The rural consumer adoption of sustainable energy: a PLS-SEM-ANN approach of conceptual model development and cross-country validation of Pakistan and Malaysia

Sustainable and alternative energy sources of biofuel and solar power panel have been revolutionizing the lives and economy of many countries. However, these changes mainly occur in the urban areas and the rural population section has long been ignored by policy makers and government in the provision of energy. It is only recently that solar and biofuel are finally making in road to provide cheap and clean energy sources to rural population. As a result, literatures on consumer behavior of rural population towards sustainable energy sources are still very scarce. The present research aims to fulfill this gap by developing a conceptual model to investigate the adoption of solar power and biofuel energy resources in the cross-cultural setting of Malaysia and Pakistan. The data was collected from the rural areas of Pakistan and Malaysia. The two-stage data analysis method of partial least squares structural equation modeling (PLS-SEM) and artificial neural network (ANN) have been applied to satisfy both linear and non-linear regression assumptions, respectively. The results show that consumer in rural areas of Pakistan are willing and possess intention to adopt both biofuel and solar power for commercial and domestic use. Additionally, the results confirm that branding, economic, and altruistic factors are important in yielding intention to use towards biofuel and solar power panel in Pakistan which are validated by the results obtained in Malaysia. Other factors such as climate change awareness, retailer services quality, and ease of use are also important. The results offer wide-ranging theoretical and managerial implications.


Introduction
Sustainable energy solutions, such as solar and biofuel energy, have the potential to overturn disasters caused by the impact of environmental pollution (Michaelides 2012;Schlager and Weisblatt 2006;Kruger 2006). The production and consumption of solar and biofuel energy do not emit anything similar to greenhouse gases (Tuomisto et al. 2012). Furthermore, such energy solutions are also viable; as in the case of biofuel, it helps governments to reduce commercial, household, and agricultural wastes by recycling into energy (Demirbas 2008), thus creating twin benefits of waste reduction and energy production (Osman et al. 2019). Due to the many benefits of solar and biofuel energy, governments, corporation, and communities have been speeding up the process of adoption of such sustainable energy solution (Stančin et al. 2020). Governments around the world have been observed giving tax breaks, investing installation and technologies, and propagating the wider adoptability of both solar and biofuel energy solutions (Mollica and Balestieri 2020). By looking at its future impact upon their own profitability, corporations have also speed up production and distribution of these energy solutions to both retail and industrial consumers (Kath et al. 2020). Communities have seen dream coming true of having access to uninterrupted, cheap, and clean supply of energy (Mousavi et al. 2021;El-Houari et al. 2020). Despite the increasing rate of adoption, the hope of complete transition towards clean energy solutions, the likes of solar and biofuel, is still elusive (Naumann and Rudolph 2020), and thus, more research and innovation are needed to fill this gap to help the world to achieve complete transition.
Generally, the sustainable energy portfolio includes solar, biofuel, wind, and geothermal sources (Chandler 2009). The government and corporations are involved in both production and distribution of energy from these sources (Tzankova 2020). However, it has been observed that communities have also been adding into productionconsumption of clean energy mainly from two sources of solar and biofuel (Hoffman and High-Pippert 2005;Walker et al. 2010;Rae and Bradley 2012). The production-consumption model of solar and biofuel are the only sources of energy utilized by the rural communities in the emerging and newly industrialized countries (Vergragt et al. 2014). The production-use of sustainable energy among the rural communities is ranging from domestic household to commercial and economic use (König et al. 2020). It is widely believed that such transition especially among the rural communities will help alleviate socioeconomic backwardness, as most developmental budget in emerging and newly industrialized countries is being spent on the urban centers (Giampietro et al. 2013;Ouedraogo 2017;Zaman and Abd-el Moemen 2017). Despite the wide-ranging positive impacts of sustainable energy production and consumption, such phenomenon remains unclear as in regard to the factors that are propelling the adoption of sustainable energy among the rural communities. In addition, it is also unclear whether the basic motivation behind production-consumption is in powering the household or in utilizing for opening and running small-scale business which can be of important sources of income.
The research on sustainable energy from the behavioral perspective is gaining pace (Saleem et al. 2021;Kowalska-Pyzalska 2018a, 2018b. However, most of the researches focused upon the household and commercial users residing in the urban centers (Kowalska-Pyzalska 2019; Şenol and Zenk 2020;Villamor et al. 2020). But it has been observed that in some rural communities, the pace of adoption of production-consumption of sustainable energy is greater than that in urban centers (Han and Wu 2018) especially in areas where the access of electricity is not available to the communities (Malik 2020;Malik and Ayop 2020) and in rural places where consumers experience frequent electricity supply breakdown (Sapkota et al. 2014). Despite the huge potential, there is scarcity of literature which can help to understand consumer behavior of rural communities behind the adoption, production, and consumption of sustainable energy (Han and Wu 2018;Clausen and Rudolph 2020;Gorb et al. 2020;Barbier 2020). Therefore, the current research is attempting to understand the consumer behavior of rural communities in both emerging and newly industrialized markets cross-culturally (Ullah et al. 2021;Lau et al. 2020) by interplaying the relationship among the factors which lead to adoption of either solar energy or biofuel energy in both household and commercial activities (Asante et al. 2020;Sarker et al. 2020).
The current research proposes several factors which have not been studied both cross-culturally and in the context of rural communities which provide basic motivation to adopt, consume, and produce the sustainable energy (Iskin et al. 2013;Bourcet 2020). Most of such kinds of factors are associated with solar and biofuel energy (Kaur and Rehman 2020;Alipour et al. 2020;Leibensperger et al. 2021) as wind power and geothermal energy sources are expensive to adopt and install. The current research also attempts to understand the difference in effect between each factor on solar or biofuel adoption over household or commercial use (Alizadeh et al. 2020). Concurrently, the current research tries to understand the mediating role of biofuel or solar energy adoption between the range of motivational factors and the use of energy for household or commercial purpose (Jabeen et al. 2021;Makki and Mosly 2020). The current research building from works of published literature has constructed a proposed range of motivational factors for the adoption of sustainable energy and leading to its production and use for both commercial and household uses (Gong et al. 2020). These factors include branding factors (Manouchehrabadi et al. 2020), retailer services factors (Ali et al. 2020), altruistic factors (Ali et al. 2020), weather and lifestyle factors (Komendantova and Neumueller 2020), product and services awareness factors (Rahman et al. 2020;Simon 2020), economic factors (Whittle et al. 2020), and ease of use factor (Masukujjaman et al. 2021).
The objective of the current research is to bridge the gap in the current literature by providing empirical evidence on the consumer behavior towards the adoption, production, and consumption of sustainable solar and biofuel energy from rural communities of emerging and newly industrialized economies of Pakistan and Malaysia. The empirical evidence from rural communities in emerging economies, such as Pakistan, is highly limited despite of the wide-range adoption (Jan 2021). Additionally, the current research also attempts to contribute to the literature towards the understanding of the phenomenon through cross-cultural lenses (Sekaran 1983). Thus, the current research has collected and compared the data from rural communities of Malaysia which is a newly industrialized country with a relatively high rate of adoption-consumption of clean energy and Pakistan, a country with a growing population and an emerging economy. The rest of this paper is followed by literature review, methodology, data analysis, conclusion, and recommendations.

Economic value
Economic factor has greatly illustrated an economic value, based on the consumption experience of Ahn and Thomas (2020). Literally, economic factors are derived when consumers seek to achieve their consumption-related objective through pricing benefit or quality excellence (Ahn and Thomas 2020). The consumers have always been observed to achieve their objective from pricing benefits (Mendoza-Abarca and Mellema 2016), superior quality (Pappas 2017), low perceived risk (Wang et al. 2019), and product-related feature (Rausch et al. 2021). Therefore, it can be said that economic factors or values that consumers seek from solar energy or biofuel can become important determining factors in developing intention (Danish et al. 2019). The literature argues that consumer economic values have the potential to be satisfied from both forms of alternative energy products and services, i.e., solar power and biofuel, from a variety of perspectives. Both solar and biofuel energy are an alternative energy solution which offers big savings in terms of money which otherwise has to be paid to the energy services provider (Pizarro-Loaiza et al. 2021), savings in generator fuel expenses (Yasmin and Grundmann 2019), lower side effect in terms of hazards and health (Brown et al. 2020), consistent supply (Oh et al. 2018), consistent quality (Botelho et al. 2016), and environmental friendliness (Deora et al. 2022). Thus, it can be argued that consumer economic values are potential factors that need to be empirically tested in the crossrural area setting.
H1: Economic factors of consumer have a significant impact on the intention to adopt solar and biofuel energy solution

Altruistic factor
The altruistic factor of consumer can be described as an individual "concern for how my own consumption behavior affects others where this experience is viewed as a self-justifying end-in-itself as when engaging in ethically desirable practices in which virtue is its own reward" (Ahn and Thomas 2020). Altruism can have a role in accelerating the adoption of alternative energy products and services that should be tested. This is based on the notion that consumers may think that their decision to adopt and consume alternative energy is a morally responsible behavior (Deora et al. 2022). Consumers have developed the idea that alternative energy consumption is not only viable economically but also viable morally (Simon 2020) as an alternative energy solution has promised to protect our environment from degradation (Adedoyin et al. 2021) and will offer continuous and uninterrupted supply of energy (Oh et al. 2018) which will have a profound positive impact upon household, especially for females and children (Jia et al. 2017). Besides that, consumers often experience a kind of intrinsic value in using and adopting alternative energy solution upon being motivated by other users and consumers (Ahn and Thomas 2020) which will further motivate their decision in light of society's view (Simon 2020). Furthermore, consumer's altruism also plays a part in a form of indirect attitude (Andersson and James 2018) which stems from the fact that consumers, in order to satisfy the inner self, can start a propagation of benefits and positive aspect to other members of society (Panda et al. 2020;Sweeney et al. 2014). Therefore, based upon such assertions, the current research has proposed that.
H2: Altruistic factors of consumer have a significant impact on the intention to adopt solar and biofuel energy solution

Product and services awareness factors
The product and services awareness refers to the extent to which a consumer of rural area knows about biofuel and solar power energy products and services (Bailey 2005). The knowledge of product includes basic features of products (Parkins et al. 2018), functions (Rad et al. 2020, usability and usages (Lakshmi et al. 2020), main brands or manufacturer (Marselina 2019), and prices (Irfan et al. 2020) while services aspect of product includes retailers' promotion techniques of installing, repairing, and provision of any extra information that can enhance basic utility of product (Pokharel 2003;Mishra et al. 2019). Product and services awareness is crucial in developing intention and adopting the alternative energy products such as biofuel and solar power (Akroush et al. 2019). Governments, environmental activists, and corporations around the world are spending a considerable amount of resources, especially in rural areas, to create awareness of solar and biofuel products ) through experimental methods of providing demonstration of the products and distributing for free (Yadav et al. 2019). The literature further argues that the road to final adoption and consumption of the solar and biofuel runs from the awareness of such products (Al-Rahmi et al. 2019). Due to a higher level of education and more advanced mass media, urban consumers tend to be more aware than rural consumers of the alternative energy sources. Therefore, the current data on the role of products and services knowledge in developing intention of rural consumers to adopt either of the alternative energy solution of solar power or biofuel is limited. So, the current research hypothesized that.
H3: Product and services awareness factors have a significant impact on the intention to adopt solar and biofuel energy solution

Climate change awareness
The weather and lifestyle factors refer to the extent of weather condition in the area in which rural peoples are currently undertaking their livelihood and extent to which weather conditions induce necessary changes in the style of conducting such livelihood (Al-Ghussain 2019). The rising temperature or global warming due to greenhouse gas effect has a negative impact upon people's livelihood (Salleh 2016), and such impact seems to be more profound in the rural areas as compared to the urban areas (Devereux 2001;Singh 2020). The commercial mobility from a village to a market, working schedule in farms and fields, and health of farmers and rural people are some of the effects well noted in the literature (Muhumuza et al. 2018). Furthermore, it is increasingly raised and argued that the effects of global warming in rural areas due to the lack of supply of electricity through traditional grid are also responsible for aggravating the situation (Schoen et al. 2017). Thus, researchers and policy makers have long ago been calling for innovations and solutions that can provide the rural communities with energy supply from electricity and gases which will bypass traditional grid-based supply (Raza et al. 2020;Zheng et al. 2010). The reason is to make rural communities more self-dependent in energy production and consumption for the support of their overall livelihood (Rosenthal et al. 2018). The solar and biofuel energy solution finally fits into such scenario where it helps rural communities to produce and consume energy themselves without any interruption of energy breakdown (Rahman et al. 2017) and use such energy solution in daily lives (Yaghoubi et al. 2019).
H4: Weather and lifestyle factors have a significant impact on the intention to adopt solar and biofuel energy solution

Ease of use factors
The ease of use can be defined as a feature of product, services, and system to provide its user easiness in performing the function of product, services, and system safely, effectively, and efficiently, hedonically (Karahanna and Straub 1999). The level of difficulty further increases with a low amount of general awareness and literacy rate (Cao and Niu 2019) as in the case of rural users. The user manuals and guides accompanied with product are always helpful, but such manual and guides become ineffective in the case of rural consumers (Gök et al. 2019) due to low literacy rate. The literature has identified that both electronic and electric products and services have a problem of being considered not easy to use (Kardooni et al. 2016), and it is the reason that many of innovative and effective products are difficult to be adopted by consumer widely due to their lack of ease of use (Al-Rahmi et al. 2019). However, considering solar and biofuel as electric and energy products, it has been both observed in the literature and practiced that solar power panel and biofuel are not intricate in terms of its installation and functions (Bandara and Amarasena 2018;Nehrir et al. 2011). The health hazards as literature indicates are always a central concern of consumers as for electric and electronic product are concerned (Chandel and Agarwal 2017) and consumers tend to pay a huge amount of money to vendors for helping them to install and guide them on its proper functioning (Acharya and Marhold 2019). But, it has been noticed that such factors are virtually invisible in the solar power and biofuel. The consumers easily install, maintain, and repair energyproducing products without paying extra amount of money to vendors (Extance 2019), and there are not any health hazards been found while installing and using solar and biofuel products and services.
H5: Ease of use factors have a significant impact on the intention to adopt solar and biofuel energy solution

Retailer services factors
The retailing or retailer services factors are referred as the extent to which a retailer is providing sales and marketing services to consumers of both solar power panel and biofuel (Lee and Chow 2020). The services by retailers include information on solar panel and biofuel products and brands, information on associated biofuel and captive products and services, easy access to vendor for installation and repairing, shipping of products from retailers' location to site of installation which can be home or commercial place, and any other additional information which can add value to consumer perception (Palmeira and Musso 2020). The literature suggests that retailers play a crucial role in enhancing consumer perception of value of products and brand (Lee et al. 2019). In the case of brand extension of line and launching a completely new product, company's manager depends upon retailer to generate the demand for the products (Eom and Kim 2018). This is also a case with innovative and radical new product. One of the important reasons of such an effective role of a retailer in consumer demand generation and satisfaction is that it is the retailer who is a bridge between company and consumer and retailers have direct contact with consumers (Lee et al. 2019). Corresponding to such assertion in the literature, the role of a retailer in the alternative energy solution of solar and biofuel is also important and similar (Luthra et al. 2015;Jelti et al. 2021). The importance of role increases further in the rural areas where a retailer is the only source of information and promotion.
H6: Retailer services factors have a significant impact on the intention to adopt solar and biofuel energy solution

Branding factors
The branding factors refer to an attempt by an equipment manufacturer and distributor of solar power and biofuel to brand it and its products and services offered to the consumers (Larsen 2017). The branding process or effort includes developing necessary elements of brand such as name, logo, color, theme, graphics, sound or jingle, packaging, and tagline (Moriarty and Franzen 2008). The literature on the branding is one highly developed (Hao et al. 2019). The empirical evidences have shown that branding is a critical factor in issues such as consumer purchase intention (Martín-Consuegra et al. 2018), consumer adoption (Al-Hajla et al. 2018), consumer loyalty (Van der Westhuizen 2018), and firm's financial success (Crass et al. 2019). The literature further enlightens us that branding is closely associated with consumer psychology (Vogel and Watchravesringkan 2017). The consumer psychological issues such as high perceived quality of products and services (Aakko and Niinimäki 2021), high perceived utility (Winzar et al. 2018), higher perceived value (Kim et al. 2019), and lower perceived risk are well mitigated through development of strong brand (Kirchoff et al. 2019). The branding activities have also been observed actively in the sustainable and alternative energy solutions (Larsen 2017). The term "green brands" is occurring in the literature frequently, and empirical evidences suggest us that branding activities in products related with environmental sustainability such as solar and biofuel have a significant and positive impact upon consumer perception and behavior (Zubair Tariq 2014). The evidences on relationship between branding and intention are well recorded in other green products except solar and biofuel, and most of these evidences are based upon urban samples. Thus, it is needed to calculate the effect of branding factors on intention and adoption in rural phenomena. H7: Branding factors have a significant impact on the intention to adopt solar and biofuel energy solution

Intention of solar power panel
The intention of consumer is referred here as a prospect of consumer who will purchase or he/she is willing to purchase a solar power panel in the near future (Ramayah et al. 2010). The past researches have been indicating that intention is always closely related to the final purchases and adoption of product and services (Peña-García et al. 2020). The intention is a powerful construct in the literature of both consumer and human behavior. As it is referred in the theory of planned behavior (TPB), people always develop an intention of adopting or discharging the certain and specific behaviors (Ajzen and Sexton 1999). Therefore, it is argued here that for final adoption and usage of solar power panel whether for the purpose electrifying home or commercial business, intention towards solar power panel is a mediating factor (Kumar et al. 2020). Previous researchers have employed a range of factors that can be instrumental in yielding the intention towards the solar power panel (Parkins et al. 2018), and many have been attracted towards the TPB which argues that attitude, subjective norms, and perceived behavioral control are responsible for developing intention generally (Liu et al. 2020). Those factors enshrined (Liu et al. 2020). Although factors enshrined in the TPB are important, empirical evidences are available in support of it within the context of solar power panel (Yadav and Pathak 2016). The current research argues that, due to the nature of solar power products itself and understanding and comparing the adoption of solar power in the context of domestic and commercial purposes, the range of other factors can also be an influence which include branding factors (Gong et al. 2020), retailers services factors (Manouchehrabadi et al. 2020), altruistic factors (Ali et al. 2020), weather and lifestyle factors (Komendantova and Neumueller 2020), product and services awareness factors (Rahman et al. 2020), economic factors (Simon 2020), and ease of use factor (Masukujjaman et al. 2021) upon intention of consumer towards solar power panel and, finally, to the domestic and commercial adoption. Thus, the current research hypothesizes that.
H8: Intention to adopt and install has a positive effect towards adoption of solar power for home H9: Intention to adopt and install has a positive effect towards adoption of solar power for commercial use H10: There is a mediating effect between factors and adoption of solar power for home H11: There is a mediating effect in between factors and adoption of solar power for commercial use

Intention of biofuel energy
The intention of consumer is referred here as a prospect of consumer who will purchase or he/she is willing to purchase a solar power panel in the near future (Ramayah et al. 2010). The past researches have been indicating that intention is always closely related to the final purchases and adoption of product and services (Peña-García et al. 2020). The literature has indicated a variety of factors employed to yield the intention towards the biofuel or biogas products and services (Yaghoubi et al. 2019) and with an obvious role of factors to yield the intention as provided in the TPB (Chin et al. 2019). Although factors enshrined in the TPB and by other researchers are important, empirical evidences are available in support of it within the context of biofuel or biogas and are strong (Yaghoubi et al. 2019;Chin et al. 2019;Bakhtiyari et al. 2017). The current research proposes to operationalize a range of factors that needed be tested such as branding factors (Gong et al. 2020), retailer services factors (Manouchehrabadi et al. 2020), altruistic factors (Ali et al. 2020), weather and lifestyle factors (Komendantova and Neumueller 2020), product and services awareness factors (Rahman et al. 2020), economic factors (Simon 2020), and ease of use factor (Masukujjaman et al. 2021) with intention of consumer towards biogas and biofuel which can help us to understand biogas and biofuel adoption towards both domestic and commercial uses. Thus, the current research hypothesizes that.
H12: Intention to adopt and install has a positive effect towards adoption of biogas and biofuels for domestic use H13: Intention to adopt and install has a positive effect towards adoption of biogas and biofuels for commercial use H14: There is a mediating effect between factors and adoption of biogas and biofuels for home H15: There is a mediating effect between factors and adoption of biogas and biofuels for commercial use

Theoretical review of adoption
The concept of adoption whether household or commercial adoption which is addressed in the next sections is derived from product adoption (Rogers 1976). The concept of new product adoption refers to a process where a consumer of new and older products started to get aware of product, understand its value, and finally start to use it for both commercial and domestic purposes where applicable (Rogers 1976;Kalish 1985). The innovation diffusion theory (Rogers 2010) which has been undergoing certain changes and modification has been a key in explaining the product adoption by consumers (MacVaugh and Schiavone 2010). The empirical researches and evidence on the product adoption under the lens of innovation diffusion theory are wide ranging which have been instrumental in understanding product adoption (Yuen et al. 2021;Zanello et al. 2016). The empirical evidences on the relationship between intention and adoption of the new or old product are also wide ranging where researchers have been employing a range of factors in the intention yield (Ahn and Park 2022;Islam 2014). Thus, it can be argued based upon the published literature that theoretical relationship exists between intention and adoption.

Household adoption of biofuel and solar
The product adoption of alternative energy solutions such as solar power panel and biofuel has remained an issue for policy makers and researcher's community (Solangi et al. 2011). As researchers and policy maker have been arguing to fight poverty and other socioeconomic issues with instrument of energy inclusiveness (Yadav et al. 2019), alternative energy solutions such as solar power panel and biofuel are playing an effective role (Giampietro et al. 2013;Zaman and Abd-el Moemen 2017). The solar and biofuel products can be adopted for domestic uses or in home uses for a variety of different purposes such as electrifying their homes, clean source of cheap sources of cooking, and different other uses (Aklin et al. 2018). The home-based adoption of solar power panel and biofuel is also increasing day by day due to the perceived usefulness and value of solution (De Felice et al. 2019). The dynamics of solar power panel and biofuel adoption in rural areas are a bit different from those in urban areas. The researches published in the literature are mostly focused upon the urban residential consumers. The rural and cross-cultural aspect of residential or domestic solar power panel and biofuel adoption is clearly missing. Furthermore, as both solar power panel and biofuel are different in nature, functioning, and price, evidences are needed to understand which product among solar power panel and biofuel is at priority of final adoption in the rural areas.

Commercial adoption of biofuel and solar
The alternative energy products such as solar power panel and biofuel have a potential to help societies to achieve a socioeconomic objective through energy inclusiveness and supply for commercial purpose (König et al. 2020). It is argued that people especially at rural areas can generate income and become employed for themselves with low-cost and clean energy solutions (Sun et al. 2021). The variety of business opportunities have been tapped by people in rural areas by using the solar power panel and biofuel solution (Gandini et al. 2021). This include completely removing electricity charges for small-scale retailer shops where electricity tariffs are higher as compared to domestic tariffs (Khan et al. 2019). Carpenter, shoe maker, barbers, tailor, and many others use solar power panel and biofuel for a small-scale production facility (Boakye et al. 2020). Furthermore, solar power panel and biofuel also contribute in the home-based venture run by women specially. However, the rural and cross-cultural aspect of commercial solar power panel and biofuel adoption is clearly missing. Furthermore, as both solar power panel and biofuel are different in nature, functioning, and price, evidences are needed to understand which product among solar power panel and biofuel is at priority of final adoption in the rural areas. The current research is a step to further develop an understanding of adoption (Fig. 1).

Research design and approach
The purpose of the current research is to establish a causal relationship between critical factors related to biofuel and solar power panel and intention to adopt that can lead to final adoption of solar power and biofuel for either domestic use or commercial use. The factors which are affecting the intention that can yield adoption are derived from previous literature and theories such as innovation diffusion (Rogers 1995) and theory of planned behavior (Ajzen 1991). Therefore, by looking at aims and objective of research, the current research has proposed a quantitative research design (Gregar 1994). The definition and operationalization of variables (factors, intention, and adoption) have been done from past literature. The current research has employed a survey questionnaire as a tool of data collection which is mostly used by previous research in studying adoption of alternative energy (Rehman Khan and Yu 2021). The unit of analysis for the current research is cross-cultural rural consumers of Pakistan and Malaysia. The decision of cross-cultural analysis of data is rooted in the literature gap. The current research proposes a two-stage method of data analysis using structural equation modeling (SEM) and artificial neural network (ANN) (SEM-ANN) (Hayat et al. 2020). The SEM is a famous tool of data analysis by an energy researcher of consumer behavior, but its results capture mainly the linear characteristic of the consumer behavior, and in practice, it is not always the case. Thus, ANN is proposed to be used as a non-linear data analysis tool to complement the results of SEM (Curry et al. 2002).

Rationale of the cross-cultural sample
The cross-cultural analysis of Pakistan and Malaysia is based upon the literature gap. The countries selected for the current cross-cultural analysis is based upon the distinction of  (Zainoddin et al. 2017;Israr et al. 2017). The major objective of their policies is uplifting rural populace from socioeconomic backwardness and achieving sustainable economic growth. Thus, the comparison between emerging and newly industrialized economies is more logical than the comparison between developed and newly industrialized or developed and emerging economies.

Population and sampling
The population of the current study includes the rural population of Pakistan and Malaysia which is 67.5% and 24%, respectively (Government of Pakistan 2019; World Bank 2019). The current research selected one state, province, and division within that state and province of Malaysia and Pakistan which is home to a majority of rural population where government and rural communities are taking part in the adoption of alternative energy consumption for both domestic and commercial uses. Table 1 illustrates the selection of state, province, and division within. The current research has employed a purposive non-probability sampling technique to collect the data. The reason of employing purposive nonprobability sampling is the scope and objective of research and researcher's judgment (Sharma 2017). The total sample drawn from cross-cultural population is approximately 400 (Saunders et al. 2009). Furthermore, consumer behavior is highly dynamic and cultural difference is one of the biggest sources of such dynamism. Thus, collection and analysis of a cross-cultural sample allows understanding of behavior which is mired by dynamism (Cavusgil and Das 1997). The literature shows an increasing amount of researcher in marketing, and other related fields prefer cross-cultural samples as these broaden the research objective and provide a better understanding of phenomena (Chen et al. 2015;Zhao et al. 2021).

Data collection and instrumentation
The current research has employed the survey questionnaire as an instrument of data collection. The questionnaire is developed at the 5-point Likert scale ranging from strongly disagree to strongly agree (Nemoto and Beglar 2014). The measure on each variable is developed from previous studies. Table 2 illustrates the number of items used to measure each of variables in the conceptual framework and sources from where items are adopted. Corresponding to a strategy of purposive sampling, the questionnaire is distributed among the participants who have knowledge of solar power and biofuel and who can understand English language with support from the researchers in case of difficulty. The verbal demonstration of the questionnaire in local language had also been given to the participants.

Data analysis
The current research has employed the two-stage method of data analysis combining SEM and ANN. The SEM is a famous tool of data analysis (Hair et al. 2011) by an Household and domestic adoption 6 each Li et al. (2015); Shabbir (2010) energy researcher of consumer behavior. However, it is more suitable and often used to analyze the linear behavior of the consumer, and in practice, consumer behavior always appears to be non-linear. To overcome this difficulty, we employ ANN for an additional analysis. Thus, ANN which is a non-linear data analysis tool is proposed to complement the results of SEM (Leong et al. 2020). The current research proposed to employ variance-based SEM which is also called as partial least squares SEM (PLS-SEM). The PLS-SEM will be used for calculating the measurement model for the purpose of establishing the validity and reliability of the instrument and data itself and the structural model for the purpose of testing the hypothesis of the current study (Hair et al. 2011). Secondly, the current research proposes to employ the forward backward propagation type of ANN which is based upon sigmoid function (Leong et al. 2020).
The neural network will be consisting of three layers: input, hidden, and output layers. The input layer is consisting of an independent variable, the output layer is consisting of a dependent variable, while the hidden layer performs the function of transformation of non-linearity between input and output layers (Kubat 2017). The software packages employed for conducting data analysis are the SmartPLS 3.4 for PLS-SEM and IBM-SPSS 21.0 for ANN and demographic analysis.

Empirical analysis
The current research has employed the two-stage data analysis method (Leong et al. 2020) where hypothesis is tested with the help of PLS-SEM techniques, and results obtained from PLS-SEM are further complemented with the help of ANN. The results are presented in the following sections.

Construct validity and reliability
The reliability of data collection instrument refers to the internal consistency of items with their respective construct, and the validity of data collection instrument refers to assuming that items of each construct are only measuring their relative phenomena rather than other (Price et al. 2015). The construct reliability in the present research has been tested through tests of Cronbach's alpha (CA) and composite reliability (CR), and the construct validity is tested through tests of average variance extracted (AVE) (Cheah et al. 2018). According to the literature of PLS-SEM, the construct validity can be assumed when each construct reports a value of 0.70 on CA and CR while the construct validity can be assumed when each construct reports a value of 0.50 on AVE (Hair et al. 2018). The results presented in Table 3 show that the data collection instrument used in collecting the data from both Malaysia and Pakistan has achieved both construct validity and reliability.

Discriminant validity: Pakistan and Malaysia
The test of discriminant validity is one of the important tests to assume the validity in the instrument of data   (Henseler et al. 2015). To assume discriminant validity, the Fornell-Larcker value of each construct must be higher on their own as compared to other constructs (Hair et al. 2018). The discriminant validity tests are presented in both Table 12 (Pakistan) and Table 13 (Malaysia) in the Appendix. The result shows that each construct has achieved discriminant validity and measuring separate phenomena.

Explanation of variance
The variance in the present research model is explained through R 2 which is also referred as a coefficient of determination. The R 2 is the proportion of variance in the dependent variable explained or is contributed by each of independent variables (Glantz et al. 1990). The R 2 is determined before testing a hypothesis, and statistical literature concludes that for a dependent variable, it is necessary that all of independent variables must contribute a variance of above 0.20 or 20% (Cohen 1992). The results shown in Table 4 clearly indicate that each of dependent variables in the model has a higher R 2 value than what is accepted which is 0.20. Thus, it is concluded that the present research has explained an appropriate and acceptable level of variance for hypothesis testing.

Model fitness
The model fitness of the research model refers to how well a set of predictor fits on the observation (Stephens 2017). The software package SmartPLS 3.0 offers to measure the goodness of fit of the research model through various indices, and one of most widely used model fit indices being followed for PLS-SEM is called standardized root mean square residual (SRMR). According to the literature of PLS-SEM, a SRMR value less than 0.10 or 0.08 is considered as a model achieving goodness of fit (Hair et al. 2019). The results shown in Table 5 confirm that models of the present research have achieved the goodness of fit as the value of SRMR is less than 0.10 or 0.08.

Structural model
The present research has employed bootstrapping based upon a procedure of 5000 subsamples to test the hypothesis (Hair et al. 2019). The present research is cross product and cross-cultural where a set of independent variables is hypothesized as a factor of yield intention towards two green energy cross products, i.e., biofuel and solar power energy in a cross-country rural market area of Malaysia and Pakistan. Therefore, to have a better visualization of results, the present research model is break down into four models. The structural models for each product, namely biofuel and solar power energy, are developed separately for each country, namely Malaysia and Pakistan. Table 4 shows the summarized results of structural models of biofuel intention towards household and commercial uses for both Pakistan and Malaysia. Table 6 shows the summarized results of structural models of solar power energy intention towards household and commercial uses for both Pakistan and Malaysia.  Table 4 shows the cross-cultural results of biofuel intention towards household and commercial adoption. The results show that, for both Pakistan and Malaysia, there is a positive effect of biofuel intention towards both household and commercial adoption. But, as far as factors of biofuel intention are concerned, for Malaysia, two factors (easy to use and product awareness) are insignificant with biofuel intention, while for Pakistan, four factors (climate change, easy to use, product awareness, and retailer services) are insignificant with biofuel intention. Table 7 shows the structural model results of solar power intention for both Pakistan and Malaysia. The results show that for Malaysia, only one hypothesis developed between product awareness and solar intention is rejected, while for Pakistan, four hypotheses (impact of climate change, easy to use, product awareness, and retailer services quality on solar power intention) are rejected.

Artificial neural network analysis
The ANN analysis is incorporated into the current study to predict the importance of each of independent variables and dependent variables (Teo et al. 2015). The ANN incorporates a non-linear relationship between exogenous and endogenous variables, and it is robust in predicting the relationship with both non-normal data and outliers (Taneja and Arora 2019). The current research is a cross-cultural and product analysis of household and commercial adoption. The current research has applied four ANN models in each country. The first two ANN models are based upon factors and intention towards biofuel and solar intention and third and fourth models towards commercial and household adoption of each country. The feed forward backward propagation (FFBP) with sigmoid function is used to train and test the ANN algorithm with the tenfold cross-validation approach to avoid an oversimplification bias in the ANN model (Leong et al. 2020).

ANN model validity
The current research has operated a tenfold ANN analysis to avoid oversimplification of the ANN model and to obtain ANN model fitness (Leong et al. 2020). The root mean square error (RMSE) indices are calculated for both training and testing data of the ANN model.  The RMSE is calculated ten times as ANN is operated tenfold, and the average of RMSE is calculated at the end. Table 8 shows the average RMSE value for different models on the data of Malaysia, and Table 9 shows the average RMSE value for different models on the data of Pakistan. The literature on RMSE statistics concludes that a RMSE value within a range of 0 and non-negative values shows the perfect fit of model, especially ANN (Heravi et al. 2004).
The RMSE results of both testing and training data of each model from Malaysia and Pakistan as shown in Tables 8 and  9 show that RMSE values of all models are within 0 and non-negative values. Therefore, it is concluded here that the ANN models used for the current study are fit statistically. The ten times RMSE calculation results of different ANN models are attached in Tables 14, 15 , 16, 17, 18, 19, 20, and 21 in the Appendix. Tables 8 and 9 only depict the average of RMSE of each model.

Sensitivity analysis
The sensitivity analysis is conducted to measure the relative importance of each input neuron or factors used in the ANN model on output neuron or factor. The results of sensitivity analysis are calculated as an average of tenfold application of the ANN model. The results are presented in Tables 8 and 9.

Sensitivity analysis: intention
The sensitivity analysis results of cross-cultural and cross product purchasing intentions are shown in Table 10. The results show that the branding factor is the most important factor in the purchasing intention of both biofuel and solar in both Pakistan and Malaysia. The factors which come after branding as far as relative importance is concerned are economic and altruism factors in every model. However, the relative importance of factors such as climate awareness, easiness in use, product awareness, and retailer services awareness changes in different models.

Sensitivity analysis: adoption
The sensitivity analysis results of commercial and household adoption of biofuel and solar power energy are presented in Table 11. The results show that solar power energy has more relative importance on commercial adoption (95.82%) as compared to household adoption (85.19%) in Malaysia, while biofuel has more relative importance on household adoption (98.21%) in comparison with commercial adoption (76.76%). Furthermore, results of Pakistan show that solar power has more relative importance (95.5%) as compared to commercial (84.62%), while biofuel has more relative importance on commercial (90%) as compared to household (81%).

Discussion
The present research attempted to study the cross-cultural consumer behavior of rural population in Pakistan and Malaysia towards the two of the most innovative energy products of the present age, i.e., solar and biofuel energy. The outcome of this study is interesting from both managerial and theoretical perspectives. The results of the present study conclude that both biofuel and solar power panel products have become increasingly important sources of energy for household and commercial activity purposes in rural areas of Malaysia and Pakistan. The results of PLS-SEM from Malaysia have concluded that two of factors, namely ease of use (P = 0.494) and product awareness (P = 0.742), do not have any significant relationship with the intention to consume energy from biofuel sources. However, all other factors such as branding (β = 0.571, P = 0.000), altruism (β = 0.288, P = 0.005), climate change awareness (β = 0.194, P = 0.037), economic (β = 0.169, P = 0.005), and retailer services (β = 0.202, P = 0.029) factors have a significant relationship with biofuel intention. The ANN results show that of the accepted hypothesis, branding, economic, altruism, climate change, and retailer services factors are important in determining the intention towards biofuel energy sources. The results conclude that biofuel energy providers need to focus much upon the branding of their products with pricing and lower economic cost at heart of such branding campaign. The results of PLS-SEM-ANN also call for the promotion of consumer altruism through various means to increase and yield purchasing intention towards biofuel. The study has confirmed the notion that climate change awareness is a real factor in purchasing intention of goods which promises environmental sustainability. Lastly, retailer services of biofuel can also play a vital role in increasing the intention towards the biofuel. Therefore, energy supply promotion to the consumers should include the aspect of retailers as well. The results with the data collected from Pakistan are different from those from Malaysia. The PLS-SEM results show that climate change (P = 0.605), ease of use (P = 0.467), product awareness (P = 0.637), and retailer services quality (P = 0.091) do not have a significant impact upon the biofuel intention. But, branding (β = 0.520, P = 0.000), altruism (β = 0.220, P = 0.044), and economic (β = 0.168, P = 0.005) factors have a significant impact upon the biofuel intention. The results of ANN are also in line with PLS-SEM where branding is the most important factor followed by economic and altruistic factors. The current literature is highly limited on reasons to support such inconsistency of results with theory. In view of the difference in development status and social culture between Pakistan and Malaysia, such inconsistencies could be attributed to various factors such as education, income, and others. Thus, more research is needed to explain these phenomena. As far as consumer's intention towards solar power is concerned, The PLS-SEM result on data from Malaysia concludes that only product awareness (P = 0.139) has an insignificant impact upon the solar power purchasing intention. The other factors such as branding (P = 0.000, β = 0.558), economic (P = 0.006, β = 0.168), altruism (P = 0.014, β = 0.262), climate change awareness (P = 0.000, β = 0.192), ease of use (P = 0.027, β = 0.228), and retailer services quality (P = 0.040, β = 0.189) all have a significant relationship with the solar power purchasing intention. The ANN results also correspond with those of PLS-SEM. Therefore, it is concluded that focusing on branding, economic, altruism, ease of use, climate change awareness, and retailer services will respectively yield purchasing intention among the rural communities in Malaysia. The results from Pakistan on solar power purchasing intention follow the same pattern of biofuel intention. The PLS-SEM results show that climate change (P = 0.449), ease of use (P = 0.500), product awareness (P = 0.759), and retailer services quality (P = 141) do not have a significant impact on solar power purchasing intention. On the other hand, branding (β = 0.512, P = 0.000), altruism (β = 0.205, P = 0.046), and economic (β = 0.169, P = 0.005) factors have a significant impact on the purchase intention for this product. The results of ANN are also in line with those of PLS-SEM where branding is the most important factor followed by economic and altruistic factors. More research is required to understand the inconsistent aspect of the Pakistani consumer behavior towards solar power purchasing intention.
Lastly, the results of PLS-SEM from Malaysia have demonstrated that solar power has a positive impact towards household (P = 0.000, β = 0.552) and commercial adoption (P = 0.000, β = 0.524), with a similar pattern of results for biofuel towards household (P = 0.000, β = 0.552) and commercial adoption (P = 0.000, β = 0.508). The data results of PLS-SEM collected from Pakistan also show similar results in which both solar and biofuel intentions have a positive and significant impact towards household and commercial adoption. Therefore, it is concluded here that consumer purchasing intention towards biofuel and solar power panel can be translated into actual adoption of biofuel and solar power energy sources for commercial and household use purpose by the rural communities in Pakistan and Malaysia. The results of ANN reveal the cross-cultural differences with reference to commercial and household adoption of biofuel and solar power panel. The results show that commercial adoption of solar panel is higher among Malaysian rural consumers than households in Pakistan. Furthermore, biofuel has higher acceptance for household use among Malaysian rural consumers while it has higher acceptance for commercial use among Pakistani rural consumers. Thus, it is concluded that both solar power and biofuel energy are an important mix of sustainable and alternative energy resources which the people in the rural areas of Pakistan and Malaysia are keen to utilize for both commercial and household purposes.

Conclusion
The present research has studied the cross-cultural behavior of rural consumer towards sustainable energy products, specifically on biofuel and solar power panels. The current research studied the behavior of rural consumers towards biofuel and solar power panels. Previous studies have been focusing on consumer behavior towards sustainable energy products and solutions by employing famous theories such as theory of planned behavior, technology acceptance model, and others. But, there is a lack of literature which asserts the consumption of sustainable energy is different from other kinds of goods and services. Sustainable and alternative energy sources are based upon a model of producing and consuming electricity without any intermediaries, especially government. Thus, a gap in the literature exists which calls for studying such consumer behavior from a very different perspective. So, the current research, guided by the results of the published literature, developed and tested a number of factors which can help to enhance the intention towards sustainable and alternative energy sources of biofuel and solar power panel. Furthermore, it has been seen from scientific and other sources of literature that alternative and sustainable energy sources have good potential for both domestic and commercial purposes. However, no scientific study has been conducted from this perspective. Thus, the current research has also incorporated and compared the prospective domestic and household adoption. Lastly, sustainable and alternative energy sources are already driving an energy revolution in the rural areas. However, the rural population in many countries, including Pakistan, has long been overlooked by the authorities on energy supply. Thus, an alternative and sustainable energy source offers an important solution to energy access for both commercial and domestic purposes. Various reports have pointed out that sustainable and alternative energy sources are reviving rural economy. Therefore, the current research offers a valuable insight on this perspective.
The current research has employed PLS-SEM-ANN to satisfy both linear and non-linear relationship assumptions. The results of data analysis have confirmed that both solar power and biofuel, the two products studied in the present research, have a significant and positive impact on both commercial and household adoptions. The results show that biofuel is more favored for commercial use in the rural areas of Malaysia while solar power panel is more preferred in Pakistan for commercial purpose. For domestic consumption, on the other hand, biofuel is preferred by the rural population in Pakistan, while in Malaysia, the choice is solar power. The results of PLS-SEM-ANN have shown that branding, economic, and altruism are powerful factors that can drive intention towards both biofuel and solar power which further leads to household and commercial adoption. The other factors such as climate change awareness, ease of use, and retailer services quality are also helpful in yielding the intention in Malaysia only. The findings of this research offer a variety of practical and theoretical implications.

Managerial implication
The present research offers three practical implications. First, the current research has confirmed from empirical evidences that consumption of solar power and biofuel is growing in rural areas of both Pakistan and Malaysia. Therefore, energy investments in this aspect need to design the marketing framework that will incorporate the needs and wants of rural consumers. Secondly, the present research offers an important insight regarding both biofuel and solar power panel sources. The research results confirm that the rural consumers are willing to adopt both biofuel and solar power for commercial and domestic use. Thus, brands that offer solar or biofuel solution should design their promotion campaign to focus on both segments of consumers. Finally, the present research suggests that providers of biofuel and solar energy solution should also take into consideration the elements of economic factors in their promotion to the rural sectors. The results of the present research corroborate the findings of earlier studies that economic factors are instrumental in yielding the intention of consumers.

Theoretical implications
The present research offers four implications for theory of innovation diffusion and planned behavior. First, while most researchers have been relying upon a linear method of regression, the current study has employed the PLS-SEM-ANN method of data analysis to satisfy both linear and nonlinear regression assumptions. Second, the current research has analyzed cross-culturally Malaysian and Pakistani crossproduct biofuel and solar power data and cross-segment domestic and household data. The results offer an interesting insight regarding the adoption of sustainable alternative energy where it is confirmed that both the theory of planned behavior and innovation diffusion are popular in such analysis. Third, the present research is conducted in the context of rural areas. Currently, the literature on sustainable and alternative energy in rural area context is highly limited. Thus, the current research contributes to the theory by providing empirical evidences. Lastly, the current research confirms that branding, economic, and altruism are important factors in yielding intention towards both solar power and biofuel energy in Malaysia and Pakistan.

Future research recommendation
The preset research offers various future research recommendations. First, although the present research has attempted to bridge a gap in the literature by collecting cross-cultural, cross-product (solar and biofuel), and rural data, as far as a wider generalization of results is concerned, future researchers can collect the data from other areas in both Pakistan and Malaysia. Secondly, both Pakistan and Malaysia are Asian countries in which a comparison between Asian and African countries would provide valuable solutions. Third, experimental setting which can employ both within-subject and between-subjects designs can help to better illustrate the causal effect of variables studied in this research. Finally, the ANN method was very helpful compared to other methods such as fuzzy-set qualitative analysis (fSQA) and can be used in future research to address both linearity and nonlinearity in the research. Tables 12,13,14,15,16,17,18,19,20,21     Data availability The data will be made available on request to corresponding author.

Declarations
Ethics approval and consent to participate We hereby give an approval that we have followed all scientific ethical guidelines while conducting our research. Consent to participate is not applicable.

Consent for publication
We hereby give consent to publish this paper in this journal.

Competing interests
The authors declare no competing interests.