Moderator Trust, Subjective Norms Influence Risk And .

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020considered it, risky way of shopping because the seller isabsent in this type of shopping [14]. Therefore, onlineshopping is a big challenge for consumers [2]. Also beside,online shopping is the same process as traditional shoppinglike problem identify, search information, evolution, buyingand post buying [15]. In present time consumer don notwant to bargain so they avoid bargaining but after all theseconveniences still consumers face a lot of risks duringonline buying such as financial risk, product risk,convenience risk, and privacy risk, so it is very important tostudy this area of research to explore.2.2 Financial RiskRisk is a sensitive impassive state-run that cannot measuretangibly. Financial risk is a primary risk considered in onlineshopping and this risk plays an important in the decisionmaking of the consumer to buying. Financial risk denotesthe possibility that a single buy online agonizes financialloss in a monitory term as he/she compensated more andmerchandise has less value [16]. Any financial loss eitherproduct does not accomplish up to expectations, badquality, fraud of credit card reduces the online shopping[17]. This risk reveals the most domineering though onlineshopping [18-20]. Moreover, financial risk is a premier fearin the consumer minds in money at buying time [21]. Thereare some studies demonstrate that financial risk does notaffect in defining the online shopping behavior [10, 22].There are inconsistent findings between financial risk andonline shopping behavior. Hence, in upcoming studies needto explore this relation [2, 23-25].H1: Financial risk has a significant negative influence ononline buying behaviorH2: Subject norms moderates between financial riskand online buying behaviorH3: Trust moderates between financial risk and onlinebuying behavior2.3 Convenience riskConvenience risk is directly associated with the buyer’smind while they go to buy something on the internet. Thisrisk is consumer perception at the time of buying throughthe internet will take a long time to reach [14]. Furthermore,when consumers think to buy online and they perceiveconvenience risk is high then consumers reluctant to buyonline. In other words, it is considered as time loss becauseit is time taking process to find the right product andcomparing one brand’s product with other brands products[26]. Meanwhile, it also involves wrong delivery, latedelivery, face problem during order placing, languageproblem all these things make consumers irritating. InPakistan only 1% of people who can operate computers.This risk effect behavior of consumer significantly due tolack of education and literacy rate. All these issues generateproblems and threats in the consumer mind [27]. Researchreveals that convenience risk influences adverse on onlineshopping behavior [26, 28]. Literature tells that conveniencerisk expressively decreases online shopping behavior [22].Notwithstanding this, literature determines that conveniencerisk does not play a role in influential online shoppingbehavior [29].H4: Convenience risk has a significant negativeinfluence on online buying behaviorH5: Subject norms moderates between ConvenienceISSN 2277-8616risk and online buying behaviorH6: Trust moderates between Convenience risk andonline buying behavior2.4 Privacy riskPrivacy risk refers to a situation when consumers perceivethat they will lose their personal information and misuse of thatinformation without any permission. This risk is also thehighest level of risk. In Pakistan, 97% of people prefer to buycash on delivery because they don’t feel secure. There arevarious kinds of privacy such as information, bodily,communication, and territorial privacy. Privacy risk significantlyeffect on online shopping behavior once consumer face riskduring online buying they reluctant to buy and avoid next time[30]. Furthermore, consumer doesn’t want to share theirname, address, credit card number, contact number becausesome retailer’s brows consumer’s information share with otherretailers and people feel insecure [31]. Online shoppingdepends on the security of personal information [32].Literature demonstrates that privacy risk significantlydecreases online shopping behavior [10, 17]. Meanwhile,some studies show that there is no relation between privacyrisk and online shopping behavior [22, 33]. The findings areinconclusive and unclear. Therefore, still, need to study thisrelationship.H7: Privacy risk has a significant negative influence on onlinebuying behaviorH8: Subject norms moderates between Privacy risk and onlinebuying behaviorH9: Trust moderates between Privacy risk and online buyingbehavior2.5 Product RiskProduct risk refers to a condition where consumers arecontingent on the information that retailer online and there is acoincidental to suffer the defeat low-quality product [6].Product risk is an insufficient potential loss to examine theproduct because in online shopping people cannot examinethe exact quality of the product, cannot touch productphysically, and sometimes product looks good in picturesmore than actual, so due to these issues people avoid to buyonline [34]. In other words, product risk indicates that theproduct fails to perform the expected performance [35].Researchers show that product risk is the utmost imperativerisk and cited the main reason to hesitate online buying.Furthermore, this risk directly linked with the decision makingof consumers but not directly communicate with retailers onlyonline communication made for transactions [36]. Also beside,research reveals that product risk significantly influences ononline shopping behavior [37]. In other words, the adverseinfluence of risk on online shopping behavior [17, 26, 38]. Incontrast, no influence of risk [2]. The results are inconclusiveand need to explore.H10: Product risk has a significant negative influence ononline buying behaviorH11: Subject norms moderates between Product risk andonline buying behaviorH12: Trust moderates between Product risk and online buyingbehavior2.6 TrustTrust refers to the perception of consumers about onlineretailers trustworthy [39]. In other words, trust in the online628IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020buying situation is customer’s alacritous to trust in retailersand take a decision in conditions where these movementsmake the customer susceptible to the retailer[40].Meanwhile, it’ buyer beliefs at another party. Trust is amainly a serious factor in online situation and customerdoes not have a direct control over the activities of the retail.Furthermore, traditional shopping is quite different than theonline buying hesitation and uncertainty committed. Lack oftrust is a big hurdle for consumers to online and retailers toattract and retain consumers as well [41].3ISSN 2277-86164. RESEARCH DESIGN AND METHODOLOGYThe researcher gives much attention to the methodologyportion in any type of research. It considers an essential partof examining the objectives. An appropriate method used tosolve the practical and theoretical problem. In the presentstudy, to solve the research problem, research objectives, andresearch nature, we used a quantitative method andquestionnaire survey, furthermore, the study was crosssectional, andTHEORETICAL FRAMEWORKSubjective NormsTrustFinancial RiskConvenience RiskOnline Shopping BehaviorPrivacy RiskProduct Risk.H13: There is a significant and positive relationship betweentrust and OSB2.7 Subjective NormsSubjective norms refer to those factors that relate torelatives, family, and friends in buying products and services[43]. In other words, subjective norms determine theperceived stress enforced by others like friends, family,neighbors, peers, etc who influence your behavior indirectlyor directly. The rationale for subjective norms is that peoplecan choose a certain thing and perform certain behavior ifone or more than one important reference thing they shoulddo. Moreover, it is documented in the vital social and peopleprefer to review others’ views and experience aboutparticular products or services and highest views productsthey want to buy [44]. The presence of a supportiveenvironment including friends, family, neighbors, peersincreases the likeliness of buying. if the particular product isuseful then they also suggest to others and they alsocogitate that product [45]. Subjective norms play asignificant role in the decision making of consumers andinfluence consumer behavior, so we cannot ignore it indetermining behavior [46].H14: There is a significant and positive relationship betweensubjective norms and OSBMoreover, buyer feeling is very important to consider theirintentions to buy. Hence, trust is significantly predictingconsumer behavior. Trust plays an important role in onlinebuying because it gives confidence to customers on retailers[42]. In addition, trust reduces uncertainty and enhance selfreliance. Literature reveals that trust establishes a crucialpsychosomatic constriction on online shopping and trustbecomes imperative with esteem to online shoppingdeductive. After data collection uses Smart PLS 3.2.8 to testthe proposed hypothesis4.1.1Data Collection MethodIn this study data collected by a questionnaire surveytechnique from students that are studying in Universities inPunjab, Pakistan. 9 universities were selected for survey whohad students more than 15 thousand. The questionnairespread only to those students who have the experience toshop online.4.1.2Questionnaire developmentA theoretical model of present the study involves sevenvariables measured by using items that were adapted fromprevious studies. The questionnaire divided into two parts, thefirst part consists of respondents’ demographics and thesecond part is consisting of four risks, subjective norms, trust,and online shopping behavior. Meanwhile to measure this partfive-Likert scale used that range agree to strongly disagree (51). In this study financial risk, product risk measured by 7items, items adapted from [27]. Convenience risk computes by9 items, that adapted from [27, 47]. privacy risk 6 itemsadapted from [48]. subject norms measured by 6 items [49].trust 5 items adapted from [50] and online shopping behavior17 items [27, 47, 51, 52] adapted from previous studies.These 17 items cover in the study of [47, 51, 52].4.1.3 Population and samplingThe population is considering the whole collection of things,entities, and events that they want to examine. In this study,population comprises universities students of PunjabPakistan. Simple random sampling practice used because itgives generalized results.629IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 20204.1.4Sample SizeIn the current study 600 sample sizes. According toresearch 30% sample size cogitate excellent for study,furthermore, sample size 5 to 10 times should greater thanvariables. Meanwhile, sample size should be greater than30.4.1.5Demographics profileAs mentioned in Table 1. The overall sum of respondentswas 550, in which 257 (46.7%) were female and 293(53.3%) were male. In education respect 10 (1.8) wereDiploma, 224 (40.7) were Bachelor degree, 280 (50.9) wereMaster’s degree, 24 (4.4) were Ph.D. and 12 (2.2) wereConstructOnline Shopping ExperienceGenderEducationCredit/ Debit CardMarital StatusSelf-SupportedAgeAverage Monthly Incomeothers. In perspective of credit/ debit card 38 (6.9) people hadcards but other 512 (93.1) without card. In marital status 508(93.1) were single, 39 (7.1) married and 3 (.5) were divorced.In self-supportive cases 54 (9.8) were self-supportive and 496(90.2) were dependent to others. In the case of age 190(34.5) students were 16-20 years, 327(59.5) were 21-25years, 9 (1.6) were 26-30 years, and 24(4.4) students weremore than 31 years. In income section 470(85.5) had noincome, 18(3.3) had 1-5000, 16(2.9) students had 5001 –10000, 16 (2.9) students had 10001 – 15000 and 38(6.9) hadmore than 15000.Demographic Characteristics of the Respondents (N 550)Table 1.Table 1CategoryYesNoMaleFemaleDiplomaBachelor DegreeMaster DegreePhDOthersYesNoSingleMarriedDivorcedYesNo16 – 20 Years21 – 25 Years26 - 30 YearsMore than 31 YearsNo income1 – 50005001 – 1000010001 – 15000More than 9619032792447018816384.2 Data AnalysisIn this paper, SmartPLS 3.2.8 to examining the theoreticalframework because it is developing second-generationtechnique [53]. According to Hair, Hair Jr, Sarstedt [54],bootstrapping is a method to develop path coefficients andfactor loadings, and to get significant standards must to runbootstrapping 5000 subsamples. SmartPLS is consideredbest due to some benefits over other tools, like no normalityissue and multicollinearity test and can use for simple andcomplex models. Moreover, the literature reveals thatSmartPLS is nest to calculating results and establishvalidities of variables ad compared to covariance-basedstructural equation modeling (CB-SEM) [53]. There are twomethods to examine the research model such asmeasurement model and structural model4.2.1 Measurement modelThe main purpose of the measurement model is to regulatehow well all the measure/ items of the constructs ladentheoretically and associated with particular constructs [55].To examine the generated hypothesis, the investigators usePartial least squares structural equation modeling (PLSSEM). In PLS-SEM, usually two approaches used the firstISSN 9measurement model and the second is the structural model.In the measurement model (outer model) factor analysisexecuted to establish model fitness. There are two methodsto assess the soundness of the theoretical models, such asreliability and validity [53]. To computing the reliability needsto establish first internal consistency reliability and it can beassessed by composite reliability (CR) of every construct,furthermore values of CR should be more than 0.6 of eachconstruct, and then evaluate the individual internal reliabilityby computing outer loading of individual items of constructsand remove all those items whose factor loading is below0.40 to improve the value of AVE and CR. Second establishconstruct validity in analyzing convergent validity anddiscriminant validity [56]. Meanwhile, convergent validitystates that measures positively connect with other measuresof the same variable [53]. Furthermore, it can be computedby AVE that value should be 0.50 at least. On the handdiscriminant validity based on empirical values to whichspecific construct of the model is different from otherconstruct of the model [53]. Meanwhile, discriminant validitymeasured by ensuring that squared AVE should be greaterthan other constructs as shown intable2630IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-8616Measurement Model fig.2631IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-8616Table 2Convergent ValidityVariablesConvenience RiskFinancial RiskOnline Shopping BehaviorPrivacy RiskProduct RiskSubjective 1SBN3SBN4SBN6TR1TR2TR3TR4TR5Factor 410.8590.8800.6710.735Table 2 shows that CR value should be greater than 0.60and it is in the present study, and AVE higher than 0.50 asrecommended [53]. Rho A values confirmed each item ofconstructs reliable.4.2.2Discriminant validityDiscriminant validity denotes a position where eachconstruct of model different from other constructs.Moreover, this 7860.5170.8810.5830.8470.6420.899R20.422certifies that items of the particular constructs are differentfrom other construct items and only relate with theirperspective [53]. Furthermore, the diagonal coefficient needsto greater than other all value in the same rows and columnsthat are shown in Table 3.Table 3.4.2.3 Discriminant validityVariablesConvenience riskCR0.793FROSBPPRFinancial risk-0.0190.813Online shopping behaviorPrivacy risk-0.1650.132Product -0.0600.0360.719Subjective -0.0140.598-0.1540.0480.170TR0.801Table 3 reveals that we meet the standards for discriminant validity as the value of a particular construct should be different fromothers suggested by [57].632IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-8616Table 4Cross LoadingsVariablesConvenience 230.425Product acy sTrustTable 4 determines the cross-loadings and meets the criteriaas suggested by [53].4.2.4 Coefficient of determination R2Table 5.Latent variablesR- SquareResultsExogenous variable OSB ( without moderator)0.102WeakExogenous variable OSB ( with moderator)0.442ModerateThe coefficient of determination R2 is used to examine theaccuracy of the model that is calculated as the squaredassociation of the analytical values and certain dependentconstruct [58]. R2 value shows all exogenous constructs howmuch effect endogenous construct collectively [58]. R2 atleast should 10%, in this study without moderator allexogenous constructs effect OSB 0.102 that is weak R2 andwith moderator 0.442 that is moderateexogenous construct and run, again excludes anothervariable and run similarly run until the last exogenousconstruct excluded. Effect size calculated by the PLSalgorithm technique4.11 Effect size F2the effect size of the variable measured by excluding one633IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-8616Table 6 Effect size F2Exogenous variableTotal effectConnivance riskFinancial risk0.0080.009Privacy riskProduct risk0.0430.012Subjective normsTrust0.0260.489Table 6 shows that some exogenous constructs have a weakeffect on endogenous contrast and some have a strong effecton endogenous contrast.4.3 Structural ModelIn this segment, we argued the direct hypotheses betweendependent and independent constructs. According to Hair Jr,Hult [53] by performing 5000 subsamples bootstrap toexamine the significant values of loading and path coefficient.4.3.1 Predictive relevance model Q2Researcher subtracts R2 and cross-validated redundancymodel to compute predictive relevance of model Q2. R2 valueexamines the level of variance that exogenous constructsexplain endogenous construct. In this study, 10.2% of onlineshopping behavior explained by financial risk, privacy risk,product risk, and convenience risk. Meanwhile, 44.2%explained by all exogenous constructs with moderators.Table 74.3.2 Predictive relevance model Q22ConstructsROSB ( without moderator)OSB ( with moderator)0.1020.442cross-validated redundancy value calculates to know thequality of the model, and it is computed by blindfoldingtechnique (PLS-SEM). According to Fornell and Cha [59]value of cross-validated redundancy greater than zero (0).4.3.3 Model FitResearchers should be wary to use model fit in PLS-SEM[53]. The standardized root means square residual (SRMR)is basedCross validated redundancy Q20.0470.195on the predicted and covariance matrix transforming of bothinto correlation matrices. SRMR value should be below 0.08or0.10 [60]. Normed fit index (NFI) calculates the Chi2 value ofthe suggested model and matches this value with standard[53].Model Fit Table 8Structured ModelEstimated 6250.625634IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-8616Structural Model direct relationship Fig 34.3.4 Hypotheses testing Direct Table 9HypothesesHypotheses PathsH1CR -- OSBH2H3FR -- OSBPRR -- OSBH4PDR -- OSBH5SBN- OSBH6TR -- OSBBeta valueSampleMeanStd. 000Significant0.5470.5410.04013.5920.000There are six direct hypotheses and all hypotheses areaccepted. Table 9 validate that convenience risk (CR)important predictor of online shopping behavior (β -0.068,t 2.158, p 0.016) and hypothesis H1 supported. Similarly,financial risk affects online shopping behavior (β -0.071, t 1.942, p 0.026) H2 is also supported. Privacy riskhas a significant negative effect on online shopping behaviorH3 supported (β -0.160, t 4.220, p 0.00). Furthermore, H4Significantalso supported that product risk affects online shoppingbehavior (β -0.084, t 1.871, p 0.031). In addition, subjectivenorms have a positive and significant effect on onlineshopping behavior and H5 supported (β 0.125, t 3.700,p 0.000). H6 also supported in the sense that trust also has apositive significant effect on online shopping behavior(β 0.547, t 13.592, p 0.000).635IJSTR 2020www.ijstr.org

INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME 9, ISSUE 01, JANUARY 2020ISSN 2277-86164.3.5 Testing Moderation EffectsStructural Model (Indirect Relationship) Figure 44.3.6 Hypothesis Testing for Moderating Relationships Table 10HypothesesHypotheses PathsBeta valueSample MeanStd. Dev.T-valuesP-valuesResultsH7CR*SBN -- OSB-0.102-0.0950.0462.2300.013SignificantH8CR*TR -- OSB0.1020.1080.0492.1010.018SignificantH9FR*SBN -- OSB-0.068-0.0610.0381.8140.035SignificantH10FR*TR -- OSB0.0050.0150.0450.1000.460Not SigH11PDR*SBN -- OSB-0.065-0.0480.0351.8580.032SignificantH12PDR*TR -- OSB0.0380.0320.0560.6750.250Not SigH13PPR*SBN -- OSB-0.067-0.0630.0371.7870.037SignificantH14PPR*TR -- OSB-0.139-0.1280.050

communication has changed the business performance and producer’s interactions with consumers. Online shopping behavior is critical in today e- . 2.1 Online shopping behavior Online shopping refers to electronic commerce to purchase . Meanwhile, it’