INTRODUCTION
BACKGROUND
The Internet has revolutionized social interactions and shopping habits over the past few decades, enabling virtual communication and online purchases without the need to visit physical stores E-commerce, particularly online shopping, has provided businesses with effective distribution channels beyond traditional methods, allowing customers to save valuable time Consequently, the online shopping sector has experienced exponential growth, evidenced by a significant rise in e-business revenues and the volume of transactions conducted via the Internet.
The Vietnamese government actively supports the growth of e-commerce, with the Ministry of Industry and Trade housing the E-commerce Development Centre This center monitors and analyzes the progress of e-commerce in the country, devising various strategies to enhance its development Additionally, they publish annual reports detailing the state of e-commerce in Vietnam.
Between 2006 and 2009, the number of enterprises in Vietnam adopting e-business rose from 8% to 12%, indicating a shift towards online shopping where consumers can purchase products via the Internet without visiting physical stores Despite this growth, traditional buying habits remain prevalent, characterized by strong buyer-seller relationships that allow for direct communication and product introductions in convenience stores These established practices, along with barriers such as inadequate electronic payment systems and network infrastructure, hinder the further development of online shopping in Vietnam.
PURPOSES AND RESEARCH QUESTIONS
E-commerce in Vietnam has government and commercial enterprise attention but the number of customers shopping online is still limited People are not yet ready to trade online Moreover, with the limited number of current customers, what should companies do to keep them and develop close relationships with them? How can companies ensure customer product repurchase or recommend others people to use them?
Research in Vietnam primarily focuses on e-commerce development from the perspective of companies, with limited attention given to customer attraction and retention in online shopping This study aims to investigate the connections between trust, commitment, loyalty, and retention—key elements identified in prior customer relationship marketing studies By examining these relationships, the research will identify factors that influence customer loyalty and retention, offering recommendations to enhance online shopping experiences in Vietnam Consequently, two research questions have been formulated to guide this inquiry.
1 What are the antecedents of trust and commitment?
2 How do trust and commitment influence customer loyalty and customer retention in online shopping in Vietnam?
METHODOLOGY
The research aims to test the hypothesis regarding the relationship between variables in online shopping in Vietnam, utilizing a deductive approach with an explanatory purpose.
The research will begin with a literature review to examine existing theories on customer relationships in marketing, aiming to identify frequently studied areas and formulate hypotheses Following the recommended research methods, a survey strategy utilizing a questionnaire will be employed to address the research questions effectively.
A data collection questionnaire was developed, comprising 28 questions adapted from various previous studies to assess both independent and dependent variables To ensure clarity and comprehension, a pilot test was conducted for participants and researchers alike.
The data collected will be analyzed using the SPSS program, starting with a check for normal distribution Subsequently, correlations and multiple regressions will be performed to assess whether the hypotheses of this research are supported or rejected, reflecting the relationships among four main areas in comparison to current studies.
DEFINITIONS
This research aims to explore the interconnections between customer trust, commitment, loyalty, and retention, which are essential elements in customer relationship management theories The study will specifically focus on the e-commerce sector, particularly online shopping, to analyze these relationships Key concepts relevant to the research will be clearly defined to ensure a comprehensive understanding of the findings.
Customer Relationship Management (CRM) is a strategic approach aimed at enhancing shareholder value by fostering meaningful relationships with key customers and segments It combines relationship marketing strategies with information technology to build profitable, long-term connections CRM leverages data to better understand customers and collaboratively create value, necessitating a cross-functional integration of processes, personnel, operations, and marketing capabilities, all supported by advanced information technology and applications.
Turban and King (2003) defined E-commerce (EC) as “the process of buying, selling, or exchanging products, services, and information via computer networks, including the Internet”
According to Mosuwe et al (2004), online shopping is characterized by customers' intentions to purchase products via the Internet from businesses engaged in e-commerce This process falls under the business-to-consumer (B2C) model, which is a key component of the broader e-commerce landscape.
Morgan and Hunt (1994) defined Trust as “the perception of confidence in the exchange partner‟s reliability and integrity”
Morgan and Hunt (1994) describe commitment as a lasting desire to uphold a valued relationship, emphasizing the customer's dedication to the company and the importance of sustaining that relationship.
Customer loyalty is defined as a strong commitment to consistently repurchase a preferred product or service in the future, leading to repeated buying of the same brand, regardless of external influences or marketing strategies that might encourage switching.
Gerpott (2001) defined Customer retention as “maintaining the business relationship established between a supplier and a customer”
LITERATURE REVIEW
CUSTOMER RELATIONSHIP MANAGEMENT
Customer Relationship Management (CRM) is a strategic approach aimed at enhancing shareholder value by fostering strong relationships with key customers and segments It combines relationship marketing strategies with information technology to establish profitable, long-term connections with customers and stakeholders By leveraging data and information, CRM facilitates a deeper understanding of customer needs and enables value co-creation Successful CRM implementation requires cross-functional integration of processes, personnel, operations, and marketing capabilities, all supported by advanced technology and applications This comprehensive definition highlights the essential activities of CRM and its integration within organizations.
Since its emergence in the mid-1990s, Customer Relationship Management (CRM) has evolved through three distinct generations According to Kumar and Reinartz (2006), the first generation focused on functional CRM aimed at boosting sales and enhancing services, emphasizing activities like sales force automation and customer support The subsequent generation introduced a customer-facing front-end approach, which aimed to address gaps in enterprise resource planning (ERP) functionality and align with business needs However, during the 1990s, CRM's goal of facilitating customer interactions—from pre-sales to post-sales—via communication channels like telephone and the internet remained largely unfulfilled.
By the end of 2002, the company adopted a strategic approach to third-generation CRM, learning from previous unsuccessful implementations This evolution shifted focus from solely customer-facing front-end systems to include back-end systems involving partners and suppliers By integrating these components with Internet technology, CRM transformed into a comprehensive company strategy rather than just a technological solution, significantly contributing to revenue growth.
Trust plays a crucial role in the success of e-commerce, particularly due to the uncertainties inherent in online transactions Researchers emphasize that trust significantly influences social and economic interactions that involve uncertainty and dependency Two key factors affected by trust in online transactions are security and privacy, highlighting the importance of clearly defining the concept of trust in this context.
E-commerce offers convenience by connecting buyers and sellers, but it also has limitations, including the absence of direct communication between them and with the products To overcome these challenges, suppliers must focus on building trustworthy relationships to enhance customer loyalty.
Teo & Liu (2007) emphasize that consumer trust is crucial in e-commerce, highlighting the need to understand its antecedents and consequences Identifying the factors that influence trust is vital for developing effective strategies to enhance it Additionally, exploring the outcomes of trust can provide insights into its significance and impact on online purchasing behavior.
In the realm of organizational trust, Mayer et al (1995) introduced a model illustrating the relationship between a trusting party and the party to be trusted In the context of e-commerce, Jarvenpaa et al (2000) investigated how customers' perceptions of an online store's reputation and size influence their trust in that store Their findings indicate that trust significantly impacts consumers' attitudes, intentions, and behaviors.
Trust is a crucial element in successful marketing relationships, defined by Morgan and Hunt (1994) as "the perception of confidence in the exchange partner's reliability and integrity." Mayer et al (1995) further elaborate that trust involves a party's willingness to be vulnerable to another's actions, based on the expectation of reliable performance, regardless of the ability to monitor or control that party Overall, confidence and reliability emerge as fundamental components of trust in these definitions.
The concept of trust is complex and encompasses various dimensions that are not fully captured by its original definition To effectively study trust, it is essential to classify it based on different beliefs derived from analyzing and comparing its definitions These classifications can vary according to factors such as attitudes, beliefs, behaviors, and tendencies, as well as the specific referents in trust—whether it pertains to something, someone, or particular characteristics like honesty.
In summary, this study adopts the definition of trust provided by Morgan and Hunt (1994), which describes trust as the perception of confidence in the reliability and integrity of an exchange partner.
This research addresses the antecedents of trust, identifying two key factors: e-retailer reputation and privacy concern Previous studies, including those by Bennett & Gabriel (2001) and Josang et al (2007), highlight the significance of e-retailer reputation in building trust, while Eastlick (2006) emphasizes the role of privacy concerns The study will further explore the relationship between these antecedents and their impact on trust.
The decision to engage in electronic commerce is influenced by various factors, with the retailer's reputation being a critical element According to Bennett & Gabriel (2001), e-retailer reputation is synonymous with brand reputation, encompassing the name, term, symbol, or design that distinguishes one retailer's goods and services from others This reputation extends beyond mere image attributes, incorporating external perceptions and subjective assessments of an organization's qualities based on its historical performance.
Research by Van and Leunis (1999) indicates that brand reputation significantly alleviates customers' risk concerns when engaging in online transactions E-retailer reputation plays a crucial role in influencing customer decisions to participate in e-commerce Additionally, studies by Bennett and Gabriel (2000) and Josang et al (2007) demonstrate a positive relationship between reputation and trust, where a strong reputation enhances customer trust This theoretical evidence is vital for constructing and testing hypotheses in the field.
A company's reputation is a fragile yet strategic asset that can be easily damaged, making it far more challenging to build than to lose Therefore, suppliers with a strong reputation must prioritize the protection of their image to mitigate potential negative impacts.
In traditional marketing, reputation is a crucial factor influencing the trust between buyers and sellers Research by Teo & Liu (2007) highlights that in online shopping, the perceived reputation of a vendor significantly impacts consumer trust This finding provides theoretical support for establishing and examining the connection between e-retailer reputation and customer trust, as outlined in hypothesis H1a.
METHOD OF STATISCAL ANALYSIS
The research data analysis begins with a correlation analysis, which examines the relationships between study variables This analysis calculates the correlation based on the standardization of covariance, specifically utilizing Pearson's correlation coefficient (r) to quantify these relationships.
The correlation coefficient has to lie between -1 and +1
A coefficient of +1 indicates a perfectly positive relationship; a coefficient of -1 indicates a perfectly negative relationship, while a coefficient of 0 indicates no linear relationship at all
The correlation coefficient is a commonly used measure of the size of an effect: values of ±0.1 present a small effect, ±0.3 a medium effect and ±0.5 a large effect
Multiple regression is a statistical method that examines the relationship between a dependent variable and multiple independent variables In this study, all regression analyses involved a dependent variable alongside several independent variables, necessitating the use of multiple regression techniques The key procedures for analyzing multiple regression are outlined in the subsequent sections.
In social science, three primary regression methods are utilized: standard (forced entry) regression, sequential (hierarchical) regression, and statistical (stepwise) regression Standard regression involves the simultaneous entry of all predictors into the model, whereas hierarchical regression allows researchers to determine the order of predictor entry Stepwise regression, on the other hand, relies on mathematical criteria to decide the order of predictors This study aims to examine the mediating relationships among variables, employing hierarchical regression to enter predictors individually under the researcher’s control, rather than simultaneously, to effectively test the hypotheses.
Analysis of regression result in this study was based on major statistics such as sums of squares (R 2 , adjusted R 2 , and R 2 Change) and regression coefficients (Bi & òi)
The sum of squares (R²) quantifies the extent to which independent variables explain the variability of the dependent variable, reflecting the percentage of outcome variation accounted for by the model It is derived by dividing the residual sum of squares (SSR) by the model sum of squares (SSM) To determine the significance of R², the p-value of the F-ratio is evaluated, with R² being considered significant if the p-value is less than 0.05 at an alpha level of 0.05.
Adjusted R² measures the proportion of variability in the dependent variable explained by independent variables when the model is based on the entire population instead of a sample This statistic is typically expected to be close to R².
R² Change is utilized to assess the difference in R² values before and after incorporating one or more independent variables into a regression equation This change is considered significant when the p-value associated with the R² Change ratio is below 0.05 at an alpha level of 0.05.
Assessing regression diagnostics is essential to determine if a model accurately fits the observed data or if it is affected by a few influential cases This evaluation involves identifying outliers and cases that significantly impact the regression results.
An outlier is a significant deviation from the main trend in data, which can skew the estimated regression coefficients and lead to model bias To minimize this bias, it is crucial to identify outliers by examining the large differences between observed data values and those predicted by the model, known as residuals Residuals represent the error in the model; smaller residuals indicate a good fit, while larger residuals suggest a poor fit Standardized residuals are commonly used to detect these outliers effectively.
According to Field (2005), one general rule for residuals:
(1) “ standard residuals with an absolute value greater than 3.29 are cause for concern because in an average sample a value high like this is unlikely to happen by chance”;
(2) “ if more than 1% of a sample has standardized residuals with an absolute value greater than 2.58 there is evidence that the level of error within our model is unacceptable”; and
(3) “ if more than 5% of cases have standardized residuals with an absolute value greater than 1.96 then there is also evidence that the model is a poor representation of the actual data”
Residuals help identify outliers by analyzing model errors, while influential cases assess whether specific data points disproportionately affect model parameters Key statistical measures for identifying influential cases include Cook's distance, leverage, Mahalanobis distance, DFBeta, and the covariance ratio (CVR).
Cook’s distance is a statistic that considers the effect of a single case on the model as a whole The values of Cook’s distance greater than 1 is may be cause for concern
Mahalanobis distance is connected to leverage values and quantifies how far cases deviate from the mean of the predictor variables In large samples, specifically with N=500 and five predictors, values exceeding 25 indicate potential issues that warrant attention.
In smaller samplers (N = 100) and with fewer predictors (namely three) values greater than 15 are problematic, and in very small sample (N0) with two predictors values greater than 11 should be examined
DFBeta measures the impact of excluding a single case on parameter estimates in a regression model By calculating DEBeta for each case and parameter, researchers can pinpoint cases that significantly influence the model's parameters Specifically, a standardized DFBeta value exceeding 2 in absolute terms signifies a considerable effect on the regression coefficients.
Regression analysis accurately reflects the sample of observed values, but its validity may not extend to a broader population To generalize the model effectively, it is essential to meet key assumptions, including independent residuals, normality of residuals, homoscedasticity, linearity, and the absence of multicollinearity.
(1) Independent residuals (Durbin-Watson test)
The first assumption for generalizing a model is that residual terms from any two observations should be uncorrelated This can be evaluated using the Durbin-Watson test, which assesses serial correlations among errors The test statistic ranges from 0 to 4, with a value of 2 indicating uncorrelated residuals Values above 2 suggest a negative correlation between adjacent residuals, while values below 2 indicate a positive correlation It's important to note that the Durbin-Watson statistic is influenced by the number of predictors in the model and the total number of observations.
Values below 1 or above 3 are significant red flags, while values around 2 may also indicate potential issues, depending on the sample and model used.
The model assumes that the residuals are random, normally distributed variables with a mean of zero, indicating that the differences between the model and observed data are typically small, with larger discrepancies occurring infrequently This assumption can be validated using two graphical methods: a histogram and a normal probability plot The histogram of the residuals should resemble a normal curve with matching mean and standard deviation Additionally, if the residuals are normally distributed, the points on the normal probability plot should align closely with the normal distribution line.
METHODOLOGY
RESEARCH STRATEGY
The deductive and inductive approaches are fundamental methodologies in research, each with distinct characteristics that guide researchers in selecting the appropriate method to address their research questions Understanding the differences between these two approaches is crucial for effective study design and analysis.
Research employing a deductive approach is characterized by its focus on explaining causal relationships between variables, necessitating quantitative measurement of all concepts involved It requires generalization through the selection of sufficiently large samples In contrast, an inductive approach involves collecting data first, followed by theory development based on data analysis, emphasizing the understanding of the meanings humans assign to events This approach relies on qualitative data and is less focused on generalization.
This study employs a quantitative research approach, emphasizing numerical observations to generalize phenomena through formal data analysis, with statistical indicators as a key component The methodology involves a survey conducted via a questionnaire, specifically designed to assess whether the collected data effectively addresses the research questions.
This study employs a quantitative cross-sectional survey methodology, which examines a specific phenomenon at a single point in time Longitudinal or experimental approaches were not utilized due to their high demands on time and resources Given the large total population, a sample was selected to test the theoretical model effectively Data collection occurred once over a brief timeframe, encompassing various contexts within the population.
Survey methodology encompasses four primary methods: self-administered questionnaires, interviews, structured record reviews, and structured observations This study aims to collect data on individual attitudes and orientations at a specific point in time while operating under limited resources Utilizing questionnaires is advantageous for identifying and describing respondents' attitudes across various phenomena In contrast, interviews demand considerable time and resources, particularly with larger samples Additionally, structured record reviews and observations focus on visual and recorded data, making them unsuitable for capturing attitudes Consequently, the self-administered questionnaire emerges as the most prevalent and effective method for data collection in social science research, fulfilling all survey requirements and proving to be the best choice for this study.
DATA COLLECTION METHOD
This research investigates the impact of trust and commitment on customer loyalty and retention in Vietnam's online shopping market Participants in the study are required to have prior experience with online shopping In Vietnam, online shopping is characterized by straightforward processes, including purchasing goods via websites and options for online payment or cash on delivery This unique aspect of e-commerce development in Vietnam stems from consumers' limited cash availability and a foundational lack of information technology infrastructure.
In Vietnam's vast population, a representative sample is essential for estimating online shopping characteristics, as there is no definitive number of online customers Analyzing data from this sample allows researchers to extrapolate findings to the entire population, effectively addressing research questions Moreover, utilizing a sample expedites the research process, which is crucial given strict deadlines.
The research uses probability samples which is most popular in the survey- based research strategy Because the research questions is concerned with online customer
This research employs multi-regression analysis to test hypotheses, following Green's (1991) rule of 50+8k to determine the minimum sample size With nine predictors—e-retailer reputation, privacy concerns, alternative attractiveness, switching cost, customer satisfaction, trust, commitment, customer loyalty, and customer retention—the required sample size is calculated as 50 + 9x8, resulting in 122 respondents Anticipating a response rate of 10%, a total of 1,220 questionnaires will be distributed to achieve the desired sample size.
The key factors in determining sample size are population representativeness and reliability In this study, a sample of 1,000 online shoppers in Vietnam, primarily university students from the National Economics University in Hanoi and working professionals, was targeted, with an anticipated response rate of 10% to 15% To ensure the sample's representativeness, a multi-regression analysis will be conducted to explore the relationships among the studied variables Additionally, three statistical methods—Cook's distance, Mahalanobis distances, and DFBeta—will be employed to identify any potential biases in the sample The findings will ultimately help assess whether the sample can accurately generalize to the entire population.
To accommodate time and resource constraints, a self-administered questionnaire was selected for a sample size of 1,000 participants To achieve the anticipated response rate of 10% to 15%, the research employed a delivery and collection method, where questionnaires were distributed to participants and collected upon completion.
The content of the questions focuses on customers‟ assessments expressed by
The survey assessed participants' levels of agreement, ranging from "extremely disagree" to "extremely agree," with statements about their last shopping experience and online shopping habits Questions 1 to 14 focused on determining factors of trust and commitment, while questions 15 to 28 explored customer loyalty and retention, aiming to uncover the relationships among trust, commitment, customer loyalty, and customer retention.
The questionnaire aimed to collect data from Vietnamese online shoppers to identify factors that influence their purchasing decisions and to evaluate the relationships among nine key variables: e-retailer reputation, privacy concerns, alternative attractiveness, switching costs, customer satisfaction, trust, commitment, loyalty, and retention Each variable was assessed through multiple questions derived from previous research to ensure clarity and comprehensibility for the participants.
The research utilizes a 28-question questionnaire divided into two sections to explore factors influencing customer loyalty and retention in online shopping within the Vietnamese market Section one comprises 14 questions focused on customers' recent online shopping experiences, examining the impact of trust and commitment as mediating variables Section two, also containing 14 questions, gathers insights on customer perceptions of e-retailers concerning trust, commitment, loyalty, and retention The study aims to identify relationships among these variables and assess the primary factors that affect customer loyalty and retention, supported by specific hypotheses outlined in the accompanying tables.
H1a E-retailer reputation positively affects customer trust
Questions 1-3 measure e-retailer reputation, questions 15-19 measure customer trust
H1b Privacy concerns negatively affect customer trust
Questions 4-6 measure privacy concern, questions 15-19 measure customer trust
H2a Alternative attractiveness negatively affects customer commitment
Questions 7-9 measure alternative attractiveness, questions 20-22 measure customer commitment
H2b Switching cost positively affects customer commitment
Questions 10-12 measure switching cost, Questions 20-22 measure customer commitment
H2c Customer satisfaction positively affects customer commitment
Questions 13,14 measure customer satisfaction, Questions 20-22 measure customer commitment
H3a Increasing customer trust leads to higher customer loyalty
Questions 15-19 measure customer trust, Questions 23-25 measure customer loyalty
H3b Increasing customer trust leads to higher customer retention
Questions 15-19 measure customer trust, Questions 26-28 measure customer retention
H4a Increasing customer commitment leads to higher customer loyalty
Questions 20-22 measure customer commitment, Questions 23-25 measure customer loyalty
H4b Increasing customer commitment leads to higher customer retention
Questions 20-22 measure customer commitment, Questions 26-28 measures customer retention
The questionnaire uses a Likert-scale of seven levels from Extremely Disagree (1) to Extremely Agree (7) to measure different variables in the model of the research and their relationships
This study investigates the connections between trust, commitment, loyalty, and retention in the context of e-commerce, with a specific focus on online shopping behaviors.
Probability sampling techniques, such as simple random, systematic, and stratified random sampling, are essential for selecting participants in surveys Random sampling ensures that every individual in the population has an equal chance of being chosen; however, accessing information about online shoppers can be challenging, making it difficult to identify the entire population As a result, convenience sampling based on participant availability becomes necessary In this study, questionnaires were distributed to friends aged over eighteen with online shopping experience in Vietnam, who were then asked to share the survey with others To confirm their eligibility, participants were initially asked, "Have you ever shopped online?" before receiving the questionnaire.
The questionnaire underwent expert review to address ethical considerations and ensure clarity for participants Although the questions were adapted from previous research, a pilot test was conducted to confirm their suitability for the target population This pilot was further evaluated by specialists in online shopping in Vietnam.
DATA ANALYSIS
Variables measured on a Likert scale are considered discrete, as they provide specific values within a defined range, such as a 5-point or 7-point scale This limitation can be a drawback, as it restricts the actual values that these variables can take, often confining them to integers like 1, 2, and so forth.
3, 4, 5, 6, or 7 and cannot be 3.15 or 5.18
As previously stated in the Literature Review chapter, and in relations to the questions in the questionnaire, the independent variables are determined as following table 3.1
(*): all questions displayed in the questionnaire are attached in the Appendix 3.3.3 Dependent Variables
The independent variables are listed and explained in the table 3.2
(*): all questions displayed in the questionnaires are attached in the Appendix
In this research, data from 219 printed paper responses were input into SPSS for primary data storage Descriptive statistics, including means, minimums, and maximums, were calculated to assess the normal distribution of the data The analysis aimed to explore the relationships among variables to address the initial research questions regarding the impact of trust and commitment on customer loyalty and retention Various results were calculated to support these findings.
Based on the mentioned hypotheses, a multivariate regression model is then built, as stated in following equation:
5 LOYALTY = ò 1 ER + ò 2 PC + ò 3 AA + ò 4 SC + ò 5 CS
6 LOYALTY = ò 1 ER + ò 2 PC + ò 3 AA + ò 4 SC + ò 5 CS + ò 6 TR + ò 7 CO
7 RETENTION = ò 1 ER + ò 2 PC + ò 3 AA + ò 4 SC + ò 5 CS
8 RETENTION = ò 1 ER + ò 2 PC + ò 3 AA + ò 4 SC + ò 5 CS + ò 6 TR + ò 7 CO
Where ER: E-retailer reputation; PC: Privacy concerns; AA: Alternative attractiveness; SC: Switching costs; CS: Customer satisfaction; TR: Trust;
CO: Commitment; CL: Customer loyalty and CR: Customer retention.
FINDINGS AND CONCLUSION
DESCRIPTIVE STATISTICS
The initial phase of data analysis involves evaluating the distribution of collected data using descriptive statistics, summarized in Table 4.1 The variables exhibit scores ranging from 1 to 7, indicating no limitations on variability Their means average around 4.0, with a maximum of 4.4581 and a minimum of 3.1689 The standard deviation averages approximately 1.2, with a maximum of 1.51073 and a minimum of 0.94741 Additionally, all absolute values of Skewness and Kurtosis are below the thresholds of 3 and 5, confirming that the variables follow a normal distribution.
Table 4.1 Descriptive Statistics of Scales
Scales N Min Max Mean Std Skewness Kurtosis
CORRELATIONS
Table 4.2 illustrates the correlation matrix among the variables, indicating that all relationships fall within an acceptable range of zero to a threshold of 0.8, warranting further study Notably, a positive correlation exists between e-retailer reputation and customer trust, suggesting that increased reputation enhances consumer trust in e-retailers Additionally, as customer trust in e-retailers grows, customer loyalty also increases.
ER PC AA SC CS TR CO CL CR
* Correlation is significant at the 0.05 level (2-tailed)
** Correlation is significant at the 0.01 level (2-tailed).
HYPOTHESIS TESTING
This section analyzes the interconnections among trust, commitment, customer loyalty, and customer retention, utilizing multi-regression analysis in SPSS Initially, it investigates the relationship between trust and its determinants Subsequently, the correlation between commitment and its related variables is explored The later sections highlight both the direct and indirect influences of various factors on customer retention and loyalty.
This section analyzes how e-retailer reputation and privacy concerns influence customer trust, with Hypothesis 1a suggesting that a strong e-retailer reputation enhances customer trust, while Hypothesis 1b posits that privacy concerns diminish it The findings are detailed in Table 4.3, which displays the regression results illustrating these relationships.
Table 4.3 The Determinants of Trust
95% Confidence Interval for B Collinearity Statistics
Durbin-Watson 1.807 a Dependent Variable: Trust (TR)
The study confirms that e-retailer reputation (ER) has a positive impact on customer trust (TR), supporting Hypothesis H1a Additionally, Hypothesis H1b is validated, indicating that privacy concerns (PC) negatively influence customer trust, as there is a significant correlation between PC and TR Overall, both e-retailer reputation and privacy concerns significantly affect customer trust.
TR at p= 000 & and 000 with coefficient B = 156 & (-.152) respectively
After conducting a multi-regression analysis, it is essential to perform regression diagnostics using Cook's distance, Mahalanobis distances, and DFBetas to identify any influential cases that may skew the results The general guideline for these diagnostics indicates that a standard residual value exceeding 3.29 raises concerns Additionally, if over 1% of the sample has standard residuals greater than 2.58, the model's error level is deemed unacceptable Furthermore, if more than 5% of the sample shows standard residuals above 1.96, it suggests that the model inadequately represents the entire population.
Table 4.4 indicates that there are 11 cases (5%) within ±2, two cases (0.91%) within ±2.5, and no cases within ±3, highlighting the necessity for regression diagnostics The analysis includes calculations of Cook's distance, Mahalanobis distances, and DFBetas The Cook's distance values for the seven cases are all significantly below 1, indicating that none exert undue influence on the regression analysis Additionally, all cases exhibit Mahalanobis distance values beneath the established threshold.
15 and the absolute values of all DFBetas are far below the threshold of 2 indicating that no case influences the regression parameters
Another important thing need to consider after running multiple-regression is to ensure the result from the sample can be generalized to the whole population
The Durbin-Watson value in the regression analysis is 1.807, which falls within the acceptable range of 1 to 3, indicating that the assumption of independent residuals is satisfied and there are no concerns regarding this sample.
The results presented in Table 4.3 indicate that all Variance Inflation Factor (VIF) values are significantly below the threshold of 10, and all tolerance statistics exceed the minimum requirement of 0.2 This confirms that multicollinearity does not adversely affect the regression model.
All significant coefficients in Table 4.3 exhibit narrow confidence intervals that do not intersect zero, suggesting that the regression estimates accurately reflect true population values Consequently, it can be concluded that the model is applicable to the entire population.
This section explores the relationship between commitment and its determinants, highlighting three key hypotheses: Hypothesis 2a suggests that alternative attractiveness negatively impacts customer commitment, Hypothesis 2b indicates that switching costs positively influence customer commitment, and Hypothesis 2c asserts that customer satisfaction enhances customer commitment The findings of the regression analysis on these relationships are detailed in Table 4.5.
Table 4.5 The Determinants of Commitment
Upper Bound Tolerance VIF (Constant) 1.098 2.091 038 063 2.134
Durbin-Watson 1.461 a Dependent Variable: Commitment (CO)
The output in Table 4.5 shows that there is a significant relationship between
The analysis reveals that alternative attractiveness (AA) negatively impacts customer commitment (CO), as indicated by a p-value of 020 and a coefficient of B = (-.180), thus supporting hypothesis H2a Additionally, hypotheses H2b and H2c are confirmed, showing that both switching costs (SC) and customer satisfaction (CS) positively influence customer commitment, with significant p-values of 000 for both correlations This suggests that increased alternative attractiveness, switching costs, and customer satisfaction all contribute to higher levels of customer commitment.
Table 4.6 reveals that there are 10 cases (4.56%) within ±2, 4 cases (1.8%) within ±2.5, and no cases within ±3, indicating the necessity for regression diagnostics The Cook's distance values for these cases are all significantly below 1, suggesting that none exert undue influence on the regression analysis Additionally, all cases exhibit Mahalanobis distance values beneath the established threshold.
15 Finally, the absolute values of all DFBetas which are far below the threshold of 2 indicate that there are no case influences the regression parameters
The value of Durbin-Watson in the regression is 1.461 (see Table 4.5)
This value is in the range between 1 and 3, close to 2, so the assumption independent residuals are thus met or no cause for concern in this sample
The analysis reveals that all Variance Inflation Factor (VIF) values are significantly below the threshold of 10, while tolerance statistics exceed the 2 threshold Additionally, all significant coefficients exhibit narrow confidence intervals that do not include zero, suggesting that the regression estimates accurately reflect true population values Therefore, this model can be effectively generalized to the entire population.
4.3.3 The determinants of Customer Loyalty
4.3.3.1 The direct determinants of Customer Loyalty
This section evaluates how customer trust and commitment impact customer loyalty, proposing that increased trust correlates with enhanced loyalty (Hypothesis 3a) and that greater commitment also fosters higher loyalty levels (Hypothesis 4a).
3.7 shows the output of regression predicting the relationships between these variables Both hypotheses H3a and H4a are accepted because the relationships between trust (TR) and customer loyalty (CL) as well as between customer commitment (CO) and customer loyalty (CL) are significant (.002 and 000 < 05 respectively)
Table 4.7 The Direct Determinants of Customer Loyalty
Upper Bound Tolerance VIF (Constant) 1.196 3.797 000 575 1.817
Durbin-Watson 1.999 a Dependent Variable: Customer Loyalty (CL)
Table 4.8 reveals that 12 cases (5.47%) fall within ±2, 6 cases (2.7%) within ±2.5, and more than ±3, necessitating regression diagnostics The Cook's distances for these cases are significantly below 1, indicating they do not exert undue influence on the regression model Additionally, all cases have Mahalanobis distance values under the threshold of 15 Lastly, the absolute DFBetas values remain well below the threshold of 2, confirming that no individual case significantly affects the regression parameters.
