Perceived risk factors assessment: during pandemic via digital buying

Purpose – Thepresentstudyisintendedtoassesstheriskfactorsassociatedwithdigitalbuying.Alsoaims to design and develop an instrument to assess the digital buyers risk factor score (DBRFS) in light of pandemic. Design/methodology/approach – Present investigation uses a quantitative approach to achieve the stated objectives. The survey instrument for the purpose of assessing risk factors associated with digital buying was developed in two phases. The present study adopts theory of planned behaviour (TPB), built based on the theory of reasoned action (TRA). The data were collected and analysed considering 500 valid responses, samplingunit being digital buyers usingsocial mediaplatforms in tyre-IIcity ofIndia. The datacollection was undertaken between June 2021 and August 2021. The instrument is designed and validated using exploratory factor analysis (EFA) followed by confirmatory factor analysis (CFA). Findings – The present research identified six perceived risk factors that are associated with digital buying; contractualrisk,social risk, psychological risk, perceivedquality risk,financialrisk and timerisk. The DBRFS of male is 3.7585, while female is 3.7137. Thus, risk taking by the male and female is at par. For the age group 15 – 30, DBRFS is 3.6761, while age group 31 – 45 noted as 3.7889 and for the 46 – 50 age groups it is measured as 3.9649. Practical implications – The marketers are expected to have the knowledge about how people responds to the pandemic. The outcome of the research helps to understand consumer behaviour but disentangling consumer ’ s “ black box ” is challenging especially during global distress. The present study outcome helps the digital shopkeepers to respond positively to meet the needs of digital buying. Originality/value


Introduction
The Internet and growth of technology across the world has changed the global business.Business operations started shifting away from traditional to advance digitalized processes.These digitalized processes gave a further boost to the e-commerce industry, making the digital environment more competitive.Despite the growing trend, there has always been a consumer market that is not involved in digital buying (Sajid et al., 2022), and this gap is huge when it comes to consumers from emerging countries, specifically India.On contrary, prolonged lockdowns resulted in responses to the pandemic causing closures of several companies.However, it brought a new wave of online shopping all over the global market.Interestingly, when businesses went bankrupt and started the closure of their processes, the online market thrived and expanded by over 30-50%.This development brought drastic changes to the way consumers used to form their intention and behave toward digitalized solutions in pre COVID-19 times.Evidence shows that the global e-commerce industry has touched phenomenal growth during COVID-19.Digital platform (Dahiya and Gayatri, 2017) revolutionized the buying process (Snowaney and Chincholkar, 2019).
The digital marketing (DM) is a system of direct marketing that links digital buyer with seller electronically using collaborative technical platforms namely emails, websites, digital forums and newsgroups, television, mobile communication (Kotler and Armstrong, 2008).The coronavirus intensely transformed the global trends (Rana et al., 2021).Due to the restrictions levied by respective governments during pandemic, digital buying has become imperative and order of the day (Ali et al., 2021).The pandemic outbreak at the beginning of 2020 had changed the ways and perception of consumers buying, subsequently marketers found it challenging to understand the digital buying intensions and to provide digital services.It is noted that there are changes in behavioural pattern of digital buyers especially in COVID-19 pandemic.Purchase behaviours among social platform users are interesting and promising research topic.
A digital platform's characteristics can positively or negatively affect buyer's intentions.These changes attributes to perceived risk factors that are associated with digital buying.The unplanned changes influence the buying intention during pandemic.The distress of complete lockdown, limited supply and scarcity of necessary products which led to panic buying compelled the impulse buying patterns (Ahmed et al., 2020).The health, safety concern, imposed restrictions, financial condition and other realities caused a notable change in digital buying behaviour (DBB) (Islam et al., 2021).The effects of trust, customer ethnocentrism, service quality and perceived risk impacted customer loyalty in the COVID-19 pandemic.The DBB is influenced by factors like risk vulnerability, severity and risk of penalties.The marketers are expected to learn lessons from COVID-19 pandemic, knowledge about consumer responses towards digital buying (Fihartini et al., 2021).It is required to report changes occurred during pandemic in information reliability, risk perception, attitudes about policy-making and communication with public with reference digital buying.Thus, the present study aims to assess perceived risk factors associated with DBB.Also attempts to quantify the perceived risk score of digital buying.

Sources on risk factors associated with digital buying
The influence of perceived usefulness, ease of use and risk factors towards digital buying matters a lot in making buying decision (Iriani and Andjarwati, 2020).The customers due to perceived risk of infections and ease of digital buying have considered to be major factors which have been examined to understand the change in DBB during the COVID-19 outbreak (Grashuis et al., 2020;Alaimo et al., 2021;Hadler et al., 2021;Kim, 2021;Prasad and Srivastava, 2021;Truong et al., 2022).Rashed Alhaimer (2022) investigates the various risk factors like risk-susceptibility, severity and formal penalties affecting DBB in Kuwait.In COVID-19 situation, as the pandemic progressed due to vaccine rollout and financial situations, the change in DBB is observed.Understanding these risk and other contributing factors is essential to predict the future of economy, especially business models for e-tailers (Nguyen et al., 2021;Almajali et al., 2021).It is better to understand these changes in DBB and associated risk factors and its antecedents (Di Crosta et al., 2021).The researchers (Nguyen et al., 2021;Almajali et al., 2021;Neger, 2020) look into issues of trust, perceived ease of use and perceived risk in other country context like Vietnam, Jordanian and Bangladesh during COVID-19.
The key factors of digital buying intention such as product delivery, cyber laws, shipping fees and communication lead to diverse behavioural changes.The intersection between COVID-19 risk factors and digital inequalities suggested to mitigate the risks associated with the pandemic (Robinson et al., 2020).
There is a need to examine the role of perceived risk of digital purchase intention (Qalati et al., 2021).The digital risk factors like product, delivery and information security risk affects digital buying during pandemic even in other countries, Im et al. (2021).The socialpsychological reactance, people freedom, interaction, work place, performance, trust and security have an impact on DBB (Naeem Akhtar et al., 2020).The antecedents associated with digital buying (hedonic motivation, satisfaction, perceived quality, financial policy, product features, information quality, economic situation and trust) which considerably influence DBB during the lockdown needs to be examined (Koch et al., 2020;Al-Hattami, 2021).Gloster et al. (2020) revealed that there is a significant association between buying digitally before and after the appearance of the COVID-19.Nurunnabi et al. (2020) restated that prioritizing mental health, anxiety and coping strategies along with psychological effects is necessary as socio-psychological pressure is acute during pandemic.Adapting coping strategies which include problem-solving, social and emotional support to promote psychological well-being, Chinna et al. (2021).The pandemic results anxiety, sensitivity-related physical concerns and contamination leading to distress, Baloch et al. (2021).Gu et al. (2021) proved that there is a need for understanding the speed of decision making by digital buying when purchasing goods and services online during pandemic.Valaskova et al. (2021) mentions that as a consequence of different restrictions in COVID-19 pandemic, digital buying and their shopping patterns have changed significantly; thus factors that influences new purchase patterns need to be identified to support retailers and marketers to develop appropriate strategies.Moon et al. (2021) states that pandemic has caused chaos since the end of 2019 and hence there is a need to study the aspects of digital buying' shopping.Islam et al. (2021) mention that pandemic has triggered substantial modifications in the habits of digital buying all over the world which led to key changes among digital buying.Given the pandemic and safety concerns, customers have considered various options for shopping (Grashuis et al., 2020;Alaimo et al., 2021;Hadler et al., 2021;Kim, 2021).As the pandemic progressed, shopping behaviour continued to change depending on possible factors related to COVID-19 situation, vaccine rollout and financial situations.Understanding these contributing factors is essential to predict the future of our economy, especially business models for retailers.The literature remains silent to explore relationship between perceived risk factors associated with digital buying although there is rapid upsurge in digital shopping.
Psychological impact such as seeking social support and acceptance coping strategies were significantly associated with the level of anxiety that prevailed during pandemic, Kamaludin et al. (2020).In addition, due to the risk of infection and possible death also exerts tremendous anxiety level pressure among university students, Khoshaim et al. (2020).The digital technology and psychological artificial intelligence solutions shall be implemented to manage anxiety levels of students, Sundarasen et al. (2020).The critical risk features of digital buying was examined considering 322 responses adopting an instrument consisting of 5 items with 5 factors, Sangameshwara et al. (2022) which needs to be studied further to confirm and validate the risk factors.
Given this background, the above mentioned literature provides the evidence on influence of pandemic across globe however; the present study is intended to assess the risk factors that affects DBB in tyre-II city of India.Therefore, present research aims to design and develop an instrument to assess the perceived risk of digital buyer's in light of pandemic and to quantity the digital buyers risk factor score (DBRFS).

Theoretical foundation for the study
The present study adopts the well proved theory on the consumer behaviour to study the digital buying during the pandemic situation.The variables considered for the current study is based on theory of planned behaviour (TPB), built based on the theory of reasoned action (TRA) ( Ramus and Asger Nielsen, 2005).TPB helps to predict an individual's intention (considering the risk factors associated with DBB) to engage in a behaviour (buying) at a specific time (pandemic) and place (Tier II city).The researchers have considered the above underpinning theory and framework while selecting the risk variables for the present study.

Research question
RQ1.What are the perceived risk factors associated with digital buying during pandemic?
RQ2.How to quantity digital buyers risk factor (DBRF) in light of pandemic?

Research objective
The purpose of this study is to analyse the perceived risk factors associated with DBB and to ascertain the risk factor score (DBRFS) in light of pandemic in tyre-II city of India.

Methodology
Present investigation uses a quantitative approach to achieve the stated objectives.The survey instrument for the purpose of assessing risk factors associated with DBB was developed in two phases as detailed below; Phase 1initial instrument development and field test Instrument development Survey instrument for the purpose of assessing perceived risk factors associated with digital buying was developed in phase 1 which is keen on developing questions to obtain feedback.Precaution was taken to evade ambiguity and double-barrelled questions.The initial instrument was shared with limited group of digital buyers through a convenience sample to obtain feedback.Further, five-point Likert-scale is used for the survey ("Strongly Disagree" to "Strongly Agree statements").The data were gathered during second wave of COVID-19 pandemic in tier-II city of Karnataka State, India between June-August 2021.The instrument consists 46 items of five drivers (to enhance the credibility of the designed instrument, minimum of five items were included for each driver.The preliminary instrument development and field test was conducted with 150 respondents via email survey.The exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) are used to validate the instrument using statistical package for the social sciences (SPSS) software and through analysis of moment structures (AMOS).The model fitness is verified.

Data analysis
The collected data were examined for any outliers and other irregularities.The CFA is adopted to examine the structure that is proposed.The objective of adopting CFA was to examine the relationship among the latent and manifest variables.The goodness-of-fit indexes are considered to evaluate proposed CFA.This investigation used the ratio of chi-square fit statistics/degree of freedom (CMIN/DF) 5 2.528, which shall be within the threshold value of 5 (Hair et al., 1998).The other fit indices include, goodness-of-fit index (GFI), adjusted goodness of fit index (AGFI), incremental fix index (IFI), Tucker-Lewis index (TLI), normed fix index (NFI) and comparative fix index (CFI) which are expected to be closer to unity, an indication of effective model fitness (Bentler, 1992) and error approximation value root mean square error of approximation (RMSEA) shall be less than 0.09, which is within the acceptable range (Hair et al., 2006).The TLI nearer to unity indicates that there is a strong relationship between covariance and variance (Schreiber et al., 2006) and TLI value greater than 0.90 shows acceptable model fitness (Hu and Bentler, 1999).

Results
The results of initial analysis (phase-1) CFA shows that the model proposed does not exhibit goodness-of-fit (CMIN/Df is 2.528; TLI 0.582; RMSEA 0.101).As the model found to confirm poor-fit, the proposed model in phase-1 is not detailed.Thus, it was desired to evaluate possible alternate structures of the data.
To attain this, EFA was used.The EFA helps to explore underlying factors without compelling a preconceived structure (Child, 1990).This six-factor model which includes risk factors like contractual risk (CR), financial risk (FR), psychological risk (PR), perceived quality risk (PQR), social risk (SR) and time risk (TR).As the six-factor model appeared to be promising, it was deemed essential to further investigate and test the model with a larger sample.

Instrument refinement
The output of phase 1 model (Six-factor) was considered as a base for updating the instrument.Further, at this stage 3 items included in the original sub-scale titled feedback in social media, advertisement in digital media, compute or mobile skills for digital buying were removed.The supplementary items were constructed to update the framework that was theorized.The revised questionnaire has 25 statements (see Table 3).Instrument is constructed using five-point Likert scale.To enhance the credibility of the designed instrument, minimum of five items were included for each driver.The instrument development and field test was conducted with 500 respondents.The EFA and CFA are used to validate the instrument using SPSS software, and through AMOS and model fitness is verified.

Data analysis
The KMO value is 0.827, indicates that the sample size is sufficient (Kaiser and Rice, 1974).Barlett's test of sphericity was hypothesised to know the relationship (strength) amongst grouped items.A small significance (less than 0.05) indicate that the data is suitable to conduct CFA (Tobias and Carlson, 1969), test results show that the p-value is less than 0.05, indicating that data is suitable to conduct EFA.To identify risk factors that influences on digital buying, to reduce the number of components and items, EFA was conducted using Varimax rotation method using principal component analysis (PCA) followed by CFA.The results are discussed in below section and tabulated in Table 1.
Based on the Eigen value in EFA six factors were extracted whose Eigen value is more than one and based on the relevance grouped factors were tilted as: Contractual Risk (CR); Financial Risk (FR); Psychological Risk (PR); Perceived Quality Risk (PQR); Social Risk (SR) and Time Risk (TR), items loading more than 0.5 were consider for the better model fit Stewart (1981).For the present study, the item loading ranges from 0.540 to 0.911, it is represented in below Table 2.

Confirmatory factor analysis (CFA)
To verify the extracted six factors along with 24 items, CFA was carried out using SPSS and AMOS software.In the CFA analysis all, the six factors along with 24 items were confirmed.
Followings are the six confirmed factors The proposed digital marketing risk factor measurement model has adequate model fitness; it is represented in Figure 1.According to Bentler (1992) and Bentler and Bonett (1987) all the fit index should be nearer to unity, error factors should be less than 0.08 and CMIN/DF should be less than 3 for the better measurement model.The proposed fit indices are CMIN/DF 5 2.334, within the threshold value of 5 (Hair et al., 1998).The other fit indices include, GFI 5 0.893, AGFI 5 0.859, IFI 5 0.932, TLI 5 0.917, NFI 5 0.887 and CFI 5 0.954 are very close to unity,

PRR
which is an indication of effective model fitness (Bentler, 1992) and error approximation value RMSEA is 0.062, which is within the acceptable range (Hair et al., 2006).The outcome of the CFA model is represented in Figure 1.The dimensions of the perceived risk factors associated with digital buying were identified as: contractual, financial, psychological, perceived quality, Social Risk and Time Risk.Thus, revised CFA with larger sample (n 5 500) was considered to understand perceived risk factors associated with digital buying during the pandemic.The constructs and dimensions of risk factors is presented in Table 3 below.
The below section details the perceived risk factors associated with digital buying.Contractual Risk-Taking account of the socioeconomic characteristics, technical characteristics of the commodity and the institutional environment are part of contractual arrangements in buyer/seller relationship.The enforcement of consumer rights through contractual agreement is vital for digital business.The digital brand trust and contractual risk are an antecedent to digital buying (Katrodia et al., 2018).The performance and contractual risk, trust and security have a significant impact on digital buying (Shahzad, 2015).
Financial Risk -The fear for loss of money, non-delivery and improper return-policy have depressing effect (Georgie, 2021).The FR has a negative effect on the intention to buy (Shahri Mejarshin et al., 2021).
Psychological Risk-The threat of health crisis, scarcity of products, fear of unknown, negative emotions and uncertainty leads to PR, Wu et al. (2020).
Perceived Quality Risk-The purchase intention was influenced by perceived quality and brand trust (Taufik et al., 2021).The seven quality dimensions are: reliability, accessibility,

Rotated component matrix
Perceived risk factors assessment ordering services, convenience, product content, assurance and credibility (Sebastianelli et al., 2008).
Social Risk factors-Attitude, subjective norms, scarcity, healthy habit of social media use, strong social ties positively influence customers' panic buying intention during COVID-19 pandemic.
Time Risk-TR-Include spend more time and effort to make decisions.Also leads delay in getting the product.Cost being deducted but order not placed The financial details might not be protected Refund initiated by the supplier but not credited Unauthorised charges on returns/taxes/ transportation Disclosure about price of product may be misleading Contractual factor Risk of guarantee/warrantee claim is confusing Kariyawasam and Guy (2008), Teo and Yu (2005), Soni andVerghese (2018), Dorward (2001), Grabner-Kraeuter (2002), Sreya and Raveendran (2016), Shahzad (2015), Bhatti et al. (2019), Korgaonkar and Karson (2007) Risk of fulfilment of Warrantee/guarantee conditions Risk of return/replacement of products in case of product failure Risk of after sales service/maintenance work Risk of reselling the product Psychological factor Digital buying reveals individual's social presence Rasty et al. (2021), Ko et al. (2004), Brewer and Sebby (2021) The quality of product and images are misleading Digital buying process does not finish one shot as Source(s):

Perceived risk factors assessment
Assessment of risk factor score for digital buyers Predictive model: regression analysis To assess the DBRFS, an attempt is made to develop the linear regression equation to estimate the score based on factors confirmed by second order CFA; CR, PR, FR, PQR, SR and TR.The findings and analysis is mentioned in below section in the form of equation.
Thus, by substituting coefficients of dependent variables, DBRFS may be obtained.The respondents were classified based on the demographic factors (age and gender).The model summary of the identified six factor regression model is mentioned Table 4.
Table 5 shows the analysis of variance (ANOVA) test statistics for the proposed six factor model, the significance level of the proposed model is 0.000, it clearly shows that none of the proposed six factors coefficient of regression is zero, that is non-zero correlation between the six factors.
Table 6 shows the regression coefficient and standardized coefficients for the proposed six factors regression model.
By substituting independent variable's mean value in the regression equation, as per descriptive statistics  Therefore, overall risk factor associated with digital buyers is 3.8325.From this it is shows that for the five points scale the risk factor score is 3.8325, that is risk factor in digital buying for the respondents are 76.65%.

Relationship between components of risk factor
To verify the magnitude and direction of the association among the components of risk factor in digital buying, Pearson correlation analysis was conducted, correlation coefficient (r) among components is represented in Table 7.
Pearson correlation analysis shows that there is a strong (r 5 0.429) correlation among PQR and CR.There is a less (r 5 0.014) correlation between PQR and FR.
The Risk Factor Score of male is 3.7585, while the same with female is 3.7137, thus, risk taking by the male and female is at par, it is represented in Table 8.In case of age wise analysis, for the age group 15-30 DBRFS is 3.6761, while for the age group 31-45 it is 3.7889 and in case of age group 46-50 it measures as 3.9649.Thus, it may be interpreted as the risk taking by the age group 15-30 is lesser than the age group 31-45 but 46-60 is comparatively higher during the pandemic, it is represented in Table 9.

Discussion and conclusions
The present research uses digital platforms including Facebook, Twitter, Linked-In and YouTube and Instagram for the study which in line with other researchers (Dwivedi et al., 2021;Chopra et al., 2021;Chaudhary et al., 2021).The digital marketers have to understand the DBB and decision making pattern.Since digital marketing emerged, there has been a huge transformation Perceived risk factors assessment in the marketing process especially during pandemic due to distress (lockdown, threat of the health crisis, product scarcity, fear and negative emotions and panic buying) which compelled traditional buyers to opt digital buying.This new trend in marketing has provided equal opportunities to businesses irrespective of size, challenges and complexity in which digital marketers have to operate.However, risk associated with digital shopping needs consideration.
The present research identified six perceived risk factors that are associated with digital buying during pandemic; CR, FR, PR, PQR, SR and TR which affects digital buying.Sreya and Raveendran (2016) also reconnoitred the same along with physical and performance risk.The present study also identifies that CR is an important risk factor.The various factors associated with CR are perceived transaction cost, uncertainty, dependability of digital stores, frequency, buyer-seller relationships, socio-economic characteristics of each party, economic and institutional environment.
As identified by present research, FR negatively affects the digital buying intention which was echoed by (Katrodia et al., 2018;Rana et al., 2021).The product and non-delivery risks negatively affect the intention of digital buyers.The perceptions of service quality and relationship-marketing are important factors in digital retail environment.From the above discussion, initiation was to be made by e-commerce companies to mitigate critical risk factors such as performance risk and FR (Guru et al., 2020).
The outcome of the present study was also prevalent among Korean and US buyers (Ko et al., 2004).The psychological and SRs were affects digital buying of South African consumers (Pentz et al., 2020).Yuen (2020) observed phenomenon of panic buying due to the fear of illness, empty shelves, price increase, social inclination, threat of the health crisis, product scarcity, fear and negative emotions in Singapore which was also witnessed in our study context.The association of consumer-perceived quality and purchase intention influences digital buying (Qalati et al., 2021), however, SR is found to be one of the factors from the present research that affects the digital buying.Saini et al. (2022) acknowledged the significance of product risk and privacy risk during the COVID-19 along with the present study findings.It is desirable to quantify risk factors associated with digital buying.
DBRFS in light of pandemic.The multiple regression equation is used to measure DBRFS.The DBRFS is measured (age and gender wise).The risk factor Score of male is 3.7585, while female is 3.7137, thus, risk taking by the male and female is at par.In case of age wise analysis, for the age group 15-30 DBRFS is 3.6761, while age group 31-45 it is 3.7889 and in case of age group 46-50 it measures as 3.9649.Thus, it may be interpreted as the risk taking by the age group 15-30 is lesser than the age group 31-45 but 46-60 is comparatively higher during the pandemic.Thus the present research findings conclude that during pandemic irrespective of gender there are no significant differences towards risk taking ability of digital buyers due to the distress caused by pandemic.

Practical implication
The study has important theoretical and managerial implications for improving digital buyer's decision model.Firstly, market researchers have been trying for decades to understand consumer behaviour but disentangling consumer's "black box" is challenging especially during global distress, our study provides further evidence on the consumer decision model for the understanding of consumer behaviour in the digital marketplace with special attention on risk associated with digital buying.Secondly, we suggest digital shopkeepers to identify the presence of their product/services in multiple social media platforms to find greater opportunity to engage with customers and showcase their brand.With more than 2 billion users worldwide, Social platforms are a perfect communication tool for sellers to convey their value to buyers and actively respond to buyers' needs and enquiries.Thirdly, digital sellers could participate in commercial communities that are regulated and trusted by community members, conducting social interactions with members so that they develop trust with sellers through continuous interactions, these members feel less exposed to risk.

Limitation and scope for further research
The present study is mainly focused on risk factors because it is imperative for social platforms wishing to monetize through s(social)-commerce to reduce risk.However, the study can be extended to other influential factors, such as merchandising approaches and online group buying.Detailed studies may be conducted to reveal key push and pull factors surrounding the use of the internet for search activities using updating technologies to facilitate purchase behaviours (cloud-based shopping apps and artificial intelligence (AI)driven chatbots).

Table 1 .
Total variance explained

Table 2 .
Table created by authors

Table 4 .
Model summary