Realizing the important and crucial role of Cloud Computing, with an enthusiasm to explore more to find solutions for the urgent problems of storing accounting data in Cloud, the group o
INTRODUCTION
Reasons of Choosing the research objective
In today's globalized world, innovation and efficiency in the workplace are essential (Aini et al., 2019) Leveraging cloud computing enables IT departments to prioritize the development of creative applications over the maintenance of software and infrastructure Common examples of cloud computing include online storage solutions like Dropbox and Google Drive, social media platforms such as Facebook and Twitter, email services like Gmail and Yahoo, and various online business applications, including Microsoft Office Live, customer relationship management (CRM) systems, and enterprise resource planning (ERP) tools.
Technology is transforming the accounting profession, primarily driven by the rise of cloud accounting Cloud computing enables organizations to access computing resources and applications from anywhere with an internet connection However, as the adoption of cloud-based services grows, so do security risks, posing significant concerns for businesses The escalating scale and complexity of cybersecurity threats are critical issues, particularly as employees connect to the internet and store data across various devices and platforms, making it challenging to monitor and protect an organization's infrastructure.
Cybercriminals are becoming increasingly sophisticated, exploiting advanced web technologies and widening the gap between enterprise security and emerging threats Both physical and network security are crucial, especially as insider threats have surfaced as significant vulnerabilities At the Vietnam Security Summit 2021, cybersecurity experts highlighted Vietnam's high rates of malware infections and cyberattacks, noting a concerning trend in rising cyber violations Data stored on cloud platforms is particularly vulnerable, as any member of an organization can inadvertently assist cybercriminals in data breaches A notable incident occurred in 2020 when Chinese hackers compromised the Vietnam Posts and Telecommunications Group (VNPT) network, exemplifying the ongoing security risks that individuals and organizations face in Vietnam.
The rapid growth of cloud-based applications has prompted both industry and academic research communities to develop innovative solutions for more consumer-friendly, cost-effective, and secure cloud systems Security and privacy remain significant obstacles to cloud adoption, largely due to vulnerabilities in various architectural components and technologies, including internet communication, web services, and virtualization Additionally, the government's "National Digital Transformation Program to 2025, with a vision to 2030" marks the formal initiation of digital transformation aimed at achieving smart governance, a digital economy, and a digital society.
The indiscriminate storage of accountants' and auditors' data raises significant concerns regarding information security, even as the industry adapts to the global technological revolution In this rapidly changing era, understanding the critical role of Cloud Computing is essential Driven by a desire to explore solutions for urgent issues related to accounting data storage in the Cloud, the authors have chosen to focus on "The Behavior of Storing Accounting Data on Public Cloud and Accounting Information Security."
Research concerns
Cloud Computing is significantly transforming the Accounting sector in Vietnam, reflecting the broader global technology revolution In 2018, Vietnam led in Cloud spending at 64.4 percent annually, according to statistical surveys The Asia Cloud Computing Association's report in the same year ranked Vietnam 14th globally, with a coverage score of 41 out of 100 for Cloud services This indicates a growing prevalence of Cloud Computing, which is increasingly replacing traditional storage methods However, despite its rising popularity, the adoption of Cloud Computing still faces numerous challenges.
The primary challenge facing accountants and auditors regarding Cloud Computing is the lack of awareness and the restrictions imposed by Cloud providers As accounting data is stored and managed in the Cloud, users are unable to actively ensure the confidentiality of this sensitive information The Cloud Computing model consists of three essential components: the central computer, server/client, and web application, all of which are susceptible to security vulnerabilities In the event of a Cloud attack, firms face significant risks, including the potential loss and manipulation of critical information.
One of the primary challenges of Cloud Computing is the reliance on a stable Internet connection, as disruptions can lead to delays or corruption in data storage Additionally, concerns surrounding information security and legal compliance make businesses hesitant to adopt Cloud Computing for their accounting activities Companies fear potential data breaches and the risk of losing information from their cloud service providers when storing sensitive data online.
On the other hand, risks from users or from the habit of users can also affect the level of information security of the enterprises.
In response to concerns regarding the use of Cloud Computing in corporate accounting, a group of authors conducted research to analyze user behavior and the potential risks to information security associated with Cloud-based data storage The study gathered insights from 500 accountants and auditors utilizing Cloud services, aiming to identify the factors influencing their behavior and to strengthen the validity of the research objectives.
Research Objectives and Research Questions
This article examines the factors influencing the storage of accounting data in the Public Cloud, focusing on the significance of information security It assesses potential risks that could compromise accounting information security and analyzes the advantages and disadvantages of using Public Cloud for data storage The research also proposes solutions to enhance Cloud capabilities for accounting practices Additionally, it explores the technological and economic factors that shape accountants' and auditors' awareness and attitudes toward Cloud adoption.
Ql:What are the factors influencing the accounting decision to adopt cloud storage for accounting information?
Q2: How do the perceived security and privacy concerns affect the decision to adopt cloud storage for accounting information among accounting professionals ?
Research Object and Research Scope
This paper examines two primary research objectives: first, it investigates the factors influencing accountants' and auditors' intentions to store accounting information in the Cloud; second, it explores the elements that impact their perceptions of information security risks.
Research Scope: Flo Chi Minh City
Survey Object: The group of authors collects insights from accountants and auditors who are working for firms, enterprises in Ho Chi Minh City, Vietnam.
The mixed research methodology utilized in the study was carried out in two stages: formal research using quantitative methods and exploratory research using qualitative methodology.
This study aims to identify the factors influencing the behavior of accounting data storage on cloud computing platforms Utilizing qualitative research methods, the theoretical framework was established based on existing literature The survey targeted accountants and auditors in Ho Chi Minh City Subsequently, a quantitative approach was employed to validate the research model and scale A questionnaire featuring 14 statements was developed, assessed using a 5-point Likert scale Following nearly two months of data collection, 200 valid questionnaires were analyzed.
The group of authors analyzed the collected data using Stata software Scale test (Cronbach's Alpha), exploratory factor analysis (EFA). l.ó.Practical applications of Research:
Despite the abundance of articles and public papers discussing Cloud Computing's definition, advantages, and disadvantages, there is a notable scarcity of scientific research on the subject in Vietnam Recognizing this gap, the authors aimed to explore the factors influencing the recognition and acceptance of Cloud Computing among firms, particularly in Ho Chi Minh City Their goal is to propose actionable solutions that enable these businesses to make informed decisions and maximize their benefits from Cloud technology.
Cloud computing enhances information security in data storage, addressing the critical need for confidentiality in accounting Given the vulnerability of accounting information to cyberattacks, natural disasters, and system failures, utilizing cloud storage provides robust protection against both software and hardware risks With regular synchronization and multiple layers of security, cloud solutions effectively safeguard sensitive data from potential threats.
Cloud Computing revolutionizes data access for accountants by eliminating the limitations of traditional computer systems, which require users to be physically present at their office With cloud-based accounting applications, users can conveniently access their financial documents, reports, bills, and invoices from anywhere with an Internet connection, providing 24/7 accessibility This flexibility enhances productivity and efficiency, allowing accountants to work remotely without the constraints of location.
This research addresses the lack of scientific studies on Cloud Computing in Vietnam by identifying key factors that affect its acceptance among firms in Ho Chi Minh City It also offers solutions to enhance decision-making and maximize the benefits of Cloud Computing By improving information security and facilitating easy data accessibility, Cloud Computing significantly boosts efficiency for accounting professionals.
1.7 Research structure: rhe study consists of 5 chapters as follow:
Chapter 2: Theoretical basis and research model: Presenting the theoretical basis, and building research models and hypotheses.
Chapter 3: Research Methods: Presenting research methods in building and checking the scale.
Chapter 4: Research results: Analyze research results to conclude research hypotheses.
Chapter 5: Conclusion and recommendations: Summary of research content, significance of the study, limitations of the study and orientation for future research.
In the era of globalization, innovation is essential for enhancing effectiveness and efficiency across all work sectors, with cloud accounting emerging as a key driver of transformation for accountants However, the rise of cloud-based applications has heightened security concerns regarding the storage of accounting data As organizations adapt to the global technological landscape, the need for robust information security practices becomes increasingly critical Recognizing the vital role of cloud computing, the authors focus on addressing the challenges of storing accounting data in public cloud environments while prioritizing solutions for safeguarding accounting information security.
Research structure
rhe study consists of 5 chapters as follow:
Chapter 2: Theoretical basis and research model: Presenting the theoretical basis, and building research models and hypotheses.
Chapter 3: Research Methods: Presenting research methods in building and checking the scale.
Chapter 4: Research results: Analyze research results to conclude research hypotheses.
Chapter 5: Conclusion and recommendations: Summary of research content, significance of the study, limitations of the study and orientation for future research.
Conclusion of Chapter 1
In the era of globalization, innovation in all work processes is essential for enhancing effectiveness and efficiency, with cloud accounting emerging as a key driver of this transformation for accountants However, the rising popularity of cloud-based applications has brought significant security concerns regarding the storage of accounting data As the global technological landscape evolves, the need for change is evident, particularly in addressing the challenges of information security Recognizing the crucial role of cloud computing, the authors focus on the critical issue of "The behavior of storing accounting data on public Cloud and accounting information security," demonstrating their commitment to finding solutions for the pressing challenges associated with cloud data storage in accounting.
THEORETICAL FOUNDATION AND RESEARCH MODEL
Overview of domestic and foreign studies related to the topic
Researchers have extensively explored various aspects of cloud security, including architecture components, attack vectors, challenges, threats, vulnerabilities, and proposed solutions Morsy et al highlighted the security challenges related to cloud architecture, service delivery models, and multi-tenancy, emphasizing issues in virtualization and isolation They proposed an integrated and adaptable configuration-based security model However, their study does not provide a comprehensive analysis of security requirements, related threats, vulnerabilities, and their corresponding responses.
Cloud services are structured around four key layers: hardware, infrastructure, platform, and application Security challenges in the cloud encompass automated security provisioning, virtual machine migration, hardware server consolidation, energy management, software frameworks, data security, and storage technologies Takabi et al highlighted unique security and privacy issues arising from data outsourcing, virtualization, heterogeneity, extensibility, shared responsibility, service level agreements (SLAs), and regulatory compliance Their proposed security solution model focuses on user authentication, identity management, access control, trust management, secured service management, and comprehensive policy management, emphasizing the need for unified organizational security management.
Hashizume et al conducted a comprehensive study on cloud-related risks, vulnerabilities, and countermeasures, emphasizing the end-user perspective on service level security They effectively outlined various risks and responses, showcasing their interconnected nature Meanwhile, Xiao and Xiao identified confidentiality, integrity, availability, accountability, and privacy-preservability as the five critical qualities for assessing vulnerabilities and security techniques Additionally, Modi et al presented a layered model of the cloud environment, addressing security challenges and enabling technologies within these layers However, there is a need for more in-depth exploration of privacy challenges and solutions, particularly regarding data storage management strategies and lifecycle management.
Fernandes et al conducted an in-depth analysis of cloud security challenges, utilizing a taxonomy of vulnerabilities, threats, and attacks Their study emphasized the significant impact of security vulnerabilities and offered practical recommendations for addressing these issues Similarly, Ramachandra et al pointed out that the cloud's complex architecture and resource-sharing model introduce unique risks for all stakeholders They stressed that security is a collective responsibility and identified existing security risks and challenges, along with strategies to mitigate them effectively.
All et al identify three primary categories of cloud security challenges: communication security, architectural security, and contractual and legal aspects They point out security concerns that affect both virtual and traditional physical networks, emphasizing the need for methods that ensure virtual machines achieve physical isolation The authors also address the importance of legal compliance due to the geographical distribution of cloud computing platforms Furthermore, they offer a concise overview of security vulnerabilities specific to mobile cloud computing.
The survey by Singh and Chatterjee offers a comprehensive overview of cloud security, detailing the essential characteristics of cloud computing along with its security challenges, threats, and responses It also explores supporting technologies, trust, privacy, and the cloud architectural framework, including various service and deployment models However, the study fails to address the implications of increased cloud usage and the growing security concerns on emerging technologies and applications such as IoT, SDN, NFV, and Big Data.
Each role within the cloud computing environment addresses unique aspects of security and privacy concerns, yet currently, there are no available opportunities This article provides a comprehensive overview of established cloud security criteria, highlighting threats faced by accountants, identified vulnerabilities, and suggested solutions for auditors who store data in the cloud Additionally, it notes the lack of discussion regarding the consequences of cloud usage As cloud-enabled applications evolve, particularly in areas like IoT, 5G, SDN, and NFV, they encounter significant security challenges This review aims to bridge existing gaps by focusing on trust-based security paradigms and proposing strategies to enhance privacy protection.
Theoretical Background
2.2.1 The technology acceptance model (TAM)
The Technology Acceptance Model (TAM), developed by Davis in 1989, effectively predicts and explains the acceptance and usage of information technology (IT) by focusing on behavioral intention and actual system use Recognized for its simplicity and robustness, TAM builds upon the Theory of Reasoned Action (TRA) and the Theory of Planned Behavior (TPB) introduced by Fishbein and Ajzen in 1975 By replacing various attitude measures from TRA, TAM emphasizes two critical factors for technology acceptance: perceived ease of use and perceived usefulness.
Perceived usefulness is the degree to which an individual believes that utilizing a specific technology can improve their performance and productivity in accomplishing tasks or goals Users are more inclined to embrace a technology when they recognize its value and benefits in enhancing their work or daily activities.
Perceived ease of use refers to the extent to which individuals feel that utilizing a technology is simple and uncomplicated When users find a technology easy to learn, navigate, and operate, they are more likely to adopt it.
The Technology Acceptance Model (TAM) has evolved through further developments, notably TAM 2 by Venkatesh and Davis (2000), which introduced key theoretical constructs such as objective norm, voluntariness, image, job relevance, output quality, result demonstrability, and perceived ease of use This model incorporates four personal anchoring factors—computer self-efficacy, perception of external control, anxiety towards computers, and computer playfulness—alongside two experience-based adjustment factors: perceived enjoyment and objective usefulness These elements reflect general beliefs about computers and their usage, significantly influencing the perceived ease of use of new systems, regardless of the system's inherent characteristics.
TAM 3, introduced by Venkatesh and Bala (2008), encompasses all determinants of perceived usefulness and perceived ease of use to examine their impact on behavioral intention and system usage.
The Technology Acceptance Model (TAM) is a straightforward framework that focuses on two key factors: perceived usefulness and perceived ease of use This simplicity makes it accessible for researchers and practitioners alike By emphasizing users' beliefs and attitudes, TAM highlights the importance of understanding user needs in technology adoption decisions, making it an essential tool for effective technology design and implementation Additionally, TAM's versatility allows it to be applied across various technological domains and user groups, facilitating the assessment and prediction of technology acceptance in diverse contexts.
2.2.2 Theory of Elaboration Likelihood Model (ELM)
The Elaboration Likelihood Model (ELM), developed by Richard E Petty and John T Cacioppo in the 1980s, is a social psychology theory that elucidates how individuals process persuasive messages, ultimately influencing their decision-making and attitude formation.
The Elaboration Likelihood Model (ELM) describes two distinct routes for processing information: the central route and the peripheral route These routes exist on a continuum that reflects the level of cognitive effort applied to a message The central route entails a detailed and systematic analysis of the message's content and logic, while the peripheral route involves a more cursory evaluation based on external cues, such as the attractiveness of the source or emotional appeals A person's motivation and ability to process information influence their likelihood of elaboration, ultimately determining which route of persuasion they will follow.
The Elaboration Likelihood Model (ELM) encompasses two distinct operations: the central route, which evaluates the merits of a message's arguments, and the peripheral route, which relies on cognitive resources and associations influenced by peripheral cues (Lin and Lin, 2018) ELM suggests that peripheral variables significantly affect personal beliefs, such as perceived usefulness and attitudes, which in turn explain technology acceptance behavior This theory has been applied in various fields, including social psychology, marketing, IT, and e-commerce, to analyze technology adoption For instance, Bhattacherjee and Sanford examined how argument quality and source credibility impact perceived usefulness and technology acceptance in IT, while Sussman and Siegal focused on the effects of argument and source quality on perceived usefulness in computer-supported channels.
The Elaboration Likelihood Model (ELM) indicates that the effectiveness of persuasion is influenced by factors such as an individual's personal connection to the message, their existing knowledge and attitudes, and potential distractions This model emphasizes the significance of both central and peripheral routes in understanding attitude change, offering a comprehensive framework for analyzing how persuasion operates.
Issue involvement, relevance, commitment, dissonance, arousal, need for cognition, etc
Self-presentation motives, demand characteristics, evaluation apprehension, source characteristics, etc
Distraction, message comprehensibility, issue familiarity, appropriate schema fear, arousal, etc
C OON I T IVE STRU CTURE - CHANGE
NATURE OF COGNITIVE PROCESSING (initial attitude, argument quality, etc.)
RETAIN OR REGAIN INITIAL ATTITUDE
Are new cognitions adopted and stored in memory? Are different responses made salient than previously?
ENDURING NEGATIVE ATTITUDE CHANGE (Boomerang)
Figure 1:: The ELM of persuasion Source: Petty & Cacioppo (1983, p 6)
2.2.3 Theory of Reasoned Action (TRA)
The Theory of Reasoned Action (TRA), developed by Martin Fishbein and Icek Ajzen in 1975, is a social cognitive theory aimed at explaining and predicting human behavior, especially in decision-making contexts As noted by Mitra Karami (2006), TRA emphasizes a consumer's action-oriented attitude over a product or service-oriented perspective The model identifies two key factors influencing behavior: attitudes, which reflect an individual's positive or negative evaluation of a behavior, and objective norms, which denote the social pressures that influence an individual's decision to engage in or refrain from a behavior.
The Theory of Reasoned Action (TRA) posits that individuals form intentions to engage in behaviors by considering their attitudes and objective norms, which significantly influence actual behavior TRA highlights the role of behavioral intentions as a crucial link between attitudes, objective norms, and behavior Although originally not designed for technology adoption research, TRA has been extensively applied in this field, demonstrating its effectiveness in understanding consumer behavioral intentions (Davis, Bagozzi, and Warshaw, 1980; Igbaria, 1993).
The Theory of Reasoned Action suggests that individuals are more inclined to perform a particular behavior when they hold a favorable attitude towards it and believe that significant others anticipate their participation in that behavior.
Cloud security is essential for protecting data, applications, and infrastructure in cloud computing environments against various threats As organizations adopt cloud services for data storage and processing, implementing strong security measures is crucial to safeguard sensitive information and maintain trust This complex issue encompasses multiple perspectives and challenges, particularly regarding the loss of physical security control and data isolation in public clouds, as highlighted in the study by D Chen and H Zhao.
(2012), underscores the importance of robust security measures to protect sensitive accounting data Cloud vendors’ efforts in implementing encryption and access controls are essential to safeguard against unauthorized access.
Multi-tenancy, a fundamental aspect of cloud computing, presents unique security challenges, particularly regarding data isolation and confidentiality In shared cloud environments, concerns arise over side-channel attacks that exploit common resources to access sensitive information To safeguard accounting data in the public cloud, encryption is essential, necessitating robust key management practices to prevent unauthorized access to encryption keys Additionally, Identity and Access Management (IAM) is crucial for cloud security, ensuring that only authorized users can access sensitive accounting information Effective IAM solutions are vital for preventing unauthorized access and managing data exposure.
Organizations using public cloud services face several critical challenges, primarily concerning the security and privacy of sensitive data, which necessitates a high level of trust in the cloud provider's security measures to prevent unauthorized access and breaches Public clouds are attractive targets for cyber attackers due to the concentration of data, risking significant damage to an organization's reputation and customer trust in the event of a breach Additionally, navigating compliance with various industry and regional data handling regulations can be complex, especially when cloud providers operate globally and store data in multiple locations Limited control over the underlying cloud infrastructure poses a concern for organizations with stringent security requirements or specific compliance needs Performance issues, such as network latency, can also affect application responsiveness, particularly for data-intensive or time-sensitive applications Lastly, organizations must consider the risk of vendor lock-in, as migrating data and applications to a public cloud can limit future flexibility and create dependency on a single provider.
Research Concepts
2.3.1 The impact of factors on storage behavior
Cloud computing is an innovative model that delivers on-demand computing solutions via the Internet, offering significant advantages such as document and data synchronization, file sharing, and efficient resource utilization It facilitates the consolidation and deployment of activities through online transactions, serving as a viable alternative or supplement to Enterprise Resource Planning (ERP) systems During the audit process involving cloud users, servers, and third-party auditors, the benefits of cloud computing become clear, including reduced storage and maintenance costs, minimized risk of data loss from hardware failures, and the convenience of accessing data from any location worldwide.
Cloud computing represents a significant transformation in the computing landscape over recent decades, offering users high-quality service through robust infrastructure Factors influencing the adoption of cloud computing include its usefulness, ease of use, ubiquity, peer influence, cost, and performance expectancy.
Hl:Perceived Usefulness positively influences behavioral intention to use cloud computing.
Performance Usefulness refers to an individual's belief that utilizing a system enhances their job performance (Venkatesh et al., 2003) This concept is rooted in several theories, including perceived usefulness from the Technology Acceptance Model (TAM), relative advantage from diffusion of innovation theory, job-fit from the model of PC utilization, outcome expectations from social cognitive theory, and extrinsic motivation from motivation models Research indicates that behavioral intention is significantly influenced by performance expectancy (Raut et al., 2018).
Social influence is defined as the extent to which individuals feel that important others expect them to adopt a new system (Venkatesh et al., 2003) It encompasses objective norms from various theories, including the Technology Acceptance Model (TAM) and the Theory of Planned Behavior (TPB), as well as social factors and image from the diffusion of innovation theory In our research, we specifically examine the influence of peers on users' decisions Previous studies have indicated that individuals' choices to store personal information in the cloud are significantly affected by their peers (Alsmadi and Prybutok, 2018) Numerous researchers have highlighted that these social factors directly impact behavioral intentions regarding technology usage (Nguyen et al., 2014; Wang et al., 2017).
H2: Perceived Ease of Use positively influences behavioral intention to use cloud computing.
Perceived Ease of Use (PEOU) is a vital factor in technology adoption, referring to users' expectations regarding the effortless nature of a system or technology (Davis, 1989) This study specifically examines users' perceptions of cloud computing, assessing how easy they believe it is to navigate and utilize this technology The easier cloud computing is perceived to be, the more likely users are to adopt these technological innovations.
User perception of how easily they can interact with technology significantly influences their intention to use it Research indicates a strong positive correlation between Perceived Ease of Use and Behavioral Intention to Use cloud computing Users who find cloud computing systems intuitive and user-friendly are more inclined to utilize them regularly.
H3: Perceived Ease of Use positively influences Perceived Usefulness.
Perceived usefulness in technology adoption is significantly influenced by job relevance, as highlighted by Venkatesh and Davis (2000) and further supported by Venkatesh and Bala (2008) The Elaboration Likelihood Model (ELM), developed by Petty and Cacioppo in 1986, provides insights into how stimuli processing affects attitude change Bhattacherjee and Sanford (2006) utilized this model to explore how work relevance moderates the relationship between attitude and perceived usefulness in IT adoption, indicating that users are more likely to embrace technology they find relevant to their jobs Research by Kim (2008) and Kim and Garrison (2009) reinforces this notion, showing that mobile technology's impact on job-related tasks shapes user attitudes and behaviors Therefore, it is essential to investigate the effect of job relevance on the relationship between perceived usefulness and the intention to use cloud computing.
H4: Peers’ Influence will have a significant positive influence on the behavioral intention toward the behavior.
Peer influence is a person's behavior or decisions influenced by the interac- tion of their peer group or social circle A person’s social circle usually includes the individual’s friends (Khare & Pandey, 2017).
Peer influence significantly affects behavior, highlighting its importance in understanding user interactions with technology This study enriches the literature on information systems and cloud computing, demonstrating that users perceive these technologies as effective in safeguarding their information The analysis of the developed model offers valuable insights for researchers in both academic and industrial settings.
Security remains a significant concern in cloud computing, primarily related to the confidentiality and integrity of information Research indicates that security issues are influenced not only by technology but also by users' perceptions and awareness levels Perceived security refers to users' beliefs in the effectiveness of cloud computing security practices to safeguard their information Studies have shown that perceived security is a critical component of risk beliefs in this domain.
H5: Perceived Privacy will have a significant positive influence on perceived risk.
Perceived privacy in cloud computing is the belief that personal information is secure, supported by strong privacy policies from cloud service providers and their third-party partners However, privacy concerns remain a significant risk in cloud computing, largely due to its dynamic environment that relies on virtualization, remote processing, and storage.
The hypothesis suggests that users who recognize robust privacy measures in cloud computing are likely to associate it with reduced risk levels when utilizing cloud services This aligns with previous research indicating that perceived privacy significantly influences perceived risk in technology adoption (Chang et al., 2017; Miltgen et al., 2013).
H6: Perceived Risks will have a significant negative influence on the behavioral intention toward the behavior.
Perceived risk plays a significant role in a firm's decision to adopt cloud computing technology, as outlined by Hsu et al (2014) Key concerns include data lock-in, confidentiality issues, inadequate service quality guarantees, bandwidth limitations, and reliability challenges (Armbrust et al., 2010; Hsu et al., 2014) Additionally, Sabi et al (2016) emphasize that technological attributes such as data security and associated risks can directly influence the technology's adoption and diffusion Furthermore, Venters and Whitley (2012) suggest that the cloud can significantly impact an organization's IT structure and interfaces, introducing critical risks in areas like identity management, governance, compliance, and security responses.
Therefore perceived risks may have a negative effect on behavioral intention to use cloud computing.
Research model and hypothesis
The authors propose a research model that integrates the theoretical foundations and summarizes internal and external studies related to financial challenges Each component of the model is interconnected, and the authors plan to conduct a test to evaluate the validity of these relationships.
Based on the theoretical basis, the research hypotheses are proposed as follows:
Hl: Perceived Usefulness positively influences behavioral intention to use cloud computing.
H2: Perceived Ease of Use positively influences behavioral intention to use cloud computing.
H3: Perceived Ease of Use positively influences Perceived Usefulness.
H4: Peers' Influence will have a significant positive influence on the behavioral intention toward the behavior.
H5: Perceived Privacy will have a significant positive influence on perceived risk.
H6: Perceived Risks will have a significant negative influence on the behavioral intention toward the behavior.
Conclusion of Chapter 2
Researchers have explored various aspects of cloud security in their published works, addressing topics such as architectural components, attack vectors, challenges, threats, vulnerabilities, observed attacks, and proposed solutions Each study delves into specific security and privacy issues within the cloud computing environment Despite the breadth of research, it seems that there are currently no viable solutions to these pressing concerns.
The three primary theoretical frameworks for understanding technology adoption are the Technology Acceptance Model (TAM), the Elaboration Likelihood Model (ELM), and the Theory of Reasoned Action (TRA) TAM is valued for its simplicity, focusing on perceived usefulness and ease of use, which facilitates its application for both researchers and practitioners This user-centered model underscores the importance of user beliefs and attitudes in technology adoption, making it effective for designing user-oriented technology Its versatility allows it to be applied across various technology domains and user groups In contrast, the ELM offers insights into persuasion processes, emphasizing the role of both central and peripheral factors in shaping attitudes Similarly, the TRA asserts that individuals are more likely to engage in a behavior if they hold a positive attitude toward it and believe that significant others expect them to do so.
Cloud computing represents a significant evolution in technology over the last two decades, providing users with top-tier services through robust cloud infrastructure Various factors, such as usefulness, ease of use, peer influence, risks, and privacy concerns, can impact the adoption of cloud computing.
Research Process
Figure 3: Research process of the group of authors
Research Methodology
The author investigates the factors influencing the adoption of cloud computing for data storage among accounting and auditing professionals in Ho Chi Minh City, with a particular focus on information security concerns.
The author employs a normative method to select a research sample focused on the impact of cloud-based accounting data storage and its implications for accounting information security The sample comprises university and college students, as well as members of accounting and auditing firms in Ho Chi Minh City.
Research Size: When using the factor analysis method in the data analysis process, the sample used in the study must be large enough Minimum sample size is
100, preferably 120, and the authors collected 614 observations (no missing data) with
The study utilized a questionnaire featuring 39 questions to gather insights from survey respondents regarding cloud computing issues After identifying 11 responses indicating "Never use cloud," the authors opted to exclude these observations, resulting in a total of 614 valid responses for analysis.
The measurement model evaluation method utilizes Stata 20 software to analyze the complete dataset, focusing on the sequential validation of measurement scales This process begins with reliability testing using Cronbach's Alpha, followed by exploratory factor analysis (EFA) to evaluate the scales' validity In EFA, Principal Components Analysis (PCA) with varimax rotation is employed, excluding variables with item-total correlations below 0.3, as per Nunnally and Burnslcin's criteria (1994) Additionally, variables with factor loadings under 0.5 are removed according to Gerbing and Anderson's recommendations (1988), ensuring that the total extracted variance is at least 50%.
Following the exploratory factor analysis (EFA), the observed variables are evaluated, and those that do not meet the necessary criteria are discarded The qualifying variables are then analyzed through multivariate regression to assess the measurement model's adequacy within the structural model (NC model) Ultimately, the hypotheses are confirmed or denied based on the analysis results.
Scale Development
3.3.1 Scale Development for Independent Variables:
The measurement scale in research involves utilizing numerical metrics to express various phenomena To investigate the factors affecting the intention to use Cloud for accounting data storage, this study introduces a quantitative measurement scale for the relevant research concepts The Likert Scale, established by Likert in 1932, is a prevalent tool in quantitative research In this approach, respondents evaluate a series of statements related to each research concept on a 5-level scale, translating their perceptions into numerical values for analysis.
5-point Likert Scale ranging from 1 to 5 is defined as follows: 1 - Strongly Disagree, 2
- Disagree, 3 - Neutral, 4 - Agree, 5 - Strongly Agree.
To effectively address customer privacy concerns in direct shopping transactions, companies must implement robust privacy protection mechanisms that ensure confidentiality of customer information, appropriate collection and use of personal data, acquisition of only necessary information, and non-disclosure or sale of personal information to third parties Building on Chen's (2007) research, this study utilizes a research scale that encompasses three key component factors.
ASECU: I carefully assess and evaluate the level of security and data integrity before selecting a tool to use.
ARELI: I consider the reliability of the security in cloud storage sendees before choosing a tool to use.
AREPU: The credibility of the provider in terms of security and privacy greatly influences my decision to use a particular tool.
The authors selected Handayani et al (2017) and the Technology Acceptance Model (TAM) by Venkatesh and Davis (2000) to effectively illustrate the concept of perceived usefulness in the context of cloud computing for data storage This framework provides a measurement scale that evaluates how users perceive the benefits of utilizing cloud technology for their storage needs.
BREADY: I assess the readiness of cloud storage services before choosing to use them.
BRECOV: 1 choose to use cloud storage services when they incorporate backup and data recovery functions in case of incidents.
BUSEFUL: Using the Cloud is very convenient for me.
BEFFIC: Utilizing the Cloud for data storage enhances my work efficiency.
Perceived Ease of Use: According to the Technology Acceptance Model
The Technology Acceptance Model (TAM) evaluates ease of use by considering factors such as learnability, usability, comprehensibility, memorability, availability of guidance, and accessibility This framework enables users to utilize cloud data storage without needing extensive technological expertise Handayani et al (2017) conducted a study using a one-way outcome-focused measurement scale, adapted from the original model.
"perceived ease of use" scale developed by Davis (1989) Building on the aforementioned studies and incorporating findings from Shailja Tripathi’s research
(2019), the authors have decided to select a research measurement scale composed of the following three observed variables:
CEASI: For me, using the Cloud is easy.
CINTER: I interact with and utilize the features of the Cloud proficiently.
COVERALL: Overall, I can easily use the Cloud.
CMIND: I feel that using the Cloud reduces the need for excessive thinking while working.
Perceived risks in online shopping and the use of bank cards, as discussed by Hsu et al (2014), encompass feelings of insecurity and the potential loss of information These risks are particularly significant for cloud computing users, where uncertainty arises from the possibility of losing or having personal or business data stolen Ogan M Yigitbasioglu (2014) introduced a measurement scale for assessing perceived risks associated with cloud computing data storage Building on previous studies, the authors developed a preliminary measurement scale that includes four observed variables.
DINFORSECU: Using cloud storage services may pose risks to information security.
DINFORRISK: Data leakage is one of the underlying risks when using cloud storage services.
DACCESS: Access rights might encounter issues when using cloud computing services.
DWEAKSECU: Current security measures are not sufficiently robust to prevent data leakage risks in cloud computing.
Peer influence: The influences of others are variables based on the social pressure, as studied by Ari & Yilmaz (2017) Additionally, the research conducted by
According to Venkatesh & CTG (2003), an individual's attitudes and beliefs within a group significantly impact their behavior regarding the use of a specific system Additionally, Sander Schovevilie (2007) identified that individuals often emulate the behaviors of those around them when they observe their usage Drawing from the draft scales of these studies, the authors developed a scale to measure the "peer influence" factor, which includes three key elements.
EREPLY: I research and consult reviews and feedback from other users about the storage tool before deciding to use it.
EPOPULAR: I have a tendency to use a specific cloud for storage if many of my acquaintances use it.
EACCQUAIN: I typically use a specific cloud if that cloud is popular among many other users.
3.3.1 Scale Development for the Dependent Variable:
This study focuses on Behavioral Intention, a concept rooted in the research of Ajzen & Fishbein (1980), which seeks to explain human behavior in planning beneficial actions Previous analyses, such as those by Hom, Katerberg, & Hulin (1979) and Sperber, Fishbein, & Ajzen (1980), have also explored this concept Handayani et al (2017) utilized a scale developed by Davis (1980) to assess the intention to use cloud-based accounting information systems Following this, the authors adopt the Handayani et al (2017) scale to evaluate the "Behavioral Intention of using Cloud to store accounting data," which includes four key variables.
FLIKE: I opt to utilize a specific Cloud platform to store my data simply because I enjoy using it.
FHABIT: I frequently use a specific Cloud platform to store my data out of habit.
FREL1: I intend to continue using the Cloud if the Cloud provider ensures its reliability and information security level.
FOBLIGA: Cloud is highly important and pertinent to forthcoming jobs and tasks, hence utilizing the Cloud is an obligation to me.
Conclusion of Chapter 3
Chapter 3 provides a strong methodological framework for this research, ensuring robust analyses and promising valuable insights into cloud adoption in accounting and auditing in Ho Chi Minh City, outlines the systematic approach used in this study to investigate factors affecting the adoption of cloud computing for accounting and auditing data storage in Ho Chi Minh City The research sample was selected using a normative method, targeting a diverse group of participants A total of 614 observations were included in the analysis after removing 11 observations related to non-cloud users.
The quantitative research methodology employs reliability testing through Cronbach’s Alpha and exploratory factor analysis (EFA) to validate the scales used in the study Principal Components Analysis (PCA) assists in selecting relevant variables, while multivariate regression analysis assesses the measurement model Key constructs such as Perceived Privacy, Perceived Usefulness, Perceived Ease of Use, Perceived Risks, and Peer Influence are introduced, each associated with specific observed variables Additionally, the chapter elaborates on the development of the Behavioral Intention construct, utilizing a Likert scale to encompass variables related to liking, habit, reliability, and obligation, thereby offering valuable insights into the factors influencing cloud adoption.
ANALYSIS AND RESULT
Research Sample Description
Between July 25 and August 5, 2023, the authors conducted a questionnaire-based study involving students and professionals in Ho Chi Minh City to explore cloud computing The survey consisted of 39 questions, resulting in 625 responses, all complete However, 11 responses indicated "never using cloud computing," leading the authors to exclude these from the analysis, ultimately focusing on 614 valid observations.
The descriptive statistics reveal key demographic insights, indicating that 60.42% of the population is female, while 35.99% is male, with 3.59% preferring not to specify their gender The largest age group comprises individuals aged 21 to 30, accounting for 49.02%, followed by those aged 0 to 20 at 26.55%, 31 to 40 at 18.24%, and those over 40 at 6.19% A significant majority, 85.02%, of the 614 valid observations reside in Ho Chi Minh City, while 14.98% live elsewhere In terms of employment, 39.25% of participants are students, 24.27% work as auditors, 25.57% are accountants, and 10.91% are engaged in other occupations.
N/A (Chose not to reveal their gender) 22 3.59
Table 1: Descriptive Analysis of the Demographic information
Currently living in HCM City
The data reveals the frequency of cloud usage among respondents, highlighting that 257 individuals utilize cloud services regularly (5-6 times per week), while 173 use it frequently (3-4 times per week) Additionally, 136 people access the cloud sometimes (3-4 times per month), and 48 individuals use it seldom (1-2 times per month).
A recent survey revealed that 84.04% of participants were familiar with Google Drive, making it the most recognized brand in cloud storage services In comparison, 60.42% of respondents were aware of OneDrive, while 34.69% recognized Amazon Drive Additionally, 31.43% of the population reported awareness of Dropbox.
A survey revealed that 288 participants, representing 46.31%, were familiar with iCloud Among them, 382 users utilized the cloud for image storage, accounting for 62.21% of the various data types stored Additionally, 362 individuals, or 58.96%, used the cloud to store video data.
A significant majority of users, 543 individuals or 88.44%, utilized cloud storage for saving various document types such as Excel, Word, and PDF files Additionally, 166 users, accounting for 27.94%, opted to store audio data in the cloud.
Table 2: Frequency of using Cloud and Common types of stored files.
Document (Excel, Word, PDF, etc)
The result of Cronbach’s alpha reliability test
The authors utilized Cronbach’s alpha coefficient to evaluate the reliability of the estimated parameters within each factor group of the dataset Variables exhibiting inter-item correlations below 0.3 were excluded, and a measurement scale was chosen if the Cronbach's alpha coefficient surpassed 0.6, as recommended by Nunnally and Bernstein.
An alpha coefficient above 0.8 demonstrates strong reliability for a measurement scale, while coefficients between 0.7 and 0.8 indicate acceptable reliability Values ranging from 0.6 to 0.7 reflect reasonable reliability, particularly for new studies.
"Perceived Privacy" measurement scale using the Cronbach's alpha coefficient approach arc as follows:
Table 3: Result of Cron bach's alpha of “Perceived Privacy” variables
Item Observations Sign Item-test
The analysis of the "Perceived Privacy" measurement scale, which includes the observed variables ASECU, ARELI, and AREPU, reveals a Cronbach's alpha coefficient of 0.7608, indicating strong reliability as it exceeds the threshold of 0.6.
Results of the reliability testing for the "Perceived Usefulness" measurement scale using the Cronbach's alpha coefficient approach are as follows:
Item Observations Sign Item-test
Table 4: Result of Cronbach’s alpha of “Perceived Usefulness1' variables
The analysis of the "Perceived Usefulness" measurement scale, which comprises four observed variables, reveals a Cronbach’s alpha coefficient of 0.7984, indicating a good level of reliability as it approaches the 0.8 threshold Furthermore, the individual Cronbach’s alpha coefficients for each observed variable are all below 0.7984 Consequently, these variables will be retained for subsequent analysis.
Results of the reliability testing for the "Perceived Ease of Use" measurement scale using the Cronbach's alpha coefficient approach are as follows:
Table 5: Result of Cron bach’s alpha of “Perceived Ease of Use" variables
The "Perceived Ease of Use" measurement scale demonstrates a Cronbach's alpha coefficient of 0.7896, surpassing the 0.6 threshold and fulfilling the necessary criteria for reliability Additionally, this coefficient is greater than the other alpha coefficients measured in the study.
Item Observations Sign Item-test
Test scale 5654685 0.7896 individual observed variables As a result, these variables will be retained and included in the subsequent analysis process.
Results of the reliability testing for the "Perceived Risks" measurement scale using the Cronbach's alpha coefficient approach are as follows:
Table 6: Result of Cronbach ’s alpha of “Perceived Risks" variables
Item Observations Sign Item-test
The Cronbach's alpha coefficient for this factor group is 0.7573, indicating strong reliability as it exceeds the acceptable threshold of 0.6 and is higher than the alpha coefficients of other observed variables Consequently, the authors have decided to retain all variables in this measurement scale and will move forward with testing the Cronbach's alpha for the "Peer Influence" scale.
Results of the reliability testing for the "Peer Influence" measurement scale using the Cronbach's alpha coefficient approach are as follows:
Item Observations Sign Item-test
Table 7: Result of Cron bach \ alpha of “Peer influence" variables
The testing of the Cronbach's alpha coefficient for the "Peer Influence" measurement scale revealed that the EREPLY variable has a coefficient of 0.7484, surpassing the factor group's coefficient of 0.7272 Consequently, to enhance reliability, the EREPLY variable will be excluded from the analysis.
Results of the reliability testing for the "Behavioral Intention" measurement scale using the Cronbach's alpha coefficient approach are as follows:
Table 8: Result ofCronbach’s alpha of “Behavioral Intention ” variables
Hem Observations Sign Item-test
The Cronbach's alpha coefficient for the "Behavioral Intention" measurement scale was found to be 0.6473, surpassing the factor group's coefficient of 0.6166 Consequently, the observed variable will be excluded from the analysis.
The authors conducted a reliability analysis using the Cronbach's alpha coefficient for five independent factors: Perceived Privacy, Perceived Usefulness, Perceived Ease of Use, Perceived Risks, and Peer Influence, alongside the dependent factor Behavioral Intention To enhance the reliability of the measurement scales, the variables FRELI and EREPLY were excluded from the study.
Factor Analysis Exploration Results (EFA)
Exploratory Factor Analysis (EFA) evaluates convergent and discriminant values on a scale, following a reliability assessment (Nguyen Dinh Tho, 2011) To effectively conduct EFA, specific conditions must be satisfied.
The KMO Test (Kaiser-Meyer-Olkin) is an essential index that assesses the strength of the correlation between measurement variables relative to their unique partial correlations, with a recommended KMO value range of 0.5 to 1 for effective Exploratory Factor Analysis (EFA) Additionally, Bartlett's Test evaluates the correlation among observed variables, with the null hypothesis stating that there is no correlation (H^: correlation = 0) A statistically significant result (Sig < 0.05) leads to the rejection of the null hypothesis, confirming the existence of relationships among the observed variables.
Evaluation of scale values following the assessment of correlations among measurement variables:
Total Variance Explained by Factors: Stop extracting factors when the eigenvalues are greater than 1 and accept when the total extracted variance is more than 50%.
The authors performed Bartlett's test and KMO test for the independent variables, and the results are presented in the table below:
Bartlett test of sphericity Kaiser-Meyer-Olkin
Table 9: Bartlett’ s test and KMO test results for independent variables
The results of Bartlett's test reveal a chi-square statistic of 5033.077 with a p-value of 0.000, indicating statistical significance as it is less than 0.05 Additionally, the KMO measure is 0.949, which is well above the threshold of 0.5 These findings confirm that the data is appropriate for Exploratory Factor Analysis (EFA) and that the independent variables exhibit significant correlations Consequently, the authors proceeded with the EFA analysis based on the observed variables of the independent factors.
Factor Eigenvalue Difference Proportion Cumulative
Table 10: Results of EFA Analysis for independent variables
An eigenvalue threshold of 8.08680, exceeding 1, suggests that only one factor is derived from the observed variables, accounting for a total extracted variance of 47.570% of the data's variability, which falls short of the 50% requirement.
Table ll:Factor Loading matrix ofEFA Analysis for independent variables
The factor rotation matrix results indicate that all variables exhibit factor loading coefficients exceeding 0.5, thereby satisfying the necessary criteria This finding highlights the robust correlation between the original scale and the derived factors.
4.3.2 Exploratory Factor Analysis Results for Dependent Variables
Bartlett's Test and KMO Test for the Dependent Variable arc as follow:
Table 12: Bartlett's test and KMO test results far the dependent variable
Bartlett test of sphericity Kaiser-Meyer-Olkin
Bartlett's test yielded a chi-square statistic of 244.228 with a significance level (p-value) of 0.000, indicating a correlation among the dependent variables and validating the appropriateness of conducting exploratory factor analysis (EFA) Consequently, the authors performed EFA on the dependent variables, and the results are detailed below.
Table l3:Results of EFA Analysis far the dependent variable
Factor Eigenvalue Difference Proportion Cumulative
The exploratory factor analysis (EFA) results indicate that the Eigenvalue of 1.75924 exceeds the threshold of 1, confirming its significance Furthermore, the total extracted variance of 58.64% effectively captures the data's variability, surpassing the 50% requirement.
Table 14: Factor Loading matrix ofEFA Analysis for the dependent variables
The factor loading matrix indicates that all dependent variables in the factor group have coefficients exceeding 0.5, meeting the necessary criteria Additionally, the extraction of one factor from the EFA analysis confirms that the "Behavioral Intention" scale has attained convergent validity.
The exploratory factor analysis (EFA) results indicate that both independent and dependent variables are appropriate for inclusion, as they satisfy KMO and Bartlett's test criteria However, the independent variables showed a lack of convergence, evidenced by the extraction of only one factor and a total variance extracted of 47.570%, which is below the 50% threshold In contrast, the dependent variables achieved convergent validity Consequently, the authors can advance to a multivariate regression analysis to evaluate the influence of independent variables on the dependent variable.
Model Fit Assessment, Regression Analysis Results, and Hypothesis Testing
The authors proceeded to calculate the representative mean values for each factor group in order to continue incorporating these representative variables into a multivariate linear regression model, as follows:
The independent variables consist of 5 representative variables: A, B, c, D, E While:
A represents the mean value of the factor group "Perceived Privacy" which includes the observed variables: ASECU, ARELI, AREPU.
The mean value for the factor group "Perceived Usefulness" is represented by B, encompassing the observed variables BREADY, BRECOV, BUSEFUL, and BEFFIC Meanwhile, the mean value for the factor group "Perceived Ease of Use" is denoted by c, which includes the observed variables CEASI, CINTER, COVERALL, and CMIND.
D represents the mean value of the factor group "Perceived Risks" which includes the observed variables: DiNFORSECU, D1NF0RRISK, DWEAKSECU, DACCESS.
E represents the mean value of the factor group "Peer Influence" which includes the observed variables: EPOPULAR, EACCỌUAIN.
F represents the mean value of the dependent variable within the factor group
"Behavioral Intention" which includes the observed variables: FLIKE, FHABIT ,
Table 15.'Pearson Correlation Matrix with Sig values
The results of correlation analysis indicate that the dependent variable F has a positive correlation with all other remaining variables: A, B, c, Đ, and E, specifically as follows:
The factor "Perceived Privacy*' has a correlation coefficient with the factor
"Behavioral Intention" of 0.5191 with a Sig coefficient of 0.0000 < 0.05 This positive correlation coefficient indicates that "Perceived Privacy" is positively correlated with
The factor "Perceived Usefulness" has a correlation coefficient with the factor
"Behavioral Intention" of 0.5 8 76 with a Sig coefficient of 0.0000 < 0.05 This positive correlation coefficient indicates that “Perceived Usefulness" is positively correlated with “Behavioral Intention."
The factor “Perceived Ease of Use" has a correlation coefficient with the factor
"Behavioral Intention" of 0.5756 with a Sig coefficient of 0.0000 < 0.05 This positive correlation coefficient indicates that "Perceived Ease of Use" is positively correlated with “Behavioral Intention."
The factor "Perceived Risks" has a correlation coefficient with the factor
"Behavioral Intention" of 0.5655 with a Sig coefficient of 0.0000 < 0.05 This positive correlation coefficient indicates that "Perceived Risks" is positively correlated with
The factor "Peer Influence" has a correlation coefficient with the factor
"Behavioral Intention" of 0.543 8 with a Sig coefficient of 0.0000 < 0.05 This positive correlation coefficient indicates that "Peer Influence" is positively correlated with
The results of the multivariate linear regression analysis and the Variance Inflation Factor (VIF) test are presented in the tables below:
Table 16:: Results of Assessment of model fitness
The model fitness assessment shows an R-squared value of 0.5952, indicating that the multivariate linear regression model explains approximately 59.18% of the variability in the dependent variable "Behavioral Intention" based on five independent variables Consequently, 40.82% of the variability is attributed to other factors and error.
Table 17:Resu!ts of the Multiple Linear Regression Analysis
Table 18: Results of Variance Inflation Factor (VIF) test
The analysis reveals that variables A, B, C, D, and E exhibit significant correlations with the "Behavioral Intention" variable, as indicated by their p-values being less than 0.05 Additionally, the Variance Inflation Factor (VIF) values for these variables are all below 5, confirming the absence of multicollinearity and suggesting no linear relationships among them.
Based on the regression results, the authors have constructed the regression equation by incorporating the regression coefficients as follows:
According to the regression equation above, the results demonstrate the impact of each factor on the usage behavior of cloud computing (CLOUD) for data storage as follows:
The "Perceived Ease of Use" significantly influences cloud computing usage for data storage, evidenced by a regression coefficient of 0.267 This is closely followed by "Perceived Usefulness" at 0.211, and "Perceived Privacy" with a coefficient of 0.115 Understanding these factors is essential for enhancing user engagement with cloud storage solutions.
Risks" with a coefficient of 0.1600578, and finally "Peer Influence" with a coefficient of 0.0792492.
The impact of each factor on the usage behavior of cloud computing for data storage can be described as follows:
"Perceived Privacy": With a regression coefficient of 0.1150943, an increase of
1 unit in "Perceived Privacy" leads to an increase of 0.1150943 units in "Behavioral
Intention," while other factors remain constant (ceteris paribus).
"Perceived Usefulness": With a regression coefficient of 0.2110881, an increase of 1 unit in "Perceived Usefulness" results in an increase of 0.2110881 units in
"Behavioral Intention", keeping other factors constant (ceteris paribus).
"Perceived Ease of Use": With a regression coefficient of 0.2674672, an increase of 1 unit in "Perceived Ease of Use" leads to an increase of 0.2674672 units in
"Behavioral Intention", assuming other factors remain constant (ceteris paribus).
"Perceived Risks": With a regression coefficient of 0.1600578, an increase of 1 unit in "Perceived Risks" results in an increase of 0.16005 78 units in "Behavioral
Intention", keeping other factors constant (ceteris paribus).
"Peer Influence": With a regression coefficient of 0.0792492, an increase of 1 unit in "Peer Influence" leads to an increase of 0.0792492 units in "Behavioral
Intention", assuming other factors remain constant (ceteris paribus).
Testing hypothesis Hl: Perceived Usefulness positively influences behavioral intention to use cloud computing With a regression coefficient of 0.2110881, t = 4.80, and a Sig coefficient of 0.000 < 0.05, the group of authors accepts hypothesis Hl.
Testing hypothesis H2: Perceived Ease of Use positively influences behavioral intention to use cloud computing With a regression coefficient of 0.2674672, t = 6.16, and a Sig coefficient of 0.000 < 0.05, the group of authors accepts hypothesis H2.
The results of testing hypothesis H3 indicate that Perceived Ease of Use has a positive influence on Perceived Usefulness, with an indirect regression coefficient of 0.7071029 and a t-value of 27.99, demonstrating statistical significance (Sig < 0.05) Additionally, the Pearson correlation coefficient of 0.7493 from the correlation matrix suggests a strong relationship, where increases or decreases in Perceived Ease of Use correspondingly affect Perceived Usefulness Consequently, the authors accept hypothesis H3.
The results of testing hypothesis FI4 indicate that peers' influence significantly positively affects behavioral intention, demonstrated by a regression coefficient of 0.792492, a t-value of 2.82, and a significance level of 0.005, which is less than the threshold of 0.05 Consequently, the authors accept hypothesis 114.
Testing hypothesis H5 reveals that perceived privacy significantly positively influences perceived risks, with an indirect regression coefficient of 0.681273 and a t-value of 22.10, yielding a significance coefficient of 0.000 The correlation matrix further supports this finding, showing a Pearson correlation coefficient of 0.6661, which indicates a direct relationship where increases or decreases in perceived privacy correspond to similar changes in perceived risks Consequently, the authors accept hypothesis H5.
The analysis of hypothesis H6 reveals that perceived risks significantly negatively affect behavioral intention, as indicated by a regression coefficient of 0.16005 78, which should ideally be negative With a t-value of 4.00 and a significance coefficient of 0.000, the authors accept hypothesis H6.
Hl: Perceived Usefulness positively influences behavioral intention to use cloud computing
H2: Perceived Ease of Use positively influences behavioral intention to use cloud computing
H3: Perceived Ease of Use positively influences Perceived Usefulness Supported Positive
H4: Peers’ Influence will have a significant positive influence on the behavioral intention toward the behavior.
H5: Perceived Privacy will have a significant positive influence on perceived risks.
H6: Perceived Risks will have a significant negative influence on the behavioral intention toward the behavior.
Table ì 9: Result of hypothesis tested using Multiple Linear Regression and Pearson Correlation Matrix.
Discussion of Results
Testing the measurement components using Cronbach’s alpha coefficient analysis and exploratory factor analysis (EFA) confirmed the reliability of the measurement scales Subsequently, the model underwent multiple regression analysis to assess and validate the proposed hypotheses.
Perceived Privacy positively influences the usage behavior of cloud computing
When utilizing cloud storage, prioritizing data security is crucial, especially in accounting The security and privacy features of cloud computing significantly influence user behavior, aligning with the findings of Chen's 2007 study.
Perceived Usefulness significantly influences the usage behavior of cloud computing, as supported by Handayani et al (2017) Mutia (2016) highlights that the advantages of cloud computing include document sharing, storage solutions, entertainment services, communication, and social networking, which enhance user productivity on cloud platforms A survey by Microsoft revealed that 92% of participants prefer cloud storage, driven by five key benefits: personalized services, network accessibility, resource storage, elastic speed, and scalable services (Najwa et al., 2020) This trend is also evident among users in Ho Chi Minh City.
Perceived Ease of Use significantly influences the usage behavior of cloud computing platforms, aligning with findings from Davis (1989) and the measurement scale by Handayani et al (2017) This concept highlights that a user-friendly experience enhances flexible interactions with cloud services, particularly in meeting data storage requirements, such as those for accounting and auditing purposes.
The analysis reveals that the Perceived Risk variable contradicts the initial hypothesis (H6), which proposed that perceived risks would negatively affect behavioral intention Instead, the regression coefficient for perceived risk is positive, indicating a positive influence on behavioral intention, contrary to expectations Despite this unexpected outcome, the significance of the coefficient supports the acceptance of the hypothesis Furthermore, Cronbach's Alpha analysis confirms the reliability of the scale for inclusion in the study.
Peer influence significantly enhances the usage behavior of cloud computing platforms This aligns with findings from researchers such as Ari & Yilmaz (2017), Venkatesh & Ctg (2003), and Sander Schoveville (2007), who highlight how the decisions of others can impact an individual's choices Additionally, an individual's cloud computing usage is often shaped by the behaviors exhibited by their peers.
Conclusion of Chapter 4
Chapter 4 presented the research findings After surveying the research subjects, the authors employed a quantitative research method and established an official scale for the research concepts The research results include: (1) Validation of the scale, assessed through Cronbach’s Alpha reliability coefficient, exploratory factor Analysis Exploration Results (EFA), and further validated through multiple regression analysis; (2) Discussions of the research findings are also provided based on the official research results.