Development and initial validation of a theory of planned behavior questionnaire: Assessment of purchase intentions towards products associated with CRM campaigns

Abstract This research develops and validates scales based on the theory of planned behavior (TPB) to measure purchase intention towards products associated with cause-related marketing (CRM) campaigns in the South Asian context. Despite few studies using global measures of TPB in specific contexts to predict behavioral intention towards CRM campaigns, this study develops and uses belief-based formative indicators that can be used as an intervention to bring about behavioral changes for positive campaign outcomes. A mixed methods approach was used, including focus group discussions and open-ended questionnaires, to collect qualitative data from 62 participants, resulting in the development of the formative indicators of the measurement instrument. The scales were then combined with global measures of reflective indicators and validated using data collected from 1035 respondents in a quantitative study. The results support the TPB theory and show that the scales have strong internal consistency, reliability, and validity. The findings indicate that behavioral beliefs (β = 0.834, p < 0.001), normative beliefs (β = 0.631, p < 0.001), and control beliefs (β = 0.725, p < 0.001) significantly impact attitude, subjective norm, and perceived behavioral control respectively. Attitude (β = 0.374, p < 0.001), subjective norms (β = 0.218, p < 0.001), and perceived behavioral control (β = 0.320, p < 0.001) significantly influence purchase intentions, with attitude having the most significant impact. The study also found that purchase intention significantly affects purchase behavior (β = 0030.530, p < 0.001). And therefore, this study strengthens the theory of planned behavior in the context of CRM campaigns, aligning with the broader field of ethical consumption.


Introduction
Contemporary society expects ethical and sustainable conduct from companies apart from financial performance (Huang et al., 2022). Consumers have recently begun to place significant value on an organization's social responsibility (Schuster et al., 2016;Woodroof et al., 2019). Such expectations compel organizations to achieve corporate objectives and enact their social responsibilities (Garg & Gupta, 2020;Sebrina et al., 2023). Cause-related marketing (CRM) is a manifestation of corporate responses to stakeholder expectations (Bae & Wright, 2020). As a result, CRM initiatives have become increasingly popular among organizations over the last three decades. Scholars define CRM as "the process of formulating and implementing marketing activities that are characterized by an offer from the firm to contribute a specified amount to a designated cause when customers engage in revenue-providing exchanges that satisfy organizational and individual objectives" (Varadarajan & Menon, 1988, p. 60). The uniqueness of a CRM campaign stems from its contribution to product value creation through active customer participation, lending a participative attribute to C-RM (Christofi et al., 2020). Customers perceive purchasing products associated with a social cause as a means of contributing to the betterment of others, which ultimately leads to societal benefits (Gilal et al., 2020).
One of the best examples of successful CRM campaigns in India is the Shiksha campaign by Procter and Gamble. It is one of India's most successful, oldest, and continuous CRM campaigns (running since 2005) (Hawkins, 2015). Procter and Gamble has substantially contributed to education through this program by leveraging their products to promote impoverished children's access to quality education (Kataria et al., 2021;Citizenship report P&G India subcontinent, 2022). The Shiksha campaign has received worldwide attention, but it has also directly affected the lives of thousands of children, demonstrating the power and effectiveness of CRM in solving social challenges (Aggarwal & Singh, 2019). Following the Shiksha campaign's positive impact and public response, several corporations have recognized the potential of CRM to push social change while also boosting their brand image (Kataria et al., 2021). As a result, many businesses have adopted CRM as an effective tool for contributing to social issues, spawning a bigger corporate social responsibility movement in India and abroad.
In today's context, CRM campaigns have evolved beyond mere product-based contributions and played a larger role in creating social awareness across various issues. Nike's "Find Your Greatness" campaign, Dove's "#Stop the Beauty Test" initiative, and Ariel's "Share the Load" campaign are notable examples of this evolution (M. Kim, 2020). These advertisements represent rising marketing communication trends that aim to raise social awareness and spark conversations regarding major societal concerns.
As we can observe growth in CRM campaigns, so do we see an increase in its research interest. Researchers have consistently attempted to understand the various contexts under which CRM campaigns drive the best outcomes (Bhatti et al., 2022). Although much CRM research has emerged from the USA, the literature from other nations has grown significantly in the last two decades (Bhatti et al., 2022). This growth in CRM research takes a different shape, particularly considering the diverse needs of consumers of different nationalities and cultures (Roggeveen & Beitelspacher, 2020). Specifically, this growth has resulted from different perceptions and social expectations (Christofi et al., 2020). Hence, researchers worldwide have been interested in understanding CRM campaign outcomes in diverse settings (Roggeveen & Beitelspacher, 2020). These outcomes determine the success of a marketing strategy. Therefore, scholars working on the consumer side of CRM research have focused on understanding consumer perceptions of a given campaign, such as their attitudes and behavioral intentions to participate in a CRM campaign (Ferraris et al., 2020).
Individuals' purchase intentions indicate their behavior and hence are widely used by marketers to make decisions (Huang & Ge, 2019; Morrison, 1979). Prior research has employed theories such as attribution theory, the elaboration likelihood model, and consistency theory to examine customer behavior responses to CRM campaigns. Surprisingly, scholars have not widely used two of the most prevalent behavioral theories, social cognitive theory (SCT) and theory of planned behavior (TPB), to explain behavioral intentions in the context of a CRM campaign (H. Kim et al., 2019;Zhang et al., 2020). This scant application is surprising because the TPB has been used to successfully determine behavioral intentions in similar areas under the broader umbrella of ethical consumption (Ahmad et al., 2020). The strength of TPB lies not only in predicting behavioral intention but also in understanding the underlying beliefs that shape such decisions. Studies that have previously used TPB to predict purchase intention towards products associated with CRM campaigns have used global measures to predict purchase intentions. By implication, they have not considered the underlying beliefs that drive such behavior, which scholars suggest is much needed in a new study context (Roggeveen & Beitelspacher, 2020). This study aimed to fill this gap. Thus, the central objective of this research is to develop and validate an instrument based on TPB variables to capture purchase intentions and, subsequently, the purchase behavior of products associated with a CRM campaign. After developing this scale, this study investigated the following research questions: (1) What are the effects of behavioral beliefs, normative beliefs, and control beliefs on the attitude, subjective norm, and PBC in forming intentions to participate in a CRM campaign?
(2) What are the effects of attitude, subjective norm, and PBC on the intention to participate in a CRM campaign?
(3) What are the effects of purchase intention on purchase behavior for products associated with CRM campaigns?
Answering the research questions mentioned above helps validate the measurement instrument we develop.
The TPB is a psychological theory that links beliefs to behavior. Based on the Theory of Reasoned Action (TRA), TPB states that behavioral achievement depends on beliefs about both motivation (intention) and ability (behavioral control). The Theory of Planned Behavior (TPB) assumes that predicting an individual's intention to engage in a behavior becomes more accurate when the individual has a positive attitude towards the behavior (attitude), perceives social support from peers (subjective norm), and believes they have the necessary capabilities to perform the behavior (perceived behavioral control) (Ajzen, 2020). Attitude, subjective norms, and perceived behavioral control (PBC) are direct or reflective indicators of a person's behavior. These are the results of the underlying beliefs that shape them. Behavioral beliefs and outcome evaluations determine attitudes. Normative beliefs and motivations for compliance determine subjective norms. Similarly, control beliefs and perceived power determine PBC. Outcome evaluation, motivation to comply, and perceived power denote the strength of each underlying belief (Ajzen, 2020). Behavioral, normative, and control beliefs are indirect or formative indicators of a person's behavior. Figure 1 presents a conceptual model of the TPB.
Ajzen added the construct of PBC to the previously proposed theory of reasoned action (TRA) while he modified TPB. This extension allowed the prediction of behaviors that were not in complete volitional control. Fishbein and Ajzen (2011) note that, although stronger intentions generally lead to an increased likelihood of performing a behavior when actual behavioral control is low (e.g., lack of essential skills or presence of environmental barriers), individuals may not carry out their intentions. This absence of intention-behavior linkage implies assessing skills, abilities, environmental barriers, and facilitators to understand when a behavior is likely to occur. Ajzen (1991) suggests that PBC influences behavior indirectly through behavioral intention when the behavior is not entirely under the volitional control of the individual. PBC can be an independent predictor of behavior to the extent that it is accurate and reflects actual ability.
Reflective (direct) indicators are sufficient to predict behavioral intention and allow examination of the predictive validity of the theory. However, it is important to understand the factors that provide the basis for behavior (perhaps to design an effective behavior change intervention) (Ajzen, 2020). In such cases, determining formative (indirect) indicators is necessary. Formative beliefs can be captured through an elicitation study to access the composites of readily accessible behavioral, normative, and control beliefs. Researchers can later use these to design items to capture formative beliefs in a self-report survey. We can develop reflective measures based on the validated measure of prior literature, which we can modify according to the context of the study. Scholars have found a high correlation between direct and indirect measures, with indirect measures providing a path for the assessment of more detailed insights into the roots of the behavior under investigation (Ajzen, 1991).
Based on the TPB assumptions and literature discussed above, the following hypotheses are proposed: H1: Behavioral beliefs positively influence consumers' attitudes towards purchasing a product associated with a CRM campaign.
H2: Normative Beliefs positively influence consumers' subjective norms towards purchasing a product associated with CRM campaigns.
H3: Control beliefs positively influence consumers' perceived behavioral control towards purchasing a product associated with a CRM campaign.
H4: Attitude significantly influences consumers' intention to buy a product associated with CRM campaigns. H5: Subjective norm significantly influence consumers' intention to buy a product associated with a CRM campaign.
H6: Perceived behavioral control significantly influences consumers' intention to buy a product associated with a CRM campaign.

H7:
The intention to purchase a product associated with a CRM campaign significantly influences the purchase behavior toward such products.
H8: Perceived behavioral control significantly influences consumers' purchase behavior towards a product associated with a CRM campaign.

Research methodology
Scholars in this study developed a TPB questionnaire based on comprehensive guidelines for structure, item wording, and scoring criteria (Ajzen, 2006). We validated the study using 16-point criteria for assessing TPB quality (Oluka et al., 2014). These criteria emphasize the importance of methodological rigor, including an elicitation study, developing indirect and direct measures, and establishing content validity. This study reports on the current research in three stages (see Figure 2). Study 1 focused on eliciting and zeroing in the content analysis of salient beliefs about purchasing a product associated with a CRM campaign. Study 2 focused on developing and validating a questionnaire that assessed indirect beliefs, direct beliefs, and intentions regarding products related to CRM campaigns. Study 3 focused on hypothesis testing to validate TPB in the context of CRM campaigns.

Study 1: Elicitation study of belief constructs
Before developing the TPB questionnaire, the prior research recommends an elicitation study to ascertain the salient beliefs of an individual towards a particular behavior (Fishbein & Ajzen, 2011). This elicitation helps establish the cognitive foundation of a population's salient beliefs to form behavioral intentions towards said behavior (Downs & Hausenblas, 2005). Elicitation studies are essential because they provide researchers with valuable information concerning people's thoughts and feelings regarding a behavior (Downs & Hausenblas, 2005). The literature recommends a minimum sample of 25 participants to sufficiently ascertain salient beliefs among the population (Godin & Kok, 1996). Hence, in this study, the authors conducted focus group discussions and administered open-ended questionnaires to elicit salient beliefs from respondents who had previously participated in a CRM campaign.

Study 1A: Focus group discussions
The authors conducted a series of six focus group discussions. Focus group discussions were preferred over in-depth interviews because of their conversational and participatory nature, which leads to the generation of additional topics and provides enhanced discussions with new insights (Samaddar & Gandhi, 2022). All the participants provided informed consent. They used online and offline modes of discussion to enable in-depth discussion and broader participation. The online discussions included participants from South Asian countries, including Bhutan, India, Nepal, and Sri Lanka. (I). Each focus group discussion lasted approximately 40 minutes. In addition, convenience sampling was used to select 37 respondents who participated in the six focus group discussions. We used a semi-structured focus group method to conduct focus group discussions. In addition, we used nine pre-designed questions to understand respondents' behavioral, normative, and perceived control beliefs towards a product associated with CRM campaigns. Following the recommendation, we worded the questions to elicit salient beliefs about the advantages and disadvantages (behavioral beliefs), who would approve or disapprove of (subjective norm), and what would make it easy or difficult (control beliefs) to purchase products associated with the CRM campaign. (See Table 1 for the discussion questions in the focus group interviews).
We conducted each subsequent focus group discussion, with a gap of three days to one week. This time gap allowed us to analyze the content after each discussion and improve moderation in subsequent focus group discussions (Krueger & Casey, 2000). Hence, we ensured that focus group discussions, open-ended questionnaires, and coding occurred simultaneously, as suggested by Krueger and Casey (2000).

Study 1B: Administering an open-ended questionnaire
In addition to the focus group discussions, we administered an open-ended questionnaire to participants to examine if any new themes emerged. Multiple sources increase the reliability of exploratory studies' findings (Kemp & Vinke, 2012). We used a convenient sampling method to recruit participants. We chose 40 undergraduate students, employed professionals, and selfemployed individuals residing in a famous university town in Karnataka, India (See Table A1 in Appendix 1 for a detailed description of the participants and the outcome of the open-ended questionnaire study). All the participants provided informed consent. Participants responded to the items in the questionnaire on a hard copy. In this connection, we created three questionnaire versions to reduce the order effect with questions in different orders. We allowed only those participants who had previously participated in the CRM campaign to participate in the openended questionnaire survey. The authors closed collecting the questionnaire responses once they received 25 responses.
We also used questions designed for focus group discussions in the open-ended questionnaires. The following statements preceded the questions:' Please take a few minutes to tell us what you think about cause-related marketing. Cause-related marketing is a strategy in which a firm commits to donating a specific amount to a non-profit organization (NPO) or social/environmental cause when customers purchase their products. For example, in its "DESH KO ARPAN" program, Tata Salt contributed ten paise for every kilo of Tata Salt sold during specific periods to the education of underprivileged children. The Child Relief and You (CRY) Foundation was non-profit partners in this campaign. Please list your thoughts in response to the questions below. There is no right or wrong answer; we are only interested in your opinion. Each is written through a separate line. Five lines were provided for each question".

Table 1. Questions used for the exploratory studies
To elicit behavioral beliefs What do you think are the advantages of buying a product associated with a cause-related marketing campaign?
What do you think are the disadvantages of buying a product associated with a cause-related marketing campaign?
Is there anything else you associate with purchasing a product associated with cause-related marketing?
To elicit normative beliefs Please list the individuals or groups who would approve or think you should buy a product associated with the cause-related marketing campaign.
Please list the individuals or groups who would disapprove or think you should not buy a product associated with the cause-related marketing campaign.
Is there anything else you associate with other people's views about purchasing a product associated with cause-related marketing?
To elicit control beliefs Please list any factors or circumstances that would make it easy or enable you to buy a product associated with the cause-related marketing campaign.
Please list any factors or circumstances that would make it difficult or prevent you from buying a product associated with the cause-related marketing campaign.
Are there any other issues that come to your mind when you think about purchasing a product associated with cause-related marketing?

Content analysis
During the focus group discussion and administration of the carefully worded open-ended questionnaire survey, content analysis of the responses provided the beliefs that underlie the TPB's indirect psychological factors. We stopped the focus group discussions upon reaching saturation when three subsequent discussions yielded no new codes (Saldaña, 2009).
Using content analytic techniques, we analyzed transcripts from focus group discussions and questionnaire responses (Miles & Huberman, 1994). We labeled the content under three criteria: behavioral beliefs, normative beliefs, and control beliefs (See Tables A2 -A4 in Appendix 1 for the quotes that emerged from the coding and the labeled codes). The lead author conducted a content analysis of the transcripts from the focus group discussions and responses to openended questionnaires to identify emerging themes. Next, the author collaborated with the second author to categorize all salient beliefs into different themes. For example, in the question "What do you think are the advantages of buying a product associated with a causerelated marketing campaign?" responses like "feeling of engagement in community building," "contributing to the community," and "making the community a better place" were included in the theme "community building." In some instances, the respondents mentioned the same responses to different questions. For example, "will have to pay more," "expensive product," and "competitive pricing" were cited as responses to questions about "disadvantage," "what would make it easy," and "what would make it difficult." To avoid repetition in such cases, we decided to frame such questions in only a single category, considering the right fit. In this case, we included the question about pricing to be a part of "control beliefs." In the case of the "approve/disapprove" questions, the responses were the same. For example, respondents mentioned "family," "friends," and "colleagues" colleagues on both lists. In the case of the "disapproval" question, terms like "skeptical friends" or "skeptical family members" were used apart from "friends" and "family." Consistent with Ajzen and Fishbein (1980;see also Francis et al., 2004), beliefs were rankordered based on the frequency of occurrence (See Tables 2 -4 to see a list of rank-ordered belief codes). Furthermore, we used 75% of the most frequently occurring beliefs to generate items that formed part of the indirect measures in the survey.

Results
The most answered questions were beliefs about advantages (n = 208; Tables 2, 3 , and 4) followed by factors or circumstances that would make it easy to buy products associated with CRM campaigns (n = 147). By contrast, only 32 responses addressed who would disapprove of buying such products. Among the responses that reflected behavioral beliefs, the code "sense of satisfaction" (n = 33) was the most commonly reported advantage, and "competitive pricing" (n = 16) was the most reported disadvantage of buying a product that supports a cause. The most prevalent response, which captured the normative beliefs of approvers and disapprovers, was "friends" (n = 21, n = 16). Finally, in response to the question on control beliefs, "transparency" (n = 31) and the product being "competitive pricing" (n = 24) were the most responded answers to factors or circumstances that made it "easy/difficult" easy/difficult' to buy such a product.

Study 2: Development and validation of indirect and direct measures, purchase intention, and purchase behavior questionnaire
The salient beliefs from Study 1 laid the foundation for the indirect measures used in Study 2. In addition to indirect measurement, we developed items for direct measures, purchase intention, and purchase behavior in this study. The direct measures are global measures that we created using items from an area similar to earlier studies (Ajzen, 1991).

Scale development for indirect beliefs
Based on this recommendation, we included more than 75% of the salient beliefs in the questionnaire. The items for each belief construct resulted from the questions in the elicitation questionnaire. The advantage/disadvantage questions provided themes for behavioral belief items, approve/disapproving questions for normative beliefs, and easy/difficult questions for control beliefs. These themes provide the basis for item generation in the conceptualization of formative constructs. The first questionnaire consisted of 25 items: 12, 6, and 7 for behavioral, normative, and behavioral beliefs, respectively.
Considering that common method bias (CMB) can be a problem in cross-sectional studies, we used procedural remedies suggested by Podsakoff et al. (2003) to overcome CMB. We conducted cognitive interviews to improve the scale item clarity. We used reverse-coded items to break monotonous and similar patterns and gain the participants' focus. In addition, we included negative questions to remove common scale properties (Herche & Engelland, 1996).

Study 2A: Content validity
We adopted the inter-rater reliability measures of the multi-rater kappa to assess the instrument's content validity. In addition, we requested that six marketing professors from three universities in Karnataka participate in the content validity process. In addition, we asked the participants to rate all the items for their relevance on a four-point scale with labels indicating "Not Relevant (1)", "Item needs some revision (2)", "Need minor revision (3)," and "Very relevant (4)." We calculated itemcontent validity index (I-CVI) for each item and Scale-Content validity Index (S-CVI) for the entire instrument" (Shrotryia & Dhanda, 2019). Scholars suggest calculating multi-rater kappa further, as the I-CVI does not consider chance agreement (Shrotryia & Dhanda, 2019). Hence, we computed multi-rater kappa for each item. As proposed, we retained all items above 0.74. Further, we calculated the S-CVI using the average method resulting in a value of 0.86, considered highly acceptable (Davis, 1992; See Table A5 in Appendix 2 for the definitions, calculations, and accepted values of I-CVI, S-CVI/ Avg, and Multi-rater Kappa techniques). Finally, we retained 20 items at the end of the content validity process (See Table A6 in Appendix 2 for the results of the inter-rater reliability study).

Study 2B: Face validity
We adopted Cognitive interviews (CIs) to assess the validity of the formative instrument. The CI enables researchers to determine whether the intended recipients understand the instrument's content (Boateng et al., 2018). Scholars argue that "cognitive interviewing has emerged as one of the more prominent methods for identifying and correcting problems with survey questions" (Beatty & Willis, 2007, p. 288). Hence, we conducted CIs with five teaching community members at a large university in Karnataka. All participants were over 18 years old and had previously purchased a product associated with the CRM campaign. In this connection, we adopted the verbal probing technique by requesting the participants to read each item and then ask questions to gauge the respondents' understanding (Beatty & Willis, 2007). Of all the items for the indirect constructs, minor modifications were incorporated for two questions and were presented again to verify the respondents' cognitive assessments. No further modifications were made to these items.

Scales for direct beliefs, purchase intention, and purchase behavior
Based on the recommendations outlined in prior research, we developed questionnaires for direct beliefs, purchase intention, and purchase behavior aspects. In this connection, we used items from previously validated TPB measures for green purchase behavior.

Attitude, subjective norm, and PBC
We adopted a seven-point semantic differential scale from Han and Kim (2010) to measure attitude; for example, buying a product associated with a cause-related marketing campaign is extremely bad (1) or extremely good (7). Subjective norm was measured using two items adapted from Chan and Lau (2002); for example, most people who are important to me would want me to purchase a product associated with a cause-related marketing campaign (strongly disagree (1)/ Behavioral Belief (BB): (strongly disagree (1)/strongly agree (7)) BB1 I feel a sense of satisfaction when I buy a product associated with a cause-related marketing campaign.

BB2
I feel good when I buy a product associated with a cause-related marketing campaign.

BB3
I feel I am contributing to community building, when I buy a product associated with a cause-related marketing campaign.

BB5
I feel a sense of empathy when I buy a product associated with a cause-related marketing campaign.

BB8
I feel I am helping people in need when I buy a product associated with a cause-related marketing campaign.

OE1
To me, experiencing a sense of satisfaction when I buy a product is.

OE2
To me, feeling good when I buy a product is

OE3
To me, contributing to the community when I buy a product is

OE5
To me, being empathetic when I buy a product is

OE8
To me, helping people in need is as a result of my purchase is Normative belief (NB): (strongly disagree (1)/ strongly agree (7)) NB1 My friends think I should buy a product associated with a cause-related marketing campaign instead of a conventional product.

NB2
My family thinks I should not buy products associated with a cause-related marketing campaign instead of conventional products.

CB1
Transparency in contribution to charity would influence my decision to buy a product associated with a cause-related marketing campaign.

CB2
The competitive price of a product associated with a cause-related marketing campaign influences my decision to buy it in place of a conventional product.

CB3
A product's competitive quality associated with a cause-related marketing campaign influences my decision to buy it instead of a conventional product.

CB4
A convenient purchase location would make it easy for me to buy a product associated with a cause-related marketing campaign.

CB7
Details of the contribution to the cause on product packaging will make it easy to buy a product associated with a cause-related marketing campaign.

PP1
Transparency in contribution to charity would influence my decision to buy a product associated with a cause-related marketing campaign.

PP2
The competitive price of a product associated with a cause-related marketing campaign influences my decision to buy it in place of a conventional product.

PP3
A product's competitive quality associated with a cause-related marketing campaign influences my decision to buy it instead of a conventional product.

PP4
A convenient purchase location would make it easy for me to buy a product associated with a cause-related marketing campaign.

PP7
Details of the contribution to the cause on product packaging will make it easy to buy a product associated with a cause-related marketing campaign.

SN1
Most people who are important to me would want me to purchase a product associated with a cause-related marketing campaign.
Chan and Lau (2002) SN2 Most people who are important to me would think I should purchase a product associated with a cause-related marketing campaign.
strongly agree (7)). PBC was measured using three items adapted from Y. Kim and Han (2010), for example, whether or not I buy a product associated with a cause-related marketing campaign in place of a conventional product is entirely up to me (strongly disagree (1)/strongly agree (7)).

Purchase intention
Purchase intention was measured by adopting items from Y. J. Kim et al. (2013). For example, I will purchase a product associated with a cause-related marketing campaign for personal use (strongly disagree (1)/strongly agree (7)).

Purchase behavior
Purchase behavior was measured using three items from Wan et al. (2012). I have been purchasing a product associated with CRM campaigns regularly (strongly disagree (1)/strongly agree (7)). Table 5 presents the final list of items used to capture all variables in the study.

Code Constructs and scale items Sources Sources
Perceived behavioral control (strongly disagree (1)/strongly agree (7)) Y. Kim and Han (2010) PCB1 Whether or not I buy a product associated with a cause-related marketing campaign instead of a conventional product is entirely up to me.

PCB2
I have resources, time and opportunities to buy products associated with a causerelated marketing campaign.

PCB3
I am confident that if I want to, I can buy a product associated with a cause-related marketing campaign instead of a conventional product.
Purchase intention (strongly disagree (1)/ strongly agree (7)) Y. J. Kim et al. (2013) PI1 I will purchase a product associated with a cause-related marketing campaign for personal use.

PI2
I am willing to purchase a product associated with a cause-related marketing campaign for personal use.

PI3
I will try to purchase a product associated with a cause-related marketing campaign.

Purchase behavior (strongly disagree (1)/ strongly agree (7))
Wan et al. (2012) PB1 I have been purchasing products associated with a cause-related marketing campaign regularly.

PB2
I have purchased a product associated with a cause-related marketing campaign for my daily needs.

PB3
I have purchased a product associated with a cause-related marketing campaign in the past six months.
Source: Authors own.

Procedure
This study was conducted in September 2022. An online survey was created using Google Forms. The questionnaire was distributed using a short link to contacts on social media via the personal account of lead researchers. All items within each part of the questionnaire were randomized to reduce order effects. As Podsakoff et al. (2003) suggested, we collected data from three waves, with a temporal separation of one month each. Temporal separation of the independent and dependent variables in a longitudinal design (Menard, 2008) is viewed particularly favourably in management research, and there is strong evidence that it reduces common method bias (e.g., Lindell & Brandt, 2000). All the participants provided informed consent. In the first wave, we collected the demographic profiles of the respondents and data for the independent constructs, including formative and reflective constructs. In the second wave, we collected data for the moderating variable purchase intention, and finally, the purchase behavior data were collected in the final round. Respondents' email Ids were collected at all three stages to consolidate the data for analysis. 1637, 1381, and 1035 individuals participated in the first, second, and third waves of data collection, respectively. Data from 1035 individuals who participated in all three survey rounds were analyzed.

Analysis
The behavioral, normative, and control beliefs (responded on a scale of 1 to 7) multiplied by the evaluation of the outcome, motivation to comply, and perceived power (set to a scale of − 3 to + 3) provided a single datum from − 21 to + 21 for formative measures. For example, "I feel a sense of satisfaction when I buy a cause-related marketing product: (1) Strongly disagree to (7) Strongly agree" was multiplied by the response to "To me, experiencing a sense of satisfaction when I buy a product is: (not at all important (−3)/extremely important (+3))." The final datum for each item reflects the underlying belief after factoring in the weight (strength) assigned to each belief.

Data screening and measurement model
Before applying the measurement model, we screened the data for possible outliers and normality to fulfill the assumptions of the general linear model. The authors replaced 12 outliers with mean values. Skewness and kurtosis were analyzed and found to be within the acceptable range of ± 1.5 (Tabachnick & Fidell, 2013). Tabachnick and Fidell (2013, p. 80) state that non-strong skewness and kurtosis violations do not lead to statistical differences in large samples. They provided a calculation for estimating a large sample size of N > 50 + 8 m, where m is the number of independent variables. In this study, we had six independent variables; hence, 50 + 8 (6) = 98 suggests a minimum of 98 participants. Our sample size was larger than 98, ensuring that there was no impact of non-strong skewness and kurtosis violations in the study.

Confirmatory factor analysis
Fulfilling the assumptions of the general linear model paves the way for a measurement model. The authors assessed the measurement model using confirmatory factor analysis (CFA). This procedure was conducted using the IBM AMOS 26. The results indicated proper data fit (χ2/df = 3.335, NFI = .897, IFI = .925, CFI = .925, TLI = .914, RMSEA = 0.056). The observed value of Root Mean Square Error Approximation (RMSEA) was 0.056, which justifies the criterion of < 0.08 (Browne & Cudeck, 1993). The other fit indices (NFI, TLI, CFI, and IFI) were above the recommended criteria of close to 0.9 and higher (Bagozzi & Yi, 1988). All item factor loadings were above the recommended level of 0.60 (Chin et al., 1997).

Scale reliability
The internal consistency of the instrument subscales was measured using Cronbach's alpha. The values for all subscales, including formative, reflective, purchase intention, and purchase behavior, were above the highly reliable and acceptable level of 0.60 (Pallant, 2001). The composite reliabilities of the scale ranged from 0.746 to 0.913, above the benchmark of 0.70 (Bagozzi & Yi, 1988;Hair et al., 2010; See Table 6 for reliability results). Hence, we can establish construct reliability for each of the sub-constructs of the scale.

Convergent validity
We estimated the convergent validity of the scale using "average variance extracted" values (Fornell & Larcker, 1981). Although most of the AVE values are acceptable, control beliefs had values below the suggested values at 0.5, and higher (Fornell & Larcker, 1981; See Table 6 for convergent validity results). However, these two values are adequate because the composite reliability for the two constructs is higher than 0.6 (Fornell & Larcker, 1981).

Discriminant validity
We calculated each construct's square root of the AVE to ensure discriminant validity. We compared the results with the correlation values for each construct. The square root of the AVE of each construct was higher than its correlation value, ensuring discriminant validity (Chin et al., 1997; See Table 7 for discriminant validity results).

Structural model: model fit and hypotheses testing
We assessed the goodness-of-fit indices of the theoretical framework using structural equation modeling. Structural equation modeling was conducted using IBM SPSS AMOS 26. The results show that the proposed theoretical framework represents an acceptable data fit (χ2/df = 3.950, GFI = .879, IFI = .901, CFI = .901, TLI = .891, RMSEA = .0630). The observed value of Root Mean Square Error Approximation (RMSEA) was 0.056, which justifies the criterion of < 0.08 (Browne & Cudeck, 1993). The other fit indices (NFI, TLI, CFI, and IFI) were above the recommended criteria of close to 0.9 and higher (Bagozzi & Yi, 1988).

Hypotheses testing
We observed significant impacts of all belief components on the outcomes. Specifically, the regression paths from behavioral belief to attitude, normative belief to subjective norm, and control beliefs to perceived behavioral control were significant, supporting hypotheses H1, H2, and H3. Hypotheses H4, H5, and H6 were supported, as attitude, subjective norm, and PBC significantly influenced purchase intention. Hypothesis H7 was also supported as purchase intention significantly supported purchase behavior. However, Hypothesis H8 was not supported, as the relationship between PCB and purchase behavior was insignificant. These results show the acceptability of the TPB variables in determining consumers' intention to purchase a product associated with a CRM campaign (See Table 8 for the results of hypothesis testing).

Discussions and implications
In this study, Ajzen Planned Behavior (TPB) was employed to examine the formation of customers' intentions and purchase behavior towards products associated with a cause-related marketing (CRM) campaign. A comprehensive elicitation study was conducted involving 62 South Asian participants, who participated in focus group discussions and completed an open-ended questionnaire. The findings of this study identified salient beliefs that served as the foundation for the formative measures of the questionnaire. This study adapted established scales from prior research on green purchase behavior to measure reflective indicators, purchase intention and purchase behavior. To verify the dimensionality, reliability, and validity of the measurement scale and test the hypotheses, data were collected in three stages from 1035 respondents across South Asia. The results of the CFA showed that the measurement instruments confirmed the dimensionality of the TPB model with an acceptable model fit. The results also indicate a sufficient level of reliability and validity. Structural equation modeling indicated an adequate level of model fit. The results showed that all hypotheses, except the relationship between PCB and purchase behavior, are significant. The study's findings have important theoretical and managerial implications.

Theoretical implications
This study confirms the application of the Theory of Planned Behavior (TPB) to understand the factors influencing purchase intentions and behavior towards products associated with CRM campaigns. The results can be interpreted to show that positive evaluation of the behavior, perceived social pressure, and ease of performing a specific behavior enhance intentions to purchase a product associated with CRM campaigns. Hence, the classical TPB model was supported by the findings of this study. Specifically, this research supports the well-established sociopsychological model TPB in determining consumers' purchase intention and behavior towards products associated with CRM products in the context of developing nations in South Asia.
Specifically, the results indicate that the belief constructs (BB*OE, NB*MC, and CB*PP) positively influence attitude, subjective norms, and PBC, influencing purchase intention and purchase behavior. This result aligns with the core tenets of the TPB, which posits that underlying beliefs influence attitudes, subjective norms, and perceived behavioral control (Ajzen, 1991). These findings contribute to the existing literature by providing empirical evidence on the relationship between belief components and the corresponding constructs in the TPB framework. This finding is consistent with the results of previous studies. For instance, Askadilla and Krisjanti (2017) found that the three belief constructs positively affect the three predictors' attitudes, subjective norms, and PCB in predicting purchase behavior toward Green Cosmetic products. Similarly, Meng and Choi (2016) found that the three belief structures positively affected the three predators of intention in the case of participation in slow tourism. The results highlight the importance of individuals' beliefs in shaping their attitudes, subjective norms, and perceived behavioral control. This suggests that consumers' decision-making processes are driven by their underlying beliefs about behavior, social influences, and perceived control. This suggests that individuals' consumption choices are driven not only by functional attributes but also by their beliefs about the social, environmental, or ethical aspects associated with the products.
Second, constructs such as attitude, subjective norms, and PBC significantly influence purchase intention and strengthen purchase behavior towards products associated with CRM campaigns. The findings provide empirical evidence for the applicability of TPB in the context of CRM campaigns. They confirm that TPB constructs (attitude, subjective norms, and PBC) are crucial in shaping consumers' purchase intention and subsequent purchase behavior towards CRM products. The results are consistent with those of previous studies on green purchase intentions. For instance, Sun and Wang (2019) found that attitude, subjective norms, and PBC positively influence consumers' intention to purchase green products.
Positive attitudes towards CRM products contribute to stronger purchase intentions and an increased likelihood of actual purchase behavior, indicating the significance of attitude-behavior consistency. Subjective norms drive consumer behavior by influencing consumers' perception of what others consider appropriate or desirable in relation to CRM products. Perceived behavioral control, including factors such as ease, resources, and self-efficacy, influences consumers' confidence in CRM product purchases, and subsequently affects their intention and behavior. These findings emphasize the importance of considering attitudes, social influence, and perceived behavioral control in understanding and predicting consumer behavior towards CRM products.
The third finding revealed a significant impact of purchase intention on behavior. Individuals with strong behavioral intentions are more likely to engage in purchasing behaviors. However, other factors, such as practical constraints, lack of opportunities, and unanticipated events, can influence behavior (Ajzen, 1985). Most studies have utilized an individual's willingness to engage in a specific behavior, referred to as behavioral intention, as the dependent variable rather than the actual behavior. However, there have been a few exceptions, with mixed results. For instance, Canova et al. (2020) showed no correlation between intention to purchase organic food and future behavior (Canova et al., 2020). On the other hand, Caliskan et al. (2021) found a substantial connection between the intention to consume organic wine and actual consumption behavior (Caliskan et al., 2021). The significant impact of purchase intention on behavior suggests that individuals with strong intentions are likely to engage in actual purchase behavior. However, the influence of other factors and mixed results from previous studies highlight the need to further explore the intention-behavior relationship.
Fourth, this study found a statistically insignificant relationship between perceived behavior control (PBC) and purchase behavior. This can be attributed to PCB being a global measure and not belief-based measure. Hence, the respondent is expected to consider all possible factors that may enhance or diminish their control over the behavior to be performed. Also, since filling out questionnaires is a rather low-involvement activity for most people, it may be unreasonable to expect such high cognitive effort (Notani, 1998). Further the insignificant relationship could have resulted from the diverse samples used in this study. Individual differences within the sample might have contributed to a lack of consistency in the relationship between PBC and purchase behavior (Notani, 1998). Additionally, the specific context in which the study was conducted could have influenced the relationship between perceived behavioral control (PBC) and purchase behavior. This could be attributed to the presence of nonvolitional behaviors (Fishbein & Stasson, 1990). Nonvolitional behaviors are actions that are not consciously controlled or guided by intentional decision-making. In certain contexts, individuals may engage in purchase behavior without actively considering or exerting control over their actions. Factors such as environmental cues, social norms, or habituation can play a role in influencing these nonvolitional purchase behaviors. Therefore, the findings may indicate that the influence of PBC on purchase behavior might be affected by the presence of nonvolitional behaviors in the study's specific context.

Managerial implications
An individual's decision to purchase products associated with CRM campaign is based mainly on beliefs relating to their benefits, which leads them to their consumption. This is consistent with similar studies on consumer perceptions of the consumption of organic food, where consumers believe that it offers protection against future health risks (Liang, 2016). This belief strongly motivates them to choose organic over conventional food options. Similarly, in the context of products associated with CRM campaigns, consumers believe that purchasing these products positively impacts society or has a specific cause. This belief further encourages them to consume CRM products, aligning with their values and desire to contribute to social betterment. This belief motivates them to choose these products over the others. Hence, companies involved in CRM campaigns should aim to build positive consumer attitudes toward such campaigns. Marketers can achieve this by tailoring their advertising messages to satisfy a person's altruistic tendencies, regardless of motives, to encourage the generous act of participating in a CRM campaign. For example, when advertisements highlight acts of kindness, they positively impact people who see them. The ads increase happiness, improve well-being, and lead to an enhanced product or brand image. Similarly, communication of benefits to the community and social changes resulting from the campaign can be communicated for effective advertising outcomes. For example, advertisements can focus on how a campaign has positively affected the community by providing exciting statistics on previous achievements.
Additionally, communication can focus on the CRM campaign as the easiest way to donate towards an espoused cause. Furthermore, marketers should focus on an empathetic marketing approach. This can be achieved by presenting the sponsored cause as the central theme of advertising communication.
Second, the data analysis results indicate that subjective norm aids in increasing the purchase intention towards products associated with CRM campaigns. It can be interpreted that people's opinions, especially family and friends, influence such situations. From a practical perspective, developing a marketing strategy that allows customers to promote products associated with CRM to acquaintances is recommended. For example, empathetic videos that can go viral with a focus on the cause can be created as part of advertising efforts. Marketing managers can use emotional narratives about the cause that appeal to users, reflect the geniality and authenticity that builds trust, and address a relevant and timely issue to obtain the best response. Such advertisements might also lead to organic or word-of-mouth marketing among target groups.
Third, the results of the study indicate that PBC dramatically influences the purchase intentions of products associated with CRM campaigns. Hence, companies involved in CRM campaigns can focus on certain crucial aspects that build the perception of ease of buying such products in the target market. Customers often exposed to CRM campaigns tend to be skeptical about them (Bae & Wright, 2020). Hence, managers should first focus on transparency in advertising communications. This could be achieved by providing statistics on the work previously done to fulfill the espoused cause. Another means of accomplishing this is by providing the customer with a code for each product's sale. The customer could then use the code in the company's dedicated webpage to know about the impact created by the purchase made by the customer. Second, companies can maintain competitive pricing compared with competitors that are not involved in similar campaigns. Traditionally, customers have never paid more attention to products associated with CRM campaigns (Koschate-Fischer et al., 2012). However, maintaining competitive pricing ensures that the customer is not overcharged because of its contribution to the cause. Third, managers must maintain a competitive quality to ensure that customers do not compromise their quality because of their contribution to the espoused cause. Fourth, efforts should be made to conveniently make products available. Place utility addresses convenience, an element of the marketplace that is becoming increasingly important to busy consumers. Lastly, efforts could be made to create brand awareness, with the espoused "cause" as a positioning strategy to create differentiation in the marketplace. Such a positioning strategy will help the customer occupy a unique space in the minds of the customer and influence brand recall.

Conclusion, limitations, and scope for future research
The results of the data analysis revealed that all proposed hypotheses within the model, except for the relationship between PCB and purchase behavior, were supported. This study is the first attempt to draw a connection between the underlying salient beliefs and their influence on the constructs that lead to outcomes, including behavioral intention and behavior towards products associated with CRM campaigns. In this respect, the results of the current study are significant in their contribution to managerial decision-making regarding the design of successful CRM campaigns. As more companies adopt CRM in South Asia, the measurement instrument can successfully test behavioral intentions towards products associated with CRM campaigns in several contexts. This research also reveals crucial underlying beliefs that shape purchase intention towards products related to CRM campaigns. This may help develop content in a campaign that resonates with the target customer.
This study had certain limitations that should be addressed in future studies. First, the samples were collected in South Asia, so it is not easy to apply the results of this study to other regions. This study used self-reported behavior to measure consumers' purchase behavior towards products associated with CRM campaigns instead of actual behavior. Behavioral studies commonly use selfreported behaviors, as behavioral information can be easily collected and can help researchers investigate behaviors that may not be observed otherwise (Kormos & Gifford, 2014). Future studies should consider actual behavior instead of self-reported behavior.
The causes supported in a CRM campaign are categorized into four categories: human, health, environmental, and animal (Christofi et al., 2020). Each cause category and type may undoubtedly result in unique underlying beliefs. Hence, in future studies, researchers could use a specific cause category or cause type to understand specific beliefs in the context of interest. This study uses convenience sampling for quantitative research, which has its own disadvantages. In future studies, random sampling can be used among the population to obtain a generalizable assessment of consumers' purchase intention towards products associated with CRM campaigns.
From a theoretical perspective, TPB does not adequately account for external influences on behavior, such as social, cultural, and environmental factors (Terry & O'Leary, 1995). This could be overcome by combining this theory with Social Cognitive Theory (SCT), which focuses on personal and environmental factors affecting behavior. While TPB provides insights into behavioral intentions, scholars have accused the TPB of not always accurately predicting actual behavior (Sutton, 1998). Implementation intention theory can increase predictive capabilities in such situations. The TPB assumes that individuals make rational decisions based on their attitudes, subjective norms, and perceived behavioral control (Ajzen, 1991). Scholars can use Dual Process Theory to account for irrational or emotional factors in decision-making.

Convenient location
If I know what the cause is . . . The cause should be known to me . . .

Sales executives
Source: Authors own.
Study 2: Development and validation of the indirect, direct, purchase intention and purchase behavior questionnaire.