums when learning about the large donations of other donors. Contrarily, Raihani et al. (2014) did not discover the effect of descriptive norms on charitable giving. Moreover, Shang and Croson (2009) uncovered that only new donors who learned about the large sum of other donations were influenced by descriptive norms. Based on the literature, I hypothesize: H2: Social norms will moderate the relationship between donation intentions and donation behavior.

Aligned and unaligned social norms
At the most basic level, the highest probability of translating intentions into behavior is when people believe that others approve of that behavior and commonly perform it. Yet how do we behave when we form intentions and then find out that people fail to match what they say and what they do? In other words, how does the combination of injunctive and descriptive norms moderate the relationship between intentions and behaviors? Researchers have not yet addressed this question in regards to charity. Rather than studying the interactions of two sets of norms, scholars either imply the primacy of injunctive norms (Cialdini et al., 2006) or aggregate the influence of injunctive and descriptive norms (Smith et al., 2012). The limited amount of studies that do look at the interactions of norms, in other domains than charity, mostly focus on intentions rather than actual behavior (Smith et al., 2012) or intention-behavior dynamics (Smith & Louis, 2008). These studies also provide conflicting findings (e.g., Rimal & Real, 2003;Smith et al., 2012). Some suggest that unaligned or conflicting norms may de-motivate behavior by signaling that it is acceptable not to act according to either set of norms. Others state unaligned or conflicting norms act as motivators by highlighting the discrepancy between what people are doing and what they should be doing (Smith et al., 2012). Smith et al. (2012) manipulated injunctive and descriptive norms toward pro-environmental behavior and found that conflicting norms led to weaker intentions to act in an environmentally friendly manner. In another study, Smith and Louis (2008) manipulated peers´ injunctive and descriptive norms and measured students´ attitudes, behavioral willingness, and actual behaviors toward various campus issues. They obtained a negative influence of norm conflict in their first study and a positive influence in the second one. Researchers suggested that their inconclusive findings might have been caused by the fact that the targeted attitude was more important for the respondents in the first study than in the second one (as cited in Cestac et al., 2014). Rimal and Real (2003) also found a positive influence of norm conflict on alcohol consumption among Texan students. Those perceiving social disapproval of alcohol use but frequent consumption by peers were the ones most likely to consume alcohol themselves. They explained the phenomenon as a function of rebellion against the social norms. Furthermore, Cestac et al. (2014) found that the respondents who received conflicting norms about speeding when driving had the highest speeding intentions.
Thus, based on past research, when social norms align their influence aggregate, generating pressure and suggesting effective behavior. Therefore, I hypothesize: H3a: Aligned social norms will moderate the influence of donation intentions on donation behavior.
On the other hand, when social norms unalign, it causes doubt about whether the normative information can be used as a reference. Thus: H3b: Unaligned social norms will not moderate the influence of donation intentions on donation behavior.

Personal involvement
Individuals are personally involved in an issue to the extent that it holds personal meaning to them (Sherif et al., 1973). Highly involved individuals perceive the focal issue as critical (Thomsen et al., 1995). Thomsen et al. (1995) referred to the Social judgment Theory (SJT) when studying attitude change and suggested the significant association between individuals´ level of involvement in an issue and their likelihood of acting in an attitude-consistent fashion. Authors proposed that attitudes of highly involved individuals are more stable over time and resistant to change by external factors. The reason is that highly involved individuals have more knowledge, as they put more effort into searching for relevant information and engage in scrutiny (Kim, 2008), critical analysis, and elaboration of this information (Walia et al., 2016). Such individuals also carefully consider the situational cues and consequences of their actions (Thomsen et al., 1995).
Based on the literature, I believe that intentions formed by careful judgment and deliberation are long-lasting and less affected by external information, such as social norms. Therefore, they should be more powerful in directing behaviors than the ones formed lightheartedly and without careful consideration. Thus: H4: Personal involvement will negatively moderate the influence of social norms on the relationship between donation intentions and behavior.
In summary, based on literature, I expect that norms and personal involvement will moderate the relationship between donation intentions and donation behavior such that those, who score high on donation intentions and perceive the social norms as supportive and aligned and hold high personal involvement, will be most likely to donate money.

The present research
The Figure 1 below illustrates the conceptual model of the study.

Participants and design
Five groups of a total of 428 respondents participated in the study. They were recruited through a crowdsourcing marketplace, www.prolific.ac (ProA), and received an hourly compensation of £7.5. All of the respondents were over the age of 18. The pool of potential participants was filtered to meet the following criteria: (1) nationality-British, (2) residence-Great Britain, and (3) language proficiency-fluent in English. Great Britain was chosen based on practical reasons: no translation services were needed for the experiment materials and ProA has a large pool of British respondents.
Repeat submissions from individuals were prevented by using the survey publishing tool, SurveyMonkey.
Thirty-six percent of respondents were male and the average age was 35 (min. = 18; max. = 82). There was no significant difference in gender (p = .335) or age (p = .423) between the five experimental groups.

Procedure
The procedure was adopted from Smith and Louis's (2008) study. Participants were assigned randomly to one of the five experimental conditions. To prevent them from matching their behavior to previously stated intentions, they were informed that the survey included a series of short questionnaires for different researchers within the School of Psychology. First, the data on behavioral intentions, personal involvement, and demographics were collected.
On the second page, injunctive and descriptive norms were manipulated by presenting the abstract, results, and pie-chart illustrations of the ostensible study (see section 3.2.3).
Afterward, participants were asked to summarize the study's findings. The manipulation check followed.  On the third page, participants completed a "destruction task" that asked them to describe emotional reactions to their job. This task was included to bolster the perception that the survey encompassed the questionnaires from multiple unrelated studies.
Participants were then taken to a page that was deliberately designed to look like a section for collecting the technical information, such as ProA IDs. At the end of this page, actual donations were measured.
Next, participants were asked to guess the hypotheses underlying the study. Nineteen participants who either guessed the hypotheses correctly, indicated they did not believe the abstract and charts depicted a real study, or did not complete the survey were excluded from the analysis.
Finally, participants were debriefed and allowed to withdraw their data but no one did.

Manipulations and manipulation checks
Manipulation occurred by providing participants with the abstract and results of an ostensible 2019 study on animal welfare by the Office for National Statistics UK (see Appendix 1). I manipulated the injunctive and descriptive norms by stating the percentages of British people who, according to the study, supported donating money to animal welfare organizations (injunctive norms) and who actually donated money (descriptive norms).
To identify a charitable cause that would produce varied results in terms of personal involvement, I conducted an extensive pretest. Consequently, animal welfare was deemed as a suitable topic for this study. By percentages, I manipulated injunctive and descriptive at the same level of specificity. These specific percentages were chosen in line with the past research (e.g., Smith et al., 2012). Additionally, I pretested the percentages to make sure they were believable, still having a strong enough effect. Together with the abstract, participants viewed two pie charts illustrating the results. This increased the manipulative power of the stimuli. After reading the text and checking the charts, participants were asked to summarize the findings of the ostensible study. This was followed by the manipulation check. Table 1. provides information on groups and manipulations.

Measures
All measures except for donation behavior used a 7-point response format (e.g., 1 = strongly disagree/ 7 = strongly agree). To reduce the response bias, some items were negatively worded and then reverse-scored. I averaged the items of the same variable to form an overall index.

Donation intentions
were measured with five items (e.g., I will donate money to animal welfare organizations in the next 4 weeks) (α = 0.86; Ajzen, 2002).
For measuring the donation behavior, participants were informed that five randomly chosen participants would receive a bonus of £50. Each participant, regardless of their answers, had an equal chance of winning the bonus. They could keep the bonus for themselves or donate all or parts of it to one of our partner animal welfare organizations or an organization of their choice. Those who did not want to donate could leave the donation tab empty or type 0. Others could type the donation amount. For analysis purposes, I coded 0-s and empty spaces as nos and any other amounts as yeses.
Manipulation checks were done using six items: three to evaluate injunctive norms (e.g., Would you say that the majority of British people approve of those who donate to animal welfare organizations?) (α = 0.9) and three to evaluate descriptive norms (e.g., Would you say that the majority of British people donate to animal welfare organizations?) (α = 0.96; Ajzen, 2002).

Manipulation Check
The analysis was done using the IBM SPSS Statistics 25.
Analysis of perceptions of the group injunctive norms using the Independent Samples T-test revealed a significant effect from the manipulation of injunctive norms (p = .000). Participants exposed to a supportive injunctive group norm were more likely to report that the majority of people in Great Britain approve of donating money to animal welfare organizations (M = 6, SD = 1.1) than participants exposed to an unsupportive injunctive group norm (M = 2.5, SD = 1.5). Analysis of perceptions of the group descriptive norms also revealed a significant effect from the manipulation (p = .000). Participants exposed to a supportive descriptive group norm were more likely to report that the majority of people in Great Britain donate money to animal welfare organizations (M = 5.8, SD = 1.3) than participants exposed to an unsupportive injunctive group norm (M = 1.8, SD = .93).

Hypotheses Testing
I hypothesized that donation intentions positively influence donation behavior (H1) and social norms moderate this relationship (H2). To test these hypotheses, I performed the logistic regression, where Group 5 served as a reference category.
The linearity of the continuous independent variables with respect to the logit of the dependent variable was checked via the procedure of Box-Tidwell. According to the Bonferroni correlation, the acceptable statistical significance level was calculated to be p < .0083 (Tabachnick et al., 2007). Thus, the continuous independent variables were found to be linearly related to the logit of the dependent variable. Multicollinearity of the independent variables was assessed via Collinearity Diagnostics. Based on the obtained VIF values of 1.054, 1.092, and 1.038, I concluded that there were no multicollinearity symptoms. There were two cases with standardized residuals of −3.223 and −3.509. The evaluation of these respondents uncovered no measurement or data entry errors. Thus, I deemed these as natural variations and decided to keep these data points.
The Chi-square model test showed that the logistic regression model was statistically significant (p = .000). This means that the prediction model fits significantly better with the data than the null model with no predictors. The model explained roughly 19% (Nagelkerke R 2 ) of the variance in actual donation behavior and correctly classified 68% of cases. Sensitivity was 74.1% and specificity was 57.1%.
The Hosmer and Lemeshow test indicated that there was not enough evidence to conclude that the model did not fit the data (p = .686). In other words, the null hypothesis of a good model fit to data was tenable.
The analysis showed that the odds of donating money to animal welfare organizations were positively related to donation intentions. The relationship was statistically significant (p = .000). Thus, I concluded that changes in the donation intentions were associated with the changes in the probability that the respondent would actually donate money. In other words, the higher the donation intentions, the more likely it is that a person will actually donate money to a charitable organization. The results indicated a significant interaction between donation intentions and social norms (p = .043). Thus, I reject the null and accept the alternative hypotheses 1 and 2 that donation intentions positively influence donation behavior and their relationship is moderated by social norms.
Hypothesis 3 states that only aligned social norms (both injunctive and descriptive norms being either supportive or unsupportive) moderate the donation intention-behavior link. In other words, I expected that aligned social norms would guide the behavior, regardless of previously stated donation intentions. On the other hand, unaligned social norms (one set of social norms being supportive, while the other one is unsupportive) would cause doubts on whether the norms could be used as a reference. In these cases, the donation intentions direct donation behavior.
The logistic regression of the interaction between donation intentions and social norms in each group separately revealed the negative interaction between donation intentions and social norms in all four groups. Thus, all conditions weakened the impact of donation intentions on actual behavior. This interaction was also statistically significant in groups 1 (p = .01) and 4 (p = .037), but not in groups 2 (p = .575) and 3 (p = .691).
Based on these results, hypotheses 3a and 3b are both partially supported; both aligned and unaligned social norms can moderate the relationship between intentions and behavior, but not in all cases. When aligned norms are unsupportive (group 3) and injunctive are supportive but descriptive are not (group 2), social norms do not moderate the intention-behavior link.
Finally, H4 states that personal involvement negatively moderates the influence of social norms. Testing H4 with logistic regression did not show a significant three-way interaction between donation intentions, norms, and personal involvement (p = .669). Thus, I failed to reject the null hypothesis 4. Table 2 provides the details of the logistic regression.

Discussion
Public donations are of utmost importance to charitable organizations; thus, it is of utmost importance to understand the factors influencing donation behavior Ling, 2012). Studying the topic, researchers mainly focus on intentions, rather than actual behavior. Although intentions often lead to certain behaviors, there might be a gap between them as intentions might be biased and not always acted upon (Ajzen, 1985). Among the possible moderators of the intention-behavior link are injunctive and descriptive social norms (Carrington et al., 2010;Fishbein & Ajzen, 2009), which may or may not align with each other. Researchers have not fully understood the influence of norms on charitable giving (Agerström et al., 2016). How two sets of norms interact and moderate donation intention-behavior relationship has received even less attention. It is critical for scholars studying intention-behavior-norms dynamics to gain more   knowledge about this issue (Agerström et al., 2016;Herath, 2013). Moreover, charities use this information to target actual behavior and increase the likelihood that both favorable and unfavorable intentions translate into monetary donations. In this experiment, I sought to understand whether and how donation intentions influence donation behavior and how different norm conditions moderate this relationship. I also checked whether personal involvement moderates the influence of norms on the intention-behavior link.
First, the results of this experiment demonstrate a strong positive correlation between donation intentions and behavior. Higher intentions did lead to higher donation behavior. This finding is consistent with previous studies by Smith and McSweeney (2007) and Kashif and De Run (2015).
Second, I found that the intention-behavior relationship may change as a function of aligned and unaligned injunctive and descriptive norms. Specifically, norms significantly decreased the intention-behavior link in two conditions where (I) both injunctive and descriptive norms were supportive of donating money, and (II) injunctive norms were unsupportive but descriptive norms were supportive. I did not find the moderating effect of social norms in conditions where (I) injunctive norms were supportive but descriptive norms were unsupportive, and (II) both sets of norms were unsupportive. The possible explanation for these results is that the pressure of injunctive norms was higher when others performed the behavior (when descriptive norms were supportive). On the other hand, when others did not perform the behavior (when descriptive norms were unsupportive) the pressure of injunctive norms was lifted, thus individuals behaved according to their intentions. To the best of my knowledge, this is the first study showing how aligned and unaligned injunctive and descriptive norms moderate an intention-behavior link in the domain of charity.
Unlike Smith and Louis (2008), this study did not uncover the moderating effect of personal involvement. Thus, I am unable to suggest whether personal involvement can serve as a moderator for the influence of social norms on the intention-behavior link. The possible reason for the insignificant result of the three-way interaction is the charitable cause itself: animal welfare. The respondents did differ in terms of their involvement with an issue but they might have perceived that the issue did not have immediate and significant consequences for their lives. Moreover, animal welfare is a relatively less critical issue in Great Britain than some other issues, such as cancer, environmental protection, and homelessness. These reasons could have resulted in the less deliberate formation of intentions, which are easily influenced by external factors, such as social norms (Kim, 2008;Walia et al., 2016).

Implications
This study contributes to the literature on charitable giving, intentionbehavior gap, and social norms. To the best of my knowledge, this is the first study that considers norms as a moderator of the intention-behavior link and orthogonally manipulates the injunctive and descriptive norms to demonstrate the moderating power of both aligned and unaligned norms. Thus, the results of this study further the understanding of whether and when intentions relate to behavior and how social norms can moderate this relationship. This information is critical for bridging intentions and behavior by incorporating injunctive and descriptive social norms in the TPB.
In addition to the theoretical contribution, this study has practical implications. Specifically, the findings are useful for charitable organizations wishing to effectively use norms to increase public donations. Consistent with previous literature (e.g., Cialdini et al., 2006;Smith et al., 2012), these results indicate that norms can be used to motivate certain behavior; in this, individuals who previously intended not to donate are changing their minds. Therefore, I recommend charitable organizations highlight the information on social norms on their websites, social media platforms, and printed materials. This can be done by indicating what percentage of people support the cause or by highlighting the supportive comments and testimonials from the public. Moreover, I found that aligned supportive norms are the most powerful moderators of the intention-behavior link. Therefore, it may be ineffective for charities to make unsupportive norms salient, in hope of increasing the empathy or obligation of personal action.

Limitations and future research suggestions
Despite theoretical and practical implications, this study has several limitations, which further research can address. First, I attempted to lead participants to believe that the experimental session involved a series of questionnaires from various studies. Additionally, the experiment included a destruction task and the results of those participants who correctly guessed the hypotheses were excluded from the analysis. Still, it is possible that previously stated intentions had a spillover effect on donation behavior. Thus, future researchers should conduct a longitudinal study, where donation intentions and behaviors will be measured at least three weeks apart.
Second, in this experiment, I only included one source of norms: in-group strangers. I encourage future investigators to study whether the moderating power of norms in the donation intention-behavior link varies based on whether the source of the norms is in-group strangers, out-group strangers, or people important to the respondents. Third, I measured the donation behavior by asking respondents whether they wished to donate the bonus. Since the respondents were unsure whether they would win the bonus or not, they might have made their donation decisions without careful deliberation. Future researchers should measure donation behavior by asking respondents to make monetary donations out of their own pockets.
Finally, in this study, I used animal welfare, which I believe is the reason why personal involvement did not interact with social norms and donation intentions. Future researchers should study the phenomenon by looking at other social issues, such as health issues, environmental protection, or homelessness.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Availability of data and material
The data that support the findings of this study are available from the corresponding author upon reasonable request.