Helping relief campaigns: Social cognitive factors underlying important issues using structural equation modelling

Abstract Natural disasters happening around the world have caused a huge burden on local and international charities in recent years. This study tended to examine and explore factors, which cause the migration of people to charitable behaviour based on Bandura’s social cognitive theory. By studying the Bandura’s social cognitive theory and the research background the conceptual model has been developed. The 199 members of Tehran’s Red Crescent Society was selected by stratified sampling and questionnaires were distributed. The validity and reliability of research were confirmed by confirmatory factor analysis and Cronbach’s α. The conceptual model was tested using structural equation modelling (SEM). The SEM results showed that perceived self-efficacy, outcome expectancies for donating to disaster campaign and intention to donate to charity has a positive direct effect on intention to donate a disaster campaign. Moreover, news exposure had no significant effect on intention to donate to a disaster campaign. Finally, results revealed that indirect insignificant effect of individualistic explanation of poverty, and awareness of a disaster on intention to donate to disaster campaign. Practical implications and limitations are subsequently discussed in light of the results.


PUBLIC INTEREST STATEMENT
Natural disasters that are happening around the world have caused a huge burden on local and international charities in recent years. This perspective article describes which factors because donors turn to giving behaviour, based on data gathered via polls, which was used to illustrate impacts of the disaster. Clearly, large-scale disasters have a profound effect on the physical and emotional health of disaster survivors. Understanding these factors can improve future disaster relief programs and policies. Designing and evaluating programs for enhancing the audience's beliefs in giving behaviours are the next step in research on donating. Exploration of post-disaster issues can also help government and non-government organizations to assist communities and individuals left in the aftermath of natural disasters.

Introduction
It is very important to identify and evaluate factors which promote public donation to charitable organizations; however, it is not planned yet. Different people have different perceptions of non-profit organizations (Awan & Hameed, 2014;Meijer, 2009). Hence, it is very important to identify those characteristics of donors which are associated with the tendency to donate (Awan & Hameed, 2014). In the first phase, public donation to charities is very important for providing financial support for poor people in local and international areas and social support in modern societies (Eayrs & Ellis, 1990). NGOs, voluntary charity organizations, either with a religious affiliation or not, emerge to play a key role in the philanthropy enterprise. Some of them, notably the Red Cross, Christian organization, Oxfam, United Nations Foundation for Children and many others offer assistance and aid to people. They raise their funds across national boundaries on an international basis. While they conspicuously provide relief to victims of disasters and haphazard events, they are also involved in improving the quality of life of poor, disadvantaged, and disabled people perennially through direct assistance or indirect means which ameliorate structural conditions, such as water and electricity facilities. Regularly, charities also launch campaigns to advocate against harmful conditions such as landmines. In support of these various charity programs, raising public funds is a main concern of charities, particularly because of economic decline in industrialized countries (Yavas et al., 1993). As successful fundraising depends on public donation, an important social project is to determine and promote factors leading to donation to charities.
The charity-linked individual behaviour increasingly attracts the attention of scholars and practitioners. Traditional charity-linked individual behaviour, when people donate used items or money to charities, has changed considerably. This has forced charities to become more aggressive in findings new ways to attract and maintain both donor and volunteer support.
In Hong Kong, both local charities and international relief organizations (IRO) are active in raising public funds. In Malaysia, there are three main ethnic groups, Malay, Chinese and Indian with a total population of 15, 6 and 2 million people, respectively. A promotional show was launched on television about twice a month by a certain charitable organization to collect public donations. It is noteworthy that a television campaign managed to collect millions of US dollars per year for a local hospital group. Undoubtedly, donation to charities has become an essential part of daily life of people in Hong Kong.
Donation is also pivotal to research on helping behaviour, which has remained a central concern in social science (Clary & Snyder, 1991). Helping and prosocial behaviour is ostensibly necessary for survival of human society. Presumably, it sustains concerns about altruism, morality, and well-being of the whole society. It is therefore an essential and recurrent topic of research. However, many studies on helping behaviour and donation, commonly conducted by students and volunteers in the laboratory (Abeler & Nosenzo, 2015;Exley, 2017;Serow, 1991;Thornton et al., 1991), substantially depart from real situations (Clary & Snyder, 1991;Wills, 1992). Thus, little is known about effective factors on public donation to charities, although this donation is common, realistic and socially significant. Furthermore, public donation to charities is considerably different from donation and helping behaviour in the laboratory context in which donation is nonspontaneous and discretionary (Clary & Snyder, 1991). In other words, it is unlikely an impulsive reaction to emergency and arousal. People often donate to charities with no knowledge that who will benefit from the donation. In the presence of multiple charity organizations, people are more likely to choose to donate to a charity. Therefore, charities should market their own distinctive images to attract different people (Eayrs & Ellis, 1990). Their different marketing efforts help promote deliberation and reasoning of people prior to donation (Borgida et al., 1992). In planning for these efforts, organizations should be aware of the factors which contribute to public donation. The present study examines the relationship between various factors underlying charitable behaviour and tends to develop a causal model. This study tends to answer this question: Do donating attitude and news exposure influence the intention to donate? Can these factors be added to the causal model of Cheung and Chan (2000) to increase variance explained of the intention to donate to a disaster relief campaign?

Intention to donate, perceived outcome and self-efficacy
In many Attitudes, personality, and behaviour (TPB) studies, behaviour intention is taken as a proxy measure of likely behaviour (e.g., Chevance et al., 2017;Nonis & Swift, 2010;Nugroho et al., 2018;Phillips & Jang, 2012;Sparks & Pan, 2009;Wang & Ritchie, 2012a, 2012b. Antecedents of intentions are better understood than antecedents of behaviour (Phillips & Jang, 2012). Intention is defined as a central condition of behaviour in TPB (Ajzen, 1988). Intentions supposedly hold motivational factors which influence behaviour and indicate the extent to which people tend to perform the behaviour (Ajzen, 1991). Ajzen (2002a) argues that, in general, the stronger the intention to behave in a certain way, the more likely the behaviour is to be performed. Intentions can therefore be useful in understanding behaviour. Because this study tends to understand donating behaviour and actual donating behaviour itself is hard to examine, the focus is on intention to donate to disaster relief campaigns and determining factors effective on intention. As TPB suggests, likelihood of engaging in a behaviour depends on strength of one's intention. Self-efficacy is widely known as one of the strongest determinants of participation in physical activity (Li et al., 2018;Rhodes & Nigg, 2011;Warner & French, 2018). The belief of people in their capabilities to reach a desired goal by their actions is called perceived self-efficacy, which plays a central role in SCT (Bandura, 1997). In the context of charity-linked behaviour, self-efficacy is associated with personal trust to the organization and whether one feels the need for donation to charity. On the other hand, self-efficacy is associated with environmental factors which often exert the pressure to donate for charity or the need for social recognition. In this study, self-efficacy is conceptualized as donation to disaster campaigns and victims of natural disasters despite the difficulty and pain which they suffer from. Another important construct which plays a key role in SCT is outcome expectancies (Bandura, 1997). Outcome expectancies, or outcome efficacy, encompass people's belief or perception on consequences of their actions (Schwarzer, 1992). Expected rewards increase the plausibility of certain behaviours, whereas expected punishments reduce the plausibility of certain actions (Bandura, 1986). Outcome expectancy may act as a mediator between selfefficacy and behaviour by increasing the likelihood of engaging in a behaviour. Outcome expectations can be physical, social, or self-evaluative in nature. Practically, they guide behavioural choices, as people adopt courses of action which are likely to result in positive outcomes. In this study, outcome efficacy is measured by pleasant experience of environment and being proud of donation.
A review of the literature shows that people are more likely to donate to charities if they feel valued for charitable giving and consider it to be appropriate to their beliefs and ideas. In other words, one's initial desire to help relief organizations or charities originates from one's intrinsic values, and these values promote one's self-efficacy-belief in one's values and abilities to assist in relief campaigns (Chevance et al., 2017;Nugroho et al., 2018). Indeed, one's perception of self-efficacy is a link between individual values and willingness to assist in relief campaigns or humanitarian activities. The findings of Clowes and Masser (2012) show that people's self-efficacy has a positive and significant effect on their willingness to engage in philanthropic and humanitarian activities. Similarly, the findings of Cheung and Chan (2000) suggest that self-efficacy is one of the important and vital cognitive factors to increase people's willingness to volunteer as well as to support disaster relief campaigns. Considering the literature, the first and second hypotheses are as follows: H 1 : Perceived self-efficacy of donation has a positive and direct effect on intention to donate to a disaster relief campaign. H2: Perceived outcome efficacy of donation has a positive and direct effect on intention to donate to a disaster relief campaign.

Moral obligation and intention to donate
Moral obligation can be defined as belief in significance of moral and altruistic conduct (Roats et al., 1995). When people act based on a moral code, their behaviour is formed by a set of effective rules on people obligation to each other (Wolfe, 1989). One's sense of what is "right" and "morally correct" consists of beliefs in personal norm related to a self-imposed obligation that people feel to "do the right thing" irrespective of what other people think. Moral obligation is operationalized as donation according to one's principles.
Moral commitment is actually a kind of moral motivation and commitment to values such as integrity, honesty, helping others, having a conscience and being altruistic in the harshest of conditions. Such a commitment makes people sensitive to what is happening and commits them to ethical action. In fact, being ethically committed enables community members to be more active in adhering to ethical norms and values, such as charitable donations, support campaigns, and voluntary associations (Winterich et al., 2009). In this context, Ortberg et al. (2001) found that the greater one's moral commitment to charitable activities, helping others, volunteering, participating in philanthropic and support campaigns, the greater the willingness to help disaster relief campaigns. Because moral commitment creates an intrinsic motivation and value and an intrinsic duty to help support campaigns. In addition, the findings of Ongley et al. (2014) also show that moral commitment is a very important factor in stimulating one's motivation for humanitarian activities and helping others in different areas. Moral commitment also enhances one's willingness to help or engage in humanitarian activities and campaigns. In fact, moral commitment first promotes one's willingness to donate to charity and subsequently enhances one's willingness to support disaster relief campaigns (Schwettmann, 2015). Accordingly, the third and fourth hypotheses are developed as follows: H 3 :Moral obligation to donate has a positive indirect effect on intention to donate to charity and thereby intention to donate to a disaster relief.
H 4: Intention to donate to charity is expected to have a positive direct effect on intention to donate to a disaster relief campaign. Rai et al. (2017) claim that understanding one's intention can influence donating action and it can also reinforce willingness to donate. Trust is an individual perception and trust level attributes to different characteristics of people. Besides, public trust plays a significant role in charities and charitable actions, because public trust as a moral voice advocates public commitment to donation and public trust contributes to improving charitable actions and secures the source of support from general public. Attitude or trust to charities is a practical and theoretical concern. Charities focus more on their image in public, rather than public attitude toward donation, in order to attain marketing advantage (Eayrs & Ellis, 1990). Trust in cooperative help organizations (CHOs) may lead to donation as an expression of support. In this study, trust in CHO is measured by good image and trust in the organization, cooperation of Dutch helping organizations, and help offered to victims of disasters.

Trust in CHO and intention to donate
Uncertainty, lack of proper planning, lack of urgent and critical requirements, severe waste of resources, additional costs, time delays in distributing items and facilities among victims, especially corruption and embezzlement due to uncontrollability and inadequate oversight are the most important causes of community mistrust in relief organizations and humanitarian campaigns (Powers & Yaros, 2013). Findings in this area indicate that one of the important factors in gaining public support for participating in relief campaigns and assisting peers is trust among actors as well as people's trust in public and humanitarian campaigns (Evers & Gesthuizen, 2011). In other words, the main motivation for people's willingness to contribute to humanitarian activities and campaigns is their trust in these campaigns. This trust is interpersonal, social, and institutional, the relationship space between individuals, with institutions and campaigns and relief organizations (Sargeant & Lee, 2004). Schwettmann (2015) show that people's trust in organizations and support campaigns has a direct and significant effect on their willingness to help these organizations. Accordingly, the following hypothesis is presented: H 5 : Trust in cooperative help organization has a positive indirect effect on intention to donate to a disaster relief campaign.
Cognitively, if one has a pleasant experience with helping campaigns or humanitarian organizations feel good about it and is more likely to contribute to disaster campaigns or relief organizations in the future. In other words, one's desire to help campaigns stems primarily from his pleasant experience (Kassie et al., 2020). When one feels that his previous help with these campaigns has been effective and has improved the situation, he will probably repeat his experiences in the future and encourage others to help. Many researchers, therefore, place greater emphasis on past experiences and consider it one of the important factors in one's willingness to contribute to disaster campaigns. In this regard, the findings of Kassie et al. (2020) show that one's desire for humanitarian activities and support of campaigns is influenced by one's past experience and attitude, and the more pleasant the previous experience, the more one's willingness to engage in charity. In addition, Hutchens et al. (2016) also found that prior experience had a direct and significant effect on willingness to contribute to public campaigns.
H 6 : Past donation to disaster relief campaigns has positive direct effect on intention to donate to a disaster relief campaign.

Awareness of natural disaster, donating attitude and intention to donate
Natural disaster awareness refers to one's understanding of the conditions, consequences, and detrimental effects of disasters. Awareness of quantity and quality of this type gives the cognition through which one can achieve a real understanding of the phenomenon to be able to make correct and responsible decisions about them (Oosterhof et al., 2009). The right decision or supportive measures are directly associated with one's understanding of why and how and consequences of natural disasters (Morgan, 2006). In fact, if there is no awareness of natural disasters, one cannot expect people in a community to be sensitive or responsible (Sanner, 2001). In other words, awareness is the starting point for one's attitude and behaviour toward action or decision on natural disasters. Disaster awareness is often seen as one of the key prerequisites for a desire to help and philanthropic activities that motivate decision to help members of a community, as this awareness represents a key factor in recognizing disasters in public memory (Hassan et al., 2018). Attitude is a combination of beliefs and emotions that prepares a person to look at different phenomena in a positive or negative way. Having a positive attitude towards philanthropic activities and helping support campaigns plays an important role in supporting behaviours and enacting their thoughts and thus facilitates the prediction or guidance of future humanitarian behaviours (James & Zagefka, 2017). People's awareness and positive attitude to natural disasters is about assessing people's risk and their intention to make a decision to support and help with campaigns. Asgary and Penfold (2011) at the University of York, Canada, studied willingness to donate to victims of a hypothetical future earthquake disaster in Vancouver using a probability estimation method. Help, particularly funding, played a key role in relief and rehabilitation. Disaster response organizations rely heavily on public donations in order to respond to disaster relief operations. About 500 people from the Toronto area were interviewed to demonstrate their willingness to help in a hypothetical earthquake in Vancouver, British Columbia. It turned out that people tended to contribute an average of 570.33 USD to disaster victims. Various factors, including past behaviour, donation, helping attitude, willingness to help fellow Canadians, and age were found to have significant effects on willingness to help.
The findings of Hassan et al. (2018) and James and Zagefka (2017) show that knowledge and attitude make people more likely to be active or to contribute to humanitarian and relief campaigns. The following hypotheses are based on the arguments raised: H 7 : Awareness of natural disaster is expected to have a positive indirect effect on intention to donate to a disaster relief campaign.
H 8 : Donating attitude is expected to have a positive direct effect on intention to donate to charity and have weak indirect effect on intention to donate to a disaster relief campaign.

Individualistic explanation of poverty, perceived need for donation and intention to donate
Individualistic explanation of poverty is a kind of perception of the phenomenon that shapes one's inner values. One's interpretation of a disaster will shape one's behaviour in the future. In other words, one's interpretation of a natural disaster is influenced by individual thoughts and beliefs that derive from one's personal values, preferences, interests, and knowledge. Individual interpretation is the process by which everyone categorizes and identifies perceptions about surrounding and in this way, gives meaning to them (De Wit et al., 2018). Perceived need for donation means the perceived need to utilize individual or group capacities of stakeholders to achieve a humanitarian group goal. In this process, conscious behaviour, collective desire, collective acceptance, choice and shared needs are important (Lundberg et al., 2019). What is important in a successful participatory process is the need to solve a problem, to recognize that problem, and to feel the need for teamwork, given the extent of knowledge and capability of people and their knowledge of existing capabilities and how to take advantage of them. Experimental findings of Iwannanda and Adiputra (2017) show that individual interpretation and perception of helping others or humanitarian activities has a direct effect on the willingness to help campaigns or support organizations (De Wit et al., 2018). In fact, one is socially responsible for adversity and natural disasters when he has a positive personal interpretation of helping others, and then shows an inner need for helping and supporting others. In this regard, the following hypotheses can be put forward: H 9 :An individualistic explanation of poverty has negative indirect effect on intention to donate to a disaster relief campaign.
H 10 :Perceived need for donation has positive indirect effect on intention to donate to a disaster relief campaign.

News exposure and intention to donate
Each of the various roles of the modern media becomes, at times, crucial to society, and sometimes even decisive. The media appears at every moment in various roles, such as informative, but in the event of major events some of these roles become more prominent (McKinnon et al., 2017). In crisis management by media, this is a media organization that is active because of high news values of critical events and has to cover critical events according to its mission. Media, in times of crisis or natural disasters, have an important role to play in encouraging people to assist in humanitarian campaigns or actions. In fact, informative media can facilitate post-disaster management and promote social cohesion and a sense of community duty for supporting relief actions and organizations (Lee & Holladay, 2017). The media will not be affected by crisis with right performance and timely response to needs of their audiences. Accuracy, honesty, speed, comprehensiveness and timeliness of news published by media can make a community susceptible to disaster management and its consequences and boost public awareness of their responsibility (Zhou & Pan, 2016).
In their "Donation to disaster relief campaigns: Underlying social cognitive factors exposed", Oosterhof et al. (2009) in Netherlands tended to improve the existing model by Cheung and Chan (2000). They also tended to study social cognitive factors which lead to donation to charity, focusing on an international relief organization. Other variables such as donating attitude and news exposure were added to the model. The expanded model showed a better fit. In this study, the greatest predictor of intention was "past donation to disaster campaigns". "News exposure" was shown to be an essential factor, as it had a direct and significant effect on "awareness of a disaster". News exposure was the only factor which had a general effect on other variables, including "Intention to donate to relief campaigns".
Findings in this area indicate that people exposed to up-to-date news are more likely to make financial and non-financial donations to support campaigns in the event of natural disasters (Lee & Holladay, 2017). Findings by Albrecht (2017) also show that trusted news has a positive and significant effect on intention to donate. Accordingly, the following hypothesis can be presented: H 11: News exposure has positive direct effect on awareness of natural disaster and thereby intention to donate to a disaster relief.

Population and sample
The studied population consisted of 411 employees in 6 departments of Red Crescent Society including Administrative and Support Department, Disaster and Rescue Department, Volunteers Department, Youth Department, Education and Research Department, and Treatment and Rehabilitation Department in different districts of Tehran. Stratified sampling was used and sample size was estimated at 199 people using Cochran formula.

Measurements
As data-gathering instrument was distributed among the population, this study was descriptive survey. As relationships between variables were studied, this study was correlational. Data collection method was primarily archival reviews in order to obtain information on background and theoretical literature; data collection instrument was questionnaire. Questions were measured on a five-point Likert scale varying from "strongly agree" (1) to "strongly disagree" (5). To analyze data, descriptive and inferential statistics were used. To describe data, frequency distribution tables, bar graph, mean and standard deviation were used; to examine hypotheses, Kolmogrov-Smirnov test, Pearson correlation coefficient, multivariate stepwise regression, linear diagram, independent t-test and pairwise t-test were used. For testing the causal model and estimating the relationships between the observed and latent variables, structural equation modelling (SEM) technique was used by Amos 18.00 software; K-S test and correlation coefficient test were used for analysis.

Reliability and validity
Cronbach's Alpha was used to measure reliability. Cronbach's alpha is the most common measure of internal consistency or reliability of the research questionnaires. A pilot survey is actually the imitation and trail of the main survey. The very purpose of doing a pilot study is to find out any flaw if it exists in the measuring instrument. Questionnaires are the measuring instrument employed in our current study. The responsiveness and applicability of the same have to be ensured through this pilot study by checking the reliability of the measurement instruments. The results of pilot survey based on 40 respondents mentioned in Table1.
According to Table 1 and 2, alpha coefficients were higher than 7.0; therefore, the questionnaire had good reliability. Experts were asked to confirm validity of the questionnaire; for this purpose, the questionnaire was given to 13 industrial experts and academic experts who suggested modifications to confirm the questionnaire.

Data normality
Kolmogorov-Smirnov test was used for valid data for checking the assumption of normality.
According to K-S test, p-value>0.05 of all variables supported null hypothesis. Thus, the variables had normal distribution and structural equations and parametric tests could be used to analyze the data.

Confirmatory factor analyses
In general, hypotheses were tested by SEM techniques using AMOS18 software. SEM offers statistical techniques in which path analysis and factor analysis are incorporated and integrated (Chehrehpak  Garson, 2006;Gholami et al., 2013;Noruzy et al., 2013). SEM is able to integrate measurement and causal models in a model to identify proper measures and parameters (Breckler, 1990). For this purpose, normality test was done first. Then, confirmatory factor analysis (CFA) and SEM were used to test the hypothesis. In statistics, confirmatory factor analysis (CFA) is a special form of factor analysis, most commonly used in social research. It is used to test whether measures of a construct are consistent with a researcher's understanding of the nature of that construct. In this study, a questionnaire was used for data collection. Content of questionnaires was validated using CFA. Measurement model represents factor loadings of the observed variables (factor) for each latent variable. Load factor is used to determine strength of the relationships between the factor (latent variable) and the observed variable. Load factor is a value between zero and one. If the load factor is less than 0.3, the relationship is weak. If load factor ranges from 0.3 to 0.6 and if it is greater than 0.6, the relationship is good. It is important to consider the model fit in CFA. The figures show common fit indexes used in measurement models for latent variables. If the ratio of chi-square to degrees of freedom (χ 2 /df) is less than 2, the model has a good fit. According to Table 3, χ 2 /df = 1.34 which is less than the acceptable level (>2); thus, the model has a good fit. RMSEA is good when it is less than 0.05. Other indices are good when they are closer to 1 (Chehrehpak et al., 2018;Ghasemi, 2011;Gholami et al., 2013;Noruzy et al., 2013).

Fit indices
Absolute fit indices are based on the difference between the observed variance and covariance on one hand, and the predicted variance and covariance according to model parameters on the other hand. High degree of freedom (close to independence model) and low chi-square (close to saturated model) indicate acceptability of the model and its goodness. The ratio of chi-square to degree of freedom, that is, relative. Results of the fit indices showed that χ2/df, RMSEA,CFI, IFI, P RATIO, PNFI and PCFI are in acceptable level means satisfactory model fit. Good fit between the model and the data means that the model confirmed.
According to Table 4, all items have suitable factor loadings (>0.3), supporting their relationship with latent variables. As shown in Table 5, effects of moral obligation to donate on donating attitude (p-value = 0.95), awareness of a natural disaster on perceived self-efficacy (p-value = 0.86) and past donation to disaster campaigns on intention to donate to a disaster relief campaign (p-value = 0.66) were rejected because their p-values were higher than 0.05.

Conclusion and hypothesis testing
This study tended to examine and explore factors which cause the migration of people to charitable behaviour based on Bandura's social cognitive theory. By discussing the Bandura's social cognitive theory and literature review the conceptual model has been developed. In theoretical foundations section intention to donate, perceived outcome and self-efficacy, moral obligation, trust in CHO, awareness of natural disaster, donating attitude, individualistic explanation of poverty and news exposure were discussed. The studied population consisted of 411 employees in 6 departments of Red Crescent Society including Administrative and Support Department, Disaster and Rescue Department, Volunteers Department, Youth Department, Education and Research Department, and Treatment and Rehabilitation Department in different districts of Tehran. Stratified sampling was used and sample size was estimated at 199 people using Cochran formula. The validity and reliability of research were confirmed by confirmatory factor analysis and Cronbach's α. According to the results alpha coefficients were higher than 7.0; therefore, the questionnaire had good reliability. The conceptual model was tested using structural equation modelling (SEM). The SEM results showed that perceived self-efficacy, outcome expectancies for donating to disaster campaign and intention to donate to charity has a positive direct effect on intention to donate a disaster campaign. Moreover, news exposure had no significant effect on intention to donate to a disaster campaign. Finally, results revealed that indirect insignificant effect of individualistic explanation of poverty, and awareness of a disaster on intention to donate to disaster campaign. Practical implications and limitations are subsequently discussed in light of the results. Results of the first hypothesis showed that perceived self-efficacy of donation has a positive and direct effect on intention to donate to a disaster relief campaign. According to Amos software output regarding the effect of perceived self-efficacy on intention to donate to a disaster relief campaign (Table 5), p-value = 0.000 which is less than 0.05 which showed that perceived self-efficacy has a positive effect on intention to donate to a disaster relief campaign. Thus first hypothesis is accepted (Figure 1 and 2).
This result is consistent with the findings of Chevance et al. (2017) and Clowes and Masser (2012). Therefore, when people believe in their abilities, skills, and knowledge in helping with disaster relief campaigns, they will be more willing to help with disaster relief campaigns. In other words, one's initial desire to help relief organizations or charities originates from one's inner values, and these values contribute to one's self-efficacy-belief in one's values and ability to contribute to relief campaigns.
Results of the first hypothesis showed that perceived outcome efficacy has positive direct effect on intention to donate to a disaster relief campaign. According to Amos software output regarding the effect of perceived outcome efficacy on intention to donate to a disaster relief campaign (Table 5), p-value = 0.005 which is less than 0.05 which revealed that effect pass coefficient of perceived outcome efficacy on intention to donate to a disaster relief campaign was positive and significant therefore, the second hypothesis is also confirmed.  Concerning hypothesis 3, which posits that moral obligation to donate has a positive indirect effect on intention to donate to charity and thereby intention to donate to a disaster relief. According to Amos software output shown in Table 5, the effect of moral obligation to donate on intention to donate to charity was tested through donating attitude. Regarding the effect of obligation on attitude (p-value = 0.95) which is more than 0.05, the hypothesis is rejected. However, the effect of attitude on intention to donate to charity was confirmed (p-value = 0.02). Therefore, moral obligation to donate has no significant direct effect on intention to donate to charity; thus, its significant indirect effect on intention to donate to disaster relief is rejected. Concerning hypothesis 4, which posits that intention to donate to charity is expected to have a positive direct effect on intention to donate to a disaster relief campaign. According to Amos software output shown in Table 5 regarding the effect of intention to donate to charity on intention to donate to a disaster relief campaign, p-value = 0.000 is less than 0.05. Moreover, path coefficient was significant and positive thus, the forth hypothesis is confirmed. The effect size is 91% which is very high.
This finding is consistent with the findings of Winterich et al. (2009). Intention to donate to charity is one of the key factors in people's willingness to help support campaigns in times of natural disasters. Because benevolence and philanthropy in helping charities shape their positive attitude towards helping humanitarian campaigns and they will be more likely to participate in natural disaster relief campaigns. In fact, the desire to help disaster relief campaigns is primarily influenced by morale and motivation of people to donate to charitable organizations and campaigns.
Results of the fifth hypothesis showed that trust in disaster help organization has a positive indirect effect on intention to donate to a disaster relief campaign. According to Amos software output shown in Table 5, the effect of trust in disaster help organization on intention to donate to Note. Paths are significant at p <.05. Dotted lines are nonsignificant paths. Bold lines are extensions to the original model by Cheung and Chan (2000).  a disaster relief campaign was tested through perceived self-efficacy. According to the results, the effect of trust in disaster help organization is confirmed on perceived self-efficacy (p-value = 0.000; which is lower than 0.05) and effect size is 51%. Moreover, the effect of self-efficacy on intention to donate to a relief campaign is confirmed (p-value = 0.000) and effect size is 23%. Hence, trust in help organization has a significant, positive indirect effect on intention to donate to a disaster relief campaign.
Concerning hypothesis 6, which posits that past donation to disaster relief campaigns has a positive direct effect on intention to donate to a disaster relief campaign. According to output shown in Table 5 regarding the effect of past donation to disaster relief campaigns on intention to donate to a disaster relief campaign, p-value = 0.66 is greater than 0.05. Therefore, the hypothesis will be confirmed.
This finding is consistent with the findings of Kassie et al. (2020). When one feels that their previous help with these campaigns has been effective and has improved the situation, they will probably repeat their experiences in the future and encourage others to help. Many researchers, therefore, place greater emphasis on past experiences and consider it one of the important factors in one's willingness to contribute to disaster campaigns. Concerning hypothesis 7, which posits that awareness of natural disaster has positive indirect effect on intention to donate to a disaster relief campaign. Results shown in Table 5, the indirect effect of awareness of a natural disaster on intention to donate to a disaster relief campaign was tested through perceived self-efficacy. According to the results, awareness of a natural disaster has an insignificant effect on perceived self-efficacy (P = 0.86). Thus, indirect effect of awareness of a natural disaster is not acceptable on intention to donate to a disaster relief campaign. Results of the eighth hypothesis showed that donating attitude has a positive direct effect on intention to donate to charity and has a weak indirect effect on intention to donate to a disaster relief campaign. According to output shown in Table 5 regarding the direct effect of donating attitude on intention to donate to charity, p-value = 0.02 is lower than 0.05. Therefore, donating attitude has a significant, direct and positive effect on intention to donate to charity. The effect size is 24%. Indirect effect of donating attitude on intention to donate to a disaster relief campaign was tested through intention to donate to charity. Given the significant effect of donating attitude on intention to donate to charity and significant effect of intention to donate to charity on intention to donate to a disaster relief campaign (p-value = 0.000), donating attitude has a significant indirect effect on intention to donate to a disaster relief campaign and effects size is 91% × 24% = 22%; therefore, the hypothesis is confirmed.
This result is consistent with the findings of Hassan et al. (2018) and James and Zagefka (2017). Therefore, a positive attitude towards helping campaigns will lead to action and practice of people for these campaigns and fulfilling their social responsibility. In fact, attitudes lead to make people more likely to be active or to contribute to humanitarian and relief campaigns.
Results of the ninth hypothesis showed that individualistic explanation of poverty has a negative indirect effect on intention to donate to a disaster relief campaign. According to Amos software output shown in Table 5, the effect of individualistic explanation of poverty on intention to donate to a disaster relief campaign was tested vie the path through the need for donation, moral obligation to donate, donating attitude, intention to donate to charity. According to the results, the effect of moral obligation to donate is insignificant on donating attitude (P = 0.95). Thus, indirect effect of individualistic explanation of poverty is not acceptable on intention to donate to a disaster relief campaign vie this path. Results of the tenth hypothesis showed that the indirect effect of perceived need for donation on intention to donate to a disaster relief campaign is tested vie the path through moral obligation to donate, donating attitude, and intention to donate to charity. The effect of moral obligation to donate is insignificant on donating attitude (P = 0.95). Therefore, indirect effect of the perceived need for donation on intention to donate to a disaster relief campaign is rejected via this path. Results of the eleventh hypothesis showed that news exposure has a positive direct effect on awareness of a natural disaster and on intention to donate to a disaster relief campaign. According to output shown in Table 5 regarding the direct effect of news exposure on awareness of a natural disaster, p-value = 0.000 is lower than 0.05. Hence, news exposure has a significant, direct and positive effect on awareness of a natural disaster. The effect size is 41%. Indirect effect of news exposure on intention to donate to a disaster relief campaign was tested through awareness of a natural disaster and perceived self-efficacy. Considering insignificant effect of awareness of a natural disaster on self-efficacy (p-value = 0.86), news exposure has an insignificant indirect effect on intention to donate to a disaster relief campaign.
This finding is consistent with the findings of Lee and Holladay (2017) and Albrecht (2017). Providing accurate and rapid news in times of crisis or natural disasters has an important role to play in encouraging people to assist in humanitarian campaigns or actions. The accuracy, honesty, speed, comprehensiveness and timeliness of news published by media can make the community more sensitive to participation in disaster management and its consequences, and strengthen the public conscious to responsibility towards them.

Discussion
The results of this study indicated the significant, positive, direct effect of self-efficacy on intention to donate to disaster relief campaign. This finding is consistent with Oosterhof et al. (2009) and Cheung and Chan (2000). In Cheung and Chan, the greatest predictor of intention to donate to IRO was self-efficacy. This shows that self-efficacy is a potent precursor to intention to donate. This is consistent with Bandura (1997), suggesting that perceived self-efficacy is a good predictor of various forms of behaviour. Perceived self-efficacy can be seen as an underlying factor of donating behaviour (Cheung and Chan 2000).
In this study, outcome expectancy, or outcome efficacy had a significant and positive direct effect on intention to donate to a disaster campaign. In Oosterhof et al. (2009), an unexpected finding was the insignificant direct effect of outcome expectancies for donating to disaster relief campaigns on intention to donate to a disaster relief campaign. The social cognitive theory highlights outcome expectancies (Bandura, 1997) and an insignificant effect found was strange. As outcome expectancies are one's beliefs or perceptions about consequences of actions (Schwarzer, 1992), it appeared plausible that people's expectation about their donation to a disaster relief campaign influences their intention to donate to disaster relief campaigns. This finding is consistent with Cheung and Chan who reported that outcome efficacy for donating to IRO had a significant direct effect on intention to donate to IRO. The results indicated a significant positive indirect effect of moral obligation to intention to donate, both to charity and to DRC. According to Oosterhof et al. (2009), once moral standards are formed, they will guide action (Bandura et al., 1996). On the other hand, Cheung and Chan showed that moral obligation generally failed to translate to support for a particular IRO, although it contributes to donation intention in general. Hence, the effect of moral obligation depends on the object, charity in general or the IRO, to which it refers. This study supports significant, positive and indirect effect of trust in CHO on intention to donate to DRC. This emphasizes findings of Oosterhof et al. (2009) andCheung andChan (2000). This direct effect can be attributed to cognitive consistency. This also explains why total effect of trust in IRO on donation intention was not significant in the presence of stronger predictors including self-efficacy and outcome efficacy.
The results of this study showed indirect insignificant effect of individualistic explanation of poverty on intention to donate to relief campaign. Individualistic explanation of poverty had a significant, negative direct effect on need for donation. Oosterhof et al. (2009) found that individualistic explanation of poverty had no significant effect on trust in CHO. They dealt with international natural disasters; in general, people agree that victims of these disasters cannot be blamed for the consequences. Cheung and Chan reported the significant total effect in light of indirect effects through self-efficacy, trust in IRO, and need for donation. Furthermore, attribution theory is relevant in that individualistic explanation of poverty tended to have direct effects on trust in IRO and need. Perceived need for donation which proved to have a significant direct effect on intention to donate both in Oosterhof et al. (2009) and in Cheung and Chan (2000), did not have a significant positive effect on intention to donate to a relief campaign. Direct effect of need for donation on moral obligation to donation was significant, while moral obligation to donate effect on donating attitude was insignificant. The results of this study supported the greatest part of the causal model of Cheung and Chan (2000) based on social cognitive theory. However, a striking finding was insignificant direct effect of past donation to disaster campaigns on intention to donate to a disaster relief campaign. According to Ajzen (2002aAjzen ( , 2002b, frequency of a performed behaviour can be a good predictor of future behaviour and an insignificant effect was strange. According to social cognitive theory, past behaviour provides the opportunity to learn self-efficacy and outcome efficacy of donation. Past donation could influence behaviour without promoting beliefs in self-efficacy and outcome efficacy. Some studies have supported the effect of prior donation on later donation (Diamond & Kashyap, 1997;Piliavin, 1989).Past donation to disaster relief campaigns, which was the greatest predictor in Oosterhof et al. (2009), was insignificant in this study. Moreover, direct and total effects of self-efficacy, trust in CHO, and outcome expectancies were significant. They are therefore crucial predictors of donation intention. This study supports the significant, positive direct effect of intention to donate to charity on intention to donate to DRC. The effect size is 0.913; this indicates that the greatest predictor of intention to donate is intention to donate to charity. This suggests that people who frequently donate to charity will be more likely to donate to disaster relief campaigns, whether charity or DRC. This is consistent with Oosterhof et al. (2009) and Cheung and Chan (2000). Therefore, intention to donate to charity would be effective on intention to donate to a disaster relief campaign, as expected.
The findings imply insignificant, indirect positive effect of awareness of a disaster on intention to donate to disaster campaign. This is inconsistent with Oosterhof et al. (2009), who found another positive relationship between awareness of natural disaster and self-efficacy of donation to a disaster relief campaign. This outcome is consistent with social cognitive theory, which claims that human functioning can be illustrated by triadic reciprocity (Bandura, 1986). In terms of reciprocity, interactions of behaviour, cognitive and personal factors and environmental events can influence each other. As awareness of a disaster is associated with awareness of environmental events, it explained self-efficacy in the expanded model. Cheung and Chan also reported significant total effect of awareness of IRO on intention to donate to IRO.
In conclusion, the effects demonstrate a causal relationship predicted by social cognitive theory. Besides, social cognitive model is supported by significant relationships between donation in general and donation to DRC. Social cognitive model takes advantage of an integration of social cognitive theory, cognitive-consistency theory, and attribution theory in defining relevant factors to donation.
Donating attitude and news exposure were added to the model. Although Cheung and Chan (2000) found that contribution of attitude was insignificant in their model, results of the present study showed the opposite. One's perception is an important dimension of charitable giving. In the modified model, donating attitude had a moderately strong direct effect on intention to donate to charity and had a weak indirect effect on intention to donate to a disaster relief campaign. This is consistent with the theory of reasoned action; according to Ajzen and Fishbein (1980), when people evaluate a behaviour positively, they will tend to perform that behaviour. According to empirical evidence, beliefs in self-efficacy and outcome efficacy also outperformed the attitude toward DRC in explaining donation intention. As attitude did not influence self-efficacy and outcome expectancies for donating to a disaster relief campaign, the finding is consistent with this the fact that behaviour can independently be influenced by cognition and attribution without mediating effect (Cheung and Chan 2000).
News exposure was a valuable additional factor, as it had a significant direct effect on awareness of a natural disaster, while it had no total effect on all other factors, including intention to donate to a disaster relief campaign. In Oosterhof et al. (2009), news exposure was the only factor that had a total effect on all other factors, including intention to donate to a disaster relief campaign. This finding is inconsistent with McCombs (1994) who argued that news is effective on many facets of people's daily life.

Conclusion
The first, second, fourth, eighth, and tenth hypotheses were confirmed and the third, fifth, sixth, seventh, ninth, and eleventh hypotheses were rejected.
The modified model, based on the causal model of Oosterhof et al. (2009), could explain the intention to donate to a disaster relief campaign. Behavioural intention is critical to assess as an outcome variable because it is proved as the immediate antecedent to behaviour. New effects, correlations and factors were included in the model. These additions led to a better fit of the model. Furthermore, the modified model explained more than one-third of variance in intention to donate to a disaster relief campaign, which was less than the two-thirds of variance explained in intention to donate to the relief campaign in the model of Oosterhof et al. (2009). The greatest predictor of intention to donate to disaster relief campaigns was intention to donate to charity, followed by self-efficacy of donating to a disaster campaign, and outcome expectancies for donating to a disaster campaign.

Limitations and future study
Support of a causal model does not mean that causal-relationships are verified as hypothesized because the design of the study is cross-sectional. Yet, causal relationships are plausible because of 1) theoretical support, 2) logical sequence, and 3) empirical goodness of fit of the causal model and significance of the causal paths. Accordingly, social cognitive theory provides the theoretical support to suggest a causal chain of awareness and exposure to beliefs and intention successively. The causal model also complies with the logic of causation in which retrospectively past (e.g., past donation) and stable event affects present and future (i.e. intention) (Davis, 1985). These theoretical props may enhance credence of the causal model and avoid specious and confusing post-hoc rhetoric found in many local studies (Shek, 1997). Another problem inherent in surveys is reliance on self-report measures, including beliefs, attitudes, intentions, and behaviour. These measures might only disclose part of the truth when the rest is inaccessible due to biases of social desirability and information processing. By using anonymous measures that people could fill out, an attempt was made to avoid socially desirable answers. Another point of attention was the inability to find out the actual donating behaviour of participants. Donating to disaster relief campaigns is an anonymous process. The study was also unable to tap actual behaviour after the survey because it involved an anonymous survey. Hence, it could only regard intention as a proxy for future behaviour and specify past behaviour as a predictor of intention and other social cognitive factors rather than an outcome. In conclusion, the present study adapted the causal model of Oosterhof et al. (2009), which entailed some limitations. However, the modified model indicated a better fit than the original model and underlying social cognitive factors proved to be useful for explaining more than one-third of the variance in intention to donate to disaster relief campaigns.
Findings of the present study imply that self-efficacy, outcome efficacy, intention to donate to charity, trust in the organization, and donating attitude are important social cognitive factors to promote public intention to donate to relief campaigns. They should offer clues to planning promotional activities for the DRC. Thoughtful planning is necessary because of its absence in current promotional activities of the DRC which fail to achieve the expected goal. The clues suggest emphasis on promoting people's self-efficacy and perceived outcome efficacy regarding donation to the DRC. They also highlight public awareness of DRC as an initial step to promote trust and beliefs in self-efficacy and outcome efficacy. Beliefs in donating attitude, need for donation, and causes of poverty are also important. The emphasis would be a remedy for current promotional activities that miss a concrete plan. For instance, existing promotional activities on television involve a mixture of singing and games which carry fun but fail to enhance the audience's social cognition about donation.
Moreover, results of the present study indicate that self-efficacy of donating to a disaster relief campaign, and intention to donate to charity have a positive direct effect on intention to donate to a disaster campaign. For planning new campaigns, this result implies that future campaigns should be similar to former campaigns in order to comply with self-efficacy of already existing donors. The question is, however, how to persuade people to take the first step in donation behaviour, and to attract these new donors without losing the old ones. Further results of the present study provide some clues to how to proceed with this difficult task.
To promote self-efficacy and outcome efficacy regarding donation to DRC, appropriate programs should emphasize the ordinary people's ability to contribute to the DRC. To enhance the former, the programs should portray that the DC is accessible and welcomes public donation, regardless of its amount. They may also employ modelling and role playing to enhance perceived self-efficacy for donation. To illustrate efficacy of the DRC in helping poor people, promotional programs should both exhibit the helping process and outcome of recipients. Feedback of recipients can be effective in demonstrating outcome efficacy. Moreover, the programs should elucidate the contribution of the DRCs to ameliorating structural factors which in turn improve quality of life of poor people. The programs can foster people's need and moral obligation to donation by having authority figures preach the importance of these issues. Nevertheless, designing and evaluating programs for enhancing the audience's beliefs in self-efficacy and outcome efficacy are the next step in research on donating to the DRC. This design and research may also benefit from past research on enhancing efficacy beliefs in other settings (Bartholomew et al., 1997;Ozer & Bandura, 1990;Spoth et al., 1995). These studies showed that intervention elements of skill enhancement and goal setting are effective for promoting people's self-efficacy for parenting, health maintenance, and defense against sexual assault. Inevitably, because of the particular nature of donation behaviour, planning and evaluation programs for enhancing the public's efficacy beliefs pose challenges to future research.
When it comes to awareness of a disaster, the media play a big role. News exposure was not the factor that had a total effect on all affected factors: intention to donate to charity, intention to donate to disaster relief campaigns, and self-efficacy of donating to a disaster relief campaign. This result does not speak for quick action, to start media campaigns when the events are still at the top of the mind of potential donors. It may not help to use so-called media-ambassadors; television personalities with an attractive media image cannot make a considerable impact.