Based on the findings of the study, all of the hypotheses were approved.Perceived Usefulness, Perceived Ease of Use, Perceived Playfulness and Environmental Concern both influence Behavi
INTRODUCTION
THEORETICAL FOUNDATIONS AND RESEARCH MODEL 20 2.1 Related Concepts
Factor
A factor is a component that constitutes objects, events, phenomena, or can be understood as a reality or condition that affects the outcome of something (Cambridge Dictionaiy).
Behavioral Intention
Intention plays a crucial role in evaluating the probability of an individual engaging in a specific behavior As noted by Ajzen (1991), intention is inherently motivational and signifies a person's preparedness to undertake a particular action, serving as a direct precursor to the behavior itself.
Behavioral Intention measures the subjective likelihood that an individual will engage in a behavior and can be seen as a specific case of belief (Fishbein & Ajzen,
1967) It is determined by an individual's attitude toward behavior and subjective norms.
A mobile application is software designed to operate on mobile devices like smartphones and tablets, requiring a compatible operating system (Wang et al., 2013) These applications are primarily distributed through platforms managed by the mobile operating system providers, including the Apple App Store, Google Play, Windows Phone Store, and BlackBerry App World (Okediran et al., 2014) Users can also find mobile apps preinstalled on their devices or download them from app stores and the Internet.
Waste sorting at source
Waste sorting at source involves classifying waste at its origin, whether from individuals, households, or organizations, before it is transported This practice is crucial for effective waste management and environmental protection, as it helps reduce pollution and promotes recycling efforts.
The primary goal of effective waste management is to minimize waste generation, optimize resource utilization, and improve the handling of hazardous waste (Tai et al., 2011; Zhang et al., 2019) Source separation plays a crucial role in the overall waste management process, serving as a foundational strategy that reduces the need for intermediate treatment facilities and enhances recovery rates (Calabro & Satira, 2020) Research by Chen et al (2020) indicates that source separation significantly decreases landfill loads and moisture in waste incineration, while also enhancing low calorific value and cutting carbon emissions by at least 66.8% Additionally, a two-stage least squares spatial model analysis by Zhao et al demonstrates that source separation effectively lowers per capita waste generation.
2020) However, challenges remain, such as public attitudes, treatment facilities, and economic benefits.
Artificial Intelligence (Al)
Artificial Intelligence (AI) has been defined in various ways over the decades, making it difficult to pinpoint a precise definition Alan Turing, in his influential 1950 paper "Computing Machinery and Intelligence," suggested that a machine can be deemed intelligent if it cannot be distinguished from a human during a conversation by an unbiased observer Today, AI typically refers to a machine's capability to communicate, reason, and act independently in both familiar and novel situations, often surpassing current methodologies Furthermore, many contemporary references to AI are frequently used interchangeably with "machine learning" or "deep learning."
Theoretical Basis
TAM, introduced by Davis (1986), is an adaptation of Theory of Reasoned Action (TRA) designed specifically to model user acceptance of information systems
In 1989, Davis, Bagozzi, and Warshaw advanced the Technology Acceptance Model (TAM) as a successor to the TRA, aiming to elucidate the key factors influencing computer acceptance and user technology behavior on a broad scale (P c Lai, 2017) The model serves not only for predictive purposes but also for understanding the reasons behind a system's lack of acceptance, enabling researchers and practitioners to implement effective corrective measures Ultimately, TAM provides a framework for analyzing how external factors shape internal beliefs, attitudes, and intentions regarding technology use.
The Technology Acceptance Model (TAM), proposed by Davis, Bagozzi, and Warshaw (1989), evaluates two crucial beliefs that influence IT acceptance: "Perceived Usefulness" (PU) and "Perceived Ease of Use" (PEU) PU refers to the extent to which an individual believes that utilizing a specific system will improve their job performance, while PEU denotes the belief that using the system will require minimal effort These beliefs significantly impact an individual's Behavioral Intention and Actual Behavior regarding technology adoption.
Figure 2.1 Technology Acceptance Model (TAM)
(Source: Davis, Bagozzi, and Warshaw, /989)
The Modified Technology Acceptance Model (TAM) is a prominent framework for evaluating user acceptance of technology, with its effectiveness supported by numerous studies, including those by Fishbein et al (1980), Davis et al (1989), Moon and Kim (2001), and Chen and Chao (2011).
Smart recycling is an innovative, technology-driven advancement in waste management (Xue et al., 2019) In their research, Liyuan Liu and Yen Hsu (2022) utilized the Modified Technology Acceptance Model (TAM) to assess public intentions towards smart recycling systems, emphasizing four key factors: Perceived Usefulness, Perceived Ease of Use, Environmental Concern, and Perceived Playfulness This study aims to apply the Technology Acceptance Model to investigate the willingness to use mobile applications that facilitate waste classification at the source through AI technology, focusing primarily on Perceived Usefulness and Perceived Ease of Use.
2.2.2 Theory of Planned Behavior (TPB)
Ajzcn and Fishbcin expanded the Theory of Reasoned Action (TRA) by introducing Perceived Behavioral Control to create the Theory of Planned Behavior, which aims to predict an individual's intention to perform a specific behavior at a given time This theory is designed to explain behaviors that individuals can control (Lamorte, 2019).
Before deciding to use a product, consumers typically navigate through five key stages: recognizing a need, searching for information, evaluating alternatives, making a purchase decision, and exhibiting purchase behavior (Kotler, 2010) Consequently, how consumers perceive and believe in the importance of attributes associated with AI-equipped mobile applications versus traditional methods can significantly influence their behavior in utilizing waste classification applications The Theory of Planned Behavior (TPB) suggests that an individual's actions are shaped by their attitudes towards the behavior and the social norms related to it.
According to the Theory of Planned Behavior (TPB), individual actions are influenced by three key factors: (1) Attitude, which reflects a person's emotional evaluation of a behavior based on the belief in its benefits; (2) Subjective norm, which encompasses an individual's perception of social pressures from significant others, such as family and friends, regarding the behavior; and (3) Perceived behavioral control, which relates to the individual's belief in their ability and resources to execute the behavior Enhanced perceived control occurs when individuals feel they possess sufficient resources to perform the desired action.
Figure 2.2 Theoiy of Planned Behavior (TPB)
Ajzen's Theory of Planned Behavior (1991) asserts that behavior is influenced by both intention and perceptions of behavioral control This theory has been extensively utilized in research exploring the connections between beliefs, attitudes, behavioral intentions, and actual behaviors across various fields Previous studies have applied TPB and its extended models to analyze recycling behavior (e.g., Botetzagias et al., 2015; Chan and Bishop, 2013; Cheung et al., 1999; Nigbur et al., 2010; Oztekin et al., 2017; Tonglet et al., 2004) In this study, we aim to leverage this framework to investigate how factors within the Technology Acceptance Model (TAM) affect individuals' intentions to use AI-powered mobile applications for waste classification at the source.
Overview of Related Studies
The study by Liu and Hsu (2020) explores the motivating factors influencing the public's intention to use smart recycling systems in China, utilizing a modified version of the Technology Acceptance Model (TAM) It highlights the roles of perceived playfulness and environmental concern as significant factors affecting user engagement with these innovative recycling solutions.
The research surveyed residents of local communities in Ningbo City equipped with smart recycling systems The questionnaire was uploaded to a survey website
I don't know!
230 were deemed valid after a data quality control process, with a valid response rale of 60.8%.
Accordingly, the authors adopted two factors from the TAM - "Perceived Usefulness" (PU) and "Perceived Ease of Use" (PEU) (Davis, Bagozzi, and Warshaw,
In 1989, the original Technology Acceptance Model (TAM) was enhanced by incorporating two additional factors, "Environmental Concern" (EC) and "Perceived Playfulness" (PP), to improve the explanatory power regarding the dependent variable "Intention to Use" (IU) These factors are central to this study's focus.
Figure 2.3 Research Model of Liu, L., Hsu, Y (2020)
This study utilized Structural Equation Modeling (SEM) to analyze the relationships between variables and the intention to use smart recycling systems The findings indicated that Perceived Ease of Use (PEU) significantly influenced Perceived Usefulness (PU) Additionally, both Environmental Concern (EC) and PU, along with personal preferences (pp), had a significant impact on Intention to Use (IU) However, PEU did not significantly affect EC.
Hypothesis testing reveals that Environmental Concern significantly influences the Intention to Use smart recycling systems, which are closely linked to environmental protection, waste reduction, and effective recycling practices Increased environmental awareness among individuals correlates with a higher likelihood of adopting smart recycling solutions Additionally, Perceived Playfulness plays a crucial role in shaping the Intention to Use, as the integration of technology and gamification elements, such as rewards, enhances user enjoyment and engagement Consequently, prioritizing Perceived Playfulness is essential for researchers and designers of smart recycling systems.
2.3.1.2 Research by Ramzan, Sidra & Liu, Chenguang & Xu, Yan & Munir, Hina & Gupta, Bhumika (2021)
The 2021 study by Ramzan et al investigates the adoption of online e-waste collection platforms among Chinese millennials, emphasizing their role in enhancing environmental sustainability The research highlights the importance of digital solutions in promoting responsible e-waste disposal practices, aiming to engage younger generations in environmental stewardship By analyzing millennials' attitudes and behaviors towards e-waste management, the study provides valuable insights into how technology can foster sustainable practices and contribute to a greener future.
China conducted a study to understand young consumers' perceptions and adoption of an online platform for e-waste collection, utilizing a conceptual framework that integrates the Theory of Planned Behavior (TPB), the Technology Acceptance Model (TAM), and Perceived Risk The research, carried out from July to November 2018, involved a survey of 807 Millennials born between 1982 and 2002 across various capital cities in China A total of 1,800 questionnaires were distributed through both paper and online channels, resulting in a final sample size of 807 valid responses and an effective recovery rate of 44.8%.
The authors utilized a 5-point Likert scale, from 1 (strongly disagree) to 5 (strongly agree), and applied the Partial Least Squares Structural Equation Modeling (PLS-SEM) method using Smart PLS 3.2.6 to analyze survey data through a two-step approach.
Figure 2.4 The researeh model of Ramzan et al (2021)
The findings reveal that the Chinese Millennial generation generally has a favorable attitude towards using an online e-waste collection platform Key factors influencing this positive intention include "Perceived Usefulness," which significantly enhances the likelihood of adopting online recycling methods when consumers recognize the benefits of the platform Additionally, "Perceived Ease of Use" plays a crucial role in shaping "Perceived Usefulness." Other important influences on the intention to use the platform include "Attitude," "Subjective Norm," and "Perceived Behavioral Control," all of which have a positive impact Conversely, "Perceived Risk" negatively affects "Perceived Usefulness," "Attitude," and the overall intention to adopt the platform.
"Attitude" was not statistically significant However, the relationship between
"Perceived Ease of Use" and "Attitude" towards accepting the online collection platform was mediated by "Perceived Usefulness".
2.3.1.3 Research by Dan Cudjoe, Huiming Zhang, Hong Wang (2023)
The study by Dan Cudjoe, Huiming Zhang, Hong Wang (2023) titled
The study on predicting residents' adoption intention for a smart waste classification and collection system in China, focusing on Shanghai as a case study, aims to identify key factors that influence Chinese residents' willingness to utilize this innovative waste management solution.
This study aimed to offer detailed policy recommendations to enhance public engagement in local smart waste classification initiatives Using Structural Equation Modeling (SEM) for data analysis, the research employed IBM SPSS 26 and AMOS 24 software SPSS 26 specifically analyzed the socio-economic characteristics of the respondents Data collection occurred between August 20, 2022, and November 26, 2022, resulting in 514 valid questionnaires after excluding inaccurate surveys A 5-point Likert scale was utilized, with responses ranging from 1 ("strongly disagree") to 5 ("strongly agree").
Dan Cudjoe, Huiming Zhang, Hong Wang (2023) also inherited the Theory of Technology Acceptance Model from Davis, Bagozzi, and Warshaw (1989), namely:
The integration of Ajzen's (1991) Theory of Planned Behavior, which includes "Subjective Norm," "Perceived Behavioral Control," and "Attitude," with the concept of "Perceived Usefulness" enhances the understanding of "adoption intention." By incorporating "Community Engagement" into this framework, the authors aim to strengthen the explanatory power of "Attitude" as a key factor influencing the intention to adopt new behaviors or technologies.
Figure 2.5 The research model of Dan Cudjoe, Huiming Zhang, Hong Wang (2023)
(Source: Dan Cudjoe, Huiming Zhang, Hong Wang, 2023)
The research findings indicate that "Perceived Usefulness", "Altitude",
"Subjective Norm" and "Perceived Behavioral Control" positively influence people's attitudes and intentions towards adopting a smart waste classification and collection system.
The study found that while community engagement does not directly shape people's intentions, it significantly influences their acceptance intention through the formation of attitudes This indicates that community engagement acts as an external factor that initially impacts individuals' attitudes, which subsequently affects their intention to adopt Furthermore, this influence may stem from the cognitive aspects of engagement, which effectively resonate with individuals on emotional and behavioral levels.
Furthermore, the study confirmed a positive relationship between community engagement and people's attitudes, leading to an intention to foster environmentally friendly behaviors.
2.3.1.4 Research by Kelly K de Wildt, Marijn H.c Meijers (2023)
A study by Kelly K de Wildt and Marijn H.C Meijers (2023) in the Netherlands investigated the effects of a mobile application on recycling behavior, focusing on knowledge, self-esteem, intentions, and actual recycling practices The research also aimed to identify whether the Original Theory of Planned Behavior (TPB), Technology Acceptance Model (TAM), or an extended TAM best predicts the app's usage Utilizing the Experience Sampling Method (ESM), the study involved 118 participants who provided feedback at specific intervals over two weeks: the first week assessed their recycling habits before app usage, while the second week evaluated their behavior after they began using the application.
The study reveals that the application successfully enhanced recycling behavior among users Nonetheless, the Theory of Planned Behavior (TPB) and the Technology Acceptance Model (TAM) proved effective only in predicting users' intentions to utilize the app, rather than their actual usage patterns This indicates that while these models are valuable for understanding intentions, they fall short in accurately forecasting real-world application usage.
Research Model and Hypotheses
2.4.1.1 The relationship between Perceived Ease of Use, Perceived Usefulness and Behavioral Intention in the Technology Acceptance Model (TAM)
Perceived Ease of Use is an individual's belief that using a particular system or technology can be done comfortably without much physical and mental effort (Davis,
Perceived Ease of Use significantly influences users' interactions with technology, as it reflects the effort required to engage with systems like smart recycling and shared e-bike platforms Research indicates that a higher Perceived Ease of Use leads to increased Perceived Usefulness, encouraging users to explore features and maintain ongoing usage (Bassiouni et al., 2019; Liyuan Liu & Yen Hsu, 2022; Ruiwei Li et al., 2022) Studies on various technologies, including electric vehicles and smart systems, further demonstrate that this ease of use positively affects users' behavioral intentions and overall satisfaction (Wu J et al., 2019; Sidra Ramzan et al., 2020; Ziqian Xia et al., 2021) Ultimately, user-friendly systems, such as smart recycling, enhance the likelihood of continued engagement (Zhang et al., 2019).
A study by Ramzan et al (2020) found that Perceived Ease of Use significantly influences the acceptance of online e-waste collection platforms, promoting environmental sustainability among Chinese youth This concept is crucial for waste classification behavior and the utilization of IoT bins (Ziqian Xia et al., 2021) Additionally, research on public acceptance of electric vehicles (Jingwen Wu et al., 2019) revealed a strong correlation between Perceived Ease of Use and the intention to adopt electric vehicles Ruiwei Li et al (2022) highlighted that Perceived Ease of Use had the most substantial effect on the intention to use shared e-bikes Furthermore, it fosters a positive user attitude towards services, thereby increasing usage intention (Fortest & Rita, 2016) Although Liu and Hsu (2022) noted that its impact on the intention to use smart recycling systems was not significant, Perceived Ease of Use still plays a role in influencing user behavior.
Therefore, this study proposes the following hypotheses:
Hl (+): Perceived Ease of Use will positively impact Perceived Usefulness in using mobile applications supporting waste classification at source using AI technology
H2 (+): Perceived Ease of Use will positively influence the Behavioral Intention to use mobile applications supporting waste classification at source using
Perceived Usefulness is defined as an individual's belief that utilizing a specific system will enhance their job performance (Davis, 1989) The implementation of smart recycling can significantly boost the efficiency of recycling processes and waste classification (Zhang et al., 2019) Extensive research has confirmed the positive influence of Perceived Usefulness on the intention to use a system, with its impact being affected by various external factors and structures (Wu et al., 2019) In simplified models that exclude preceding variables, Perceived Usefulness has been shown to greatly influence both the intention to use and sometimes attitudes towards the system (Chen).
Perceived Usefulness significantly influences the intention to use a model when accounting for other preceding variables, with external factors impacting this relationship (Cheng & Huang, 2013; Lu et al., 2016).
A study by Wu et al (2019) highlights that Chinese consumers' intention to use autonomous electric vehicles is positively influenced by their perceived usefulness, particularly regarding environmental benefits This suggests that individuals who appreciate these ecological advantages are more inclined to adopt or purchase such vehicles, reflecting a growing trend towards enhancing environmental quality Furthermore, the positive correlation between perceived usefulness and the intention to use has been validated in various technologies, including electric vehicles, bike-sharing systems, and smart recycling initiatives (Sidra Ramzan et al., 2020; Ruiwei Li et al., 2022; Liyuan Liu & Yen Hsu).
In 2022, research highlighted the significance of Perceived Usefulness in shaping positive attitudes and enhancing the intention to utilize online e-waste collection platforms (Sidra Ramzan, ChenGuang Liu, Yan Xu, Hina Munir, and Bhumika Gupta, 2020) Consequently, the team has put forth a hypothesis to explore this relationship further.
H3 (+): Perceived Usefulness will positively impact the Behavioral Intention to use mobile applications supporting waste classification at source using AI technology
2.4.I.2 The relationship between Perceived Playfulness and Behavioral Intention:
A motivating factor to promote interaction between humans and technology
Users frequently make decisions based on emotions rather than logic, which can significantly influence their acceptance of technology (Zhang & Li, 2005) To address this, Moon and Kim expanded the Technology Acceptance Model (TAM) to include technology service products (Moon & Kim, 2001) Further research has recognized Perceived Playfulness as both a personal trait and a psychological state, highlighting its importance in user engagement (Webster et al., 1993; Ahn et al., 2007).
Playfulness serves as a fundamental intrinsic motivation when individuals engage with new systems (Venkatesh & Bala, 2008) This multifaceted concept encompasses enjoyment, psychological arousal, and personal interest (Csikszentmihalyi, 2014) According to Moon and Kim (2001), Perceived Playfulness is characterized by three key aspects: an individual's focused attention on the interaction, a sense of curiosity about the experience, and genuine enjoyment derived from the interaction.
Smart recycling systems leverage technology to achieve recycling goals, and incorporating gamification techniques, such as rewards, enhances user enjoyment and encourages usage Consequently, prioritizing perceived playfulness is essential for researchers and designers of these systems This aligns with Bakker et al (2020), who noted that play is inherently satisfying and beneficial Therefore, it is crucial to explore how playful elements can boost the intention to use smart recycling systems Additionally, Koo et al (2015) found that some individuals participate in environmentally friendly behaviors primarily for enjoyment.
Previous research has consistently demonstrated that Perceived Playfulness significantly influences the Intention to Use new systems (Moon & Kim, 2001; Terzis et al., 2012) In particular, Padilla-Melendez et al (2013) confirmed a positive correlation between Perceived Playfulness and users' intention to engage with new technologies Furthermore, Liyuan Liu and Yen Hsu (2022) found that among various factors examined, perceived enjoyment exerted the most substantial effect on the public's intention to utilize smart recycling systems.
Perceived playfulness significantly enhances human-computer interaction, particularly in the context of mobile applications that utilize AI technology for waste classification at the source This study introduces the variable of perceived enjoyment, leading to the formulation of a new hypothesis.
H4 (+): Perceived Playfulness will positively impact the behavioral intention to use mobile applications supporting waste classification at source using AI technology
2.4.1.3 The relationship between Environmental Concern and Behavioral Intention: A significant predictive factor for environmental actions
Waste classification plays a crucial role in minimizing environmental pollution and enhancing recycling efforts (Luo et al., 2020) Understanding how organizational interactions shape human behavior can significantly aid managers in anticipating and influencing individual actions (Lin Shen et al., 2019) Research by Ellen et al highlights the importance of examining individual intentions within specific organizational contexts.
Research by Lin Shen et al (2019) indicates that individuals with heightened environmental concern are more inclined to participate in environmentally protective behaviors They expanded on the Theory of Planned Behavior (TPB) model, highlighting the significant influence of environmental concern on pro-environmental actions The authors define environmental concern as a comprehensive perspective on environmental issues, suggesting that when young people acknowledge their impact on the environment, they are more motivated to adopt eco-friendly practices.
A study by Liu et al (2021) highlights the positive correlation between perceptions of environmentally friendly behaviors and the intention to adopt eco-friendly electric vehicles in China Furthermore, Cui et al (2021) emphasized the role of Environmental Concern as a crucial factor influencing individuals' recycling activities, reinforcing its significance in promoting sustainable practices (Felix et al.).
2018) According to Meng et al (2019), environmental concern influences people’s actions related to waste classification and recycling Dinh Thai and Nguyen Van Thich
In 2022, research highlighted Environmental Concern as a significant factor influencing the Theory of Planned Behavior (TPB), demonstrating its strong impact on the attitudes of students in Ho Chi Minh City towards waste classification actions aimed at environmental protection The proposed model will establish a connection between Environmental Concern and the intention to utilize mobile applications that support source waste classification through AI technology, leading to the formulation of a new hypothesis.
H5 (+): Environmental Concern will positively impact the Behavioral intention to use mobile applications supporting waste classification at source using
2.4.1.4 From Intention to Behavior - Core relationships of the Theory of Planned Behavior model (TPB)
Research Procedure
(Source: Proposed by the group of authors)
The research process is essential for maintaining the quality and scientific rigor of a study, as it involves specific steps that guide the research in the right direction while enabling effective control and evaluation of results A well-defined process minimizes common errors and supports the advancement of future research The research team has established a structured approach for the topic, which is outlined in the following detailed steps.
To ensure meaningful research outcomes, it is essential to define specific research objectives The authors aim to: (1) identify the factors influencing the intention to use an application designed for waste classification at the source, (2) measure these factors' impact on the intention to utilize the application, and (3) offer conclusions and recommendations for management to enhance public awareness of waste classification and promote the use of technology in addressing everyday challenges.
The authors conducted a thorough search for secondary data sources published in both domestic and international journals and academic research forums They synthesized and analyzed research closely related to the topic to assess its applicability and relevant concepts Based on these findings, the research team developed a proposed research model.
Step 3: Develop A Proposed Research Model And Establish A Scale
The research team developed a preliminary scale table by carefully selecting appropriate scales based on theoretical foundations and original scales from previously published studies This selection process adhered to three key criteria: first, the compatibility of the original scale with the research topic; second, the scales were sourced exclusively from research articles published within the last five years; and third, priority was given to scales from reputable journals ranked Q1 according to Scopus.
The initial model and measurement scales developed in step 03 are merely preliminary To establish official and suitable measurement scales, the research team performed qualitative research, which included in-depth interviews with two experts in relevant fields and focused discussions with target groups The participants of the survey encompassed various stakeholders.
A survey involving six participants of varying ages will inform the adjustment of the scale based on expert input and target group feedback This ensures that any elements deemed inappropriate will be refined to align with the research context Ultimately, a finalized scale table will be established to facilitate the subsequent quantitative research phase.
The team designed the online survey in the form of Google Form and sent it to
A survey was conducted involving 310 consumers born between 1995 and 2005, utilizing non-probability and convenient sampling methods The data collection employed a 5-point Likert scale to assess various observed variables, providing essential information for subsequent analysis and evaluation.
Step 6: Analyze Data, Verify Research Results
The research team utilized SmartPLS software to screen, process, and analyze data, following a systematic approach that included Cronbach's Alpha reliability testing, Composite Reliability (CR), and Average Variance Extracted (AVE) verification They assessed Heterotrait-Monotrait Correlation (HTMT) coefficients and conducted non-parametric Bootstrap analysis The model evaluation focused on indirect effects, multicollinearity testing, and R² and Q² coefficient determination, culminating in a linear structural model analysis This comprehensive methodology enabled the team to critically examine the scales, hypotheses, and research model The findings are presented with an emphasis on their managerial implications, highlighting their relevance to sustainable consumer behavior in the fashion industry.
Following the quantitative research phase, the authors have reached several conclusions regarding the topic, highlighting both theoretical and practical implications The research team also proposes management implications, while acknowledging the study's limitations and suggesting directions for future research.
Qualitative Research
(Source: Proposed by the group of authors) Step 1: Build a draft scale for research concepts through reference research articles
Prior to initiating the research process, the team dedicated time to exploring and reviewing credible research articles relevant to the topic This thorough examination enabled the team to consult and refine scales that were suitable for the research subject, ultimately leading to the development of draft scales for the concepts involved.
Step 2: Develop a questionnaire to prepare for interviews with experts and feedback groups
The group utilized a draft scale to create a questionnaire aimed at an expert group and focus group, seeking to gather diverse perspectives on the topic to enhance and refine the scales.
Step 3: Conduct face-to-face or online interviews with experts and feedback groups to adjust the appropriate scale
To develop a comprehensive official scale table for our research, the team conducted in-depth interviews with two experts knowledgeable in technology and AI waste classification software Their valuable insights and practical feedback significantly contributed to the research process.
The group conducted interviews with six participants, comprising high school students, university students, and working individuals residing in Ho Chi Minh City This diverse selection of subjects offers accessible insights and an objective perspective on the topic being researched.
To conduct interviews, the group first selects an appropriate subject and then proactively reaches out to invite them for a scheduled interview The interview takes place only when respondents genuinely agree and are willing to participate Prior to the interview, the research team sends relevant documents and materials, such as research models and scale tables, to the subjects for their review.
Before the interview begins, the research team will introduce the topic and ensure that each respondent is treated with respect The interviewer will create a positive and comfortable environment, encouraging participants to openly share their thoughts and opinions.
The research team will present a detailed scale table to respondents, outlining each question, factor, and related concepts Throughout the interview process, the team is available to address any questions from respondents and encourages them to rate each factor as they see fit, without the need for modifications.
Respondents have the opportunity to contribute new insights and additional data for each observed variable to enhance the scale The research team remains committed to continuous improvement, actively listening and documenting ideas to develop innovative scales.
Step 4: Synthesize contributions and agree on an official scale to use for quan titative research
To construct the questionnaire, the research team will retain observed variables that garner strong agreement and positive evaluations Conversely, those variables that receive inadequate feedback or are poorly assessed will be evaluated for potential removal or replacement.
After discussing and agreeing on edits, the group will create the final scale for the next research process - qualitative research.
The research team gathered insights from expert interviews to refine and enhance the measurement scale The resulting scale, which received broad consensus among participants, was utilized to develop a clear and effective questionnaire for quantitative research.
The qualitative research findings will inform the development of a revised scale for quantitative research, incorporating necessary adjustments and enhancements The research team has updated the scale (see Appendix 2) and finalized the official version prior to conducting the quantitative survey (refer to Appendix 3).
After conducting a direct discussion, experts have the opinion that using the original scale is not really suitable for financial analysis because the scale referring to
Electronic waste does not align with the principles of source waste classification Experts recommend that the group reevaluate the original scale to ensure it is more realistic and relevant to the topic at hand.
The observed variables PU1, PU2, PU3, and PU4 will be transformed into a new scale that reflects the belief in the usefulness of AI waste classification for effective waste management Specifically, participants will express their agreement with statements such as: "I believe that the application of AI waste classification will be beneficial for waste management" (PU1), "I believe that using an AI waste classification application will reduce the manual processing workload" (PU2), and "I believe that using an AI waste classification application will enhance my motivation to classify waste" (PU3) This comprehensive adjustment ensures alignment with the current research context.
Most respondents expressed agreement with the questions in the scale, although the variable "PEU2" was perceived as challenging to comprehend Additionally, the term "interaction" was found to lack flexibility in its expression.
In this adjustment, the variable 'TEU2' was modified to reflect that using the AI waste classification application does not require extensive thought, while all other variables remained unchanged.
Quantitative Research
In this research, the team employed non-probability sampling methods, including judgment sampling, multistage sampling, seed development, and quota sampling Judgment sampling relies on the researcher's ability to evaluate subjects that align with the research topic Multistage sampling involves dividing the population into strata based on key characteristics, such as age and gender, and selecting samples from each stratum in a specific ratio The seed development method begins with initial survey points to inform the selection of subsequent points, particularly useful when the population is unknown but certain characteristics are available Lastly, quota sampling is conducted according to predetermined proportions for specific demographic groups, ensuring representation of the overall population.
To determine an appropriate sample size for exploratory factor analysis (EFA), Hair et al (1998) suggest that it should be at least five times the number of observed variables Similarly, Hoang Trong and Chu Nguyen Mong Ngoc (2008) recommend a minimum ratio of 4 to 5 times the number of observed variables Given that this study includes 22 observed variables, the research team followed the guidelines set by Tabachnick and Fidell (2007) to calculate the sample size, ensuring accuracy and consistency in their findings.
In this study, the minimum sample size was calculated to be 226, based on the formula 50 + 8p, where p represents the number of observed variables To meet this requirement, the research team distributed 320 survey forms, ultimately receiving 308 valid responses from participants residing in Ho Chi Minh City These responses were then used to conduct exploratory factor analysis (EFA).
For forms of research with limited time, resources and costs (Nguyen Dinh Tho,
In 2011, a combined judgment and convenience sampling method was deemed appropriate for research purposes The research team utilized Google Forms to create a questionnaire, which was distributed to respondents through social media platforms This approach enhances the representativeness and validity of the sample by targeting individuals and groups who possess knowledge about the research topic.
Judgment method: the group sent questionnaires to groups on social networking sites.
The seed development method involves the selection of key candidates, chosen based on the team's experience and established relationships These selected individuals then identify and invite additional candidates to participate, leveraging their own networks to ensure a comprehensive selection process.
The research team employed a norm method, stratifying survey subjects into four age groups representing Generations Y and Z, revealing notable differences in education levels, financial capabilities, technology access, and social issue perspectives among these groups Additionally, subjects were categorized by gender, with the General Statistics Office (2023) reporting a nearly even gender distribution in Vietnam's population: 49.61 million men (49.9%) and 49.85 million women (50.1%) The detailed demographic characteristics of the survey subjects are illustrated in the accompanying table.
Table 3.1 Distribution of demographic factors of survey subjects
Demographic characteristics Number of sample Ratio (%)
(Source: Proposed hy the authors)
The research team conducted a survey with a sample of 330 participants, utilizing three distinct methods to identify and segment the target group To ensure data quality, they filtered out responses that were incomplete or lacked answers, ultimately achieving a refined sample size of 300 valid responses.
The research team conducted a direct survey at prestigious higher education institutions in Ho Chi Minh City, including the University of Economics Ho Chi Minh City (UEH), Ho Chi Minh City University of Education (HCMUE), and the University of Social Sciences and Humanities (USSH) from January 24 to January 29, 2024 This approach aimed to gather diverse opinions from students across various fields of study, with a target sample size of 100 participants Data collection was facilitated through paper forms and QR codes linking to the survey.
The research team conducted a survey at Crescent Mall in District 7, Ho Chi Minh City, from 5 PM to 9 PM on January 27, 2024, a Saturday This timing was chosen due to the mall's status as one of the busiest commercial centers in the South, attracting a diverse range of shoppers who visit for leisure and shopping on weekends The team aimed to engage with interviewees of various age groups and educational backgrounds to gather comprehensive insights.
The study focused on individuals aged 28 to 43, primarily those who were employed, married, and financially independent The research team conducted interviews exclusively with this demographic, targeting a sample size of 70 participants To facilitate participation, subjects were provided with a paper application form or a QR code that linked directly to the interview form.
The third method employed for data collection is an indirect online survey using a questionnaire created with Google Forms The authors utilize the seed development technique, selecting participants from their network of relatives and friends They distribute the survey link through social media platforms such as Zalo and Facebook, encouraging respondents to complete the questionnaire and share the link with their own connections to enhance response rates The online data collection period is set from January 24 to January 29, 2024, with a target sample size of 150 respondents.
Based on the three sample collection methods above, the study ensures the similarity and diversity of the survey subjects.
3.3.3 Method to test the research model
Testing the research model is a crucial step in the research process, as it assesses the accuracy and reliability of the findings This process ensures that the model is correctly constructed, capable of predicting and explaining data, and robust enough to identify relationships between variables This study employs various methods to effectively test the research model.
3.3.3.1 Test the reliability of the scale
Internal consistency refers to the extent to which observed variables within a scale correlate and represent the same concept The level of internal consistency is commonly measured using Cronbach's alpha coefficient A high Cronbach's alpha indicates that the observed variables in a group are closely related and effectively measure the intended construct.
A strong correlation within a scale indicates high consistency, reflected by a Cronbach's Alpha coefficient ranging from 0 to 1 A value of 0 signifies no correlation among the observed variables, while a value of 1 indicates perfect correlation Higher Cronbach’s Alpha values enhance the reliability of the scale Nunnally (1978) asserts that a reliable scale should achieve a Cronbach’s Alpha of 0.7 or above, a view supported by Hair (2009), who emphasizes that scales must meet this threshold for both unidimensionality and reliability.
3.33.2 Test the scale of factors using composite reliability (CR) and convergent validity (AVE)
Cronbach’s Alpha is a conservative measure of reliability, often yielding lower reliability values, while composite reliability (CR) tends to overestimate internal consistency, resulting in higher reliability estimates Therefore, it is essential to report both metrics when assessing the internal reliability of scales, as the true reliability typically falls between the lower limit of Cronbach’s Alpha and the upper limit of CR Additionally, convergent validity, indicated by the average variance extracted (AVE), assesses the strength of positive correlations among observed variables of a latent variable Researchers should evaluate convergent validity by examining the external loadings of these observed variables alongside the AVE.
DATA ANALYSIS AND RESULTS
Descriptive Statistics
After screening 332 collected survey responses for relevance, 319 valid responses were retained Data cleaning was conducted to eliminate missing answers and patterned responses, resulting in 308 samples eligible for analysis Detailed characteristics of the survey subjects are presented in Table 4.1.
Table 4.1 Descriptive Table of the Study Sample
Factor Valid Frequency Valid Percent (%)
High school students 35 11,36 un i vers ity/co 11ege/ vocational students 180 58,44
Al-based waste sorting apps?
(Source: Data processed by the authors)
Chart 4.1 Pie Chart Showing the Gender Distribution of the Study Sample
(Source: Data processed by the authors)
A recent survey of 308 respondents revealed a nearly equal gender distribution, with 153 males (50.32%) and 155 females (49.68%) This balanced ratio underscores the relevance of the findings, highlighting that both men and women equally recognize the need and intention to utilize AI-based waste sorting applications.
Chart 4.2 Pie Charl Showing the Age range Distribution of the Study Sample
(Source: Data processed by the authors)
In terms of age, the research group divided the respondents into four groups corresponding to the ages of two generations: Gen Z/Zoomcrs (1997-2012) aged 12 to
27 and Gen Y/Millennials (1981-1996) aged 28 to 43.
A recent survey revealed that out of 308 respondents, 212 identified as Gen Z/Zoomers, making up 68.83% of the total Within this group, 34 respondents were aged 12-17, representing 11.04% of the total, while 178 were aged 18-27, accounting for 57.79% Additionally, 96 respondents were from Gen Y/Millennials, constituting 31.17% of the overall sample.
Of these, 65 were aged 28-35 (21.1%) and 31 were aged 36-43 (10.06%).
The study predominantly involved young participants from Generation Z and a segment of Generation Y, making it ideal for exploring new technology products like AI-based waste sorting applications.
Uni versify/college'vocational students
Chart 4.3 Pie Chart Showing the Occupation Distribution of the Study Sample
(Source: Data processed by the authors)
According to the survey data, out of 308 respondents, 35 were high school students (11.3 6%), 180 were university/college/vocational students (5 8.44%), and 93 were employed (30.19%).
The survey predominantly included university, college, and vocational students, highlighting a significant age distribution among respondents Most participants aged 12-17 are high school students, while those between 18-27 are enrolled in higher education Additionally, respondents aged 28-43 are typically employed and married.
The relationship between age and occupation enhances the practicality, diversity, and objectivity of survey findings Both age and occupational demographics show a significant interest in utilizing AI-based waste sorting applications.
Chart 4.4 Pie Chart Showing the Waste Sorting Implementation Level of the Study
(Source: Data processed by rhe authors)
The survey results indicate a generally positive attitude towards waste sorting among respondents, with 47.4% reporting that they sort waste "Occasionally." This highlights a significant level of engagement in waste management practices among the participants.
"Regularly" and "Always" is also relatively positive, at 27.92% and 5.84%, respectively Besides, a small number of respondents said they "Rarely" or even
"Never" sort waste, accounting for 15.26% and 3.57%, respectively.
Most respondents demonstrate a strong awareness of waste sorting, which encourages them to explore and utilize applications designed for convenient waste management, ultimately enhancing the reliability of their responses.
4.1.5 Knowledge of AT-based Waste Sorting Apps
Chart 4.5 Pie Chart Showing the Knowledge of Al-based Waste Sorting Apps of the
(Source: Data processed by the authors)
A significant 214 respondents, representing nearly 70% of the total, reported being aware of or having learned to use AI-based waste sorting apps This high level of awareness indicates a positive trend in public engagement with these technologies, enhancing the reliability of the survey results.
Evaluate the scale
Scale reliability assesses the reliability value of each component within a scale, focusing on the indicator values of the scale's components that represent various aspects of the measured factor To ensure accuracy, scale reliability must adhere to established standard values.
The Cronbach's Alpha test result for the factors that must meet the requirements is greater than 0.6.
The total correlation coefficient of observed variables is greater than 0.3.
The Cronbach’s Alpha value if each observed variable of the above factors is eliminated is smaller than the Cronbach's Alpha value of the scale.
Scale standards that achieve reliable value will be included in exploratory factor analysis (EFA) to measure convergent validity The testing results are described in the table below:
Table 4.2 Scale Reliability Test by Cronbach’ Alpha
Scale Mean if Item Deleted
Scale Variance if Item Deleted
Cronbach’s Alpha if Item Deleted Perceived Usefulness (PU): Alpha = 0,847
Perceived Ease of use (PEU): Alpha = 0,808
(Source: Data analysis from SmartPLS4)
The analysis indicates that all variables exhibit Cronbach's Alpha values between 0.6 and 0.9, confirming the scale's reliability Additionally, the total correlation coefficients for the measurement components exceed the acceptable threshold of 0.3, and the coefficients for Cronbach's Alpha if Item Deleted are consistently lower than the Cronbach Alpha values for each observed variable.
Thus, the observed variables all meet the requirements and the scale is reliable and these variables are used in the next EFA analysis.
4.2.2 Test the scale of factors by Composite Reliability (CR) and Convergent Validity (AVE)
To evaluate the reliability of the scale, it is essential to examine two key metrics: the Composite Reliability Coefficient (CR), which should be at least 0.70, and the Average Variance Extracted (AVE), which must reach a minimum threshold of 50% (Hair et al.).
2009) Test results showed that CR ranged from 0.846 to 0.923, all meeting the requirements > 0.70, and AVE ranged from 57.8% to 79.9%, meeting the requirements
> 50% This shows that the scale ensures reliability.
Table 4.3 AVE and CR lest results
(Source: Data processing results from SmartPLS4)
Convergent validity testing
According to Hair et al (2014), for effective measurement in statistical analysis, the outer loading coefficient of variables should be 0.70 or higher, indicating a significant relationship with variance An outer loading coefficient above 0.70, with its square being 0.5, is ideal, while values between 0.4 and 0.7 may be acceptable but warrant careful consideration before exclusion (Henseler et al., 2009) The assessment of convergent validity hinges on both the outer factor loading index and the Average Variance Extracted (AVE).
After checking the outer loading indices, no observed factors were removed, so all observed variables met the requirements.
BE EC IN PEU pp PU
(Source: Data processing results from SmartPLS4)
Discriminant validity
Discriminant validity assesses whether a concept is genuinely distinct from other research concepts based on empirical evidence By calculating the discriminant value, researchers can demonstrate that a specific concept is unique and represents a different phenomenon within the model compared to other concepts.
The Heterotrait-Monotrait (HTMT) conceptual coefficient approach predicts the true correlation between two concepts if measured perfectly, indicating strong correlations or disattenuated correlations A high correlation, close to 1, suggests a lack of discriminant validity, with an HTMT coefficient exceeding 0.9 indicating insufficient distinction between the concepts For concepts generally perceived as distinct, a lower acceptance threshold of around 0.85 is acceptable Additionally, the statistical confidence interval of the HTMT should not include the value 1 for all concept combinations to ensure valid discriminant validity.
Table 4.5 Heterotrait-Monotrait (HTMT) test results
BE EC IN PEU pp PU
(Source: Data processing results from SmartPLS4)
The study's findings indicate that all Heterotrait-Monotrait (HTMT) correlation coefficients are below 0.9, with the highest coefficient being 0.574 This suggests that the concepts within the research models are distinct and do not overlap in meaning.
Hypothesis testing
The analysis of the internal structural model follows the assessment of the external model, focusing on the relationships between latent variables Path coefficient values for endogenous latent variables and R-squared analysis were utilized, with a significance level set at 5% (p < 0.05) Research results were evaluated using non-parametric Bootstrap analysis, employing a repeated sampling approach (Bootstrapping) with 5,000 samples to ensure robust findings.
Table 4.6 Results of testing the hypotheses Original sample (O)
(Source: Data processing results from SmartPLS4)
The analysis presented in Table 4.6 indicates that all effects are significant, with p-values below 0.05 The standardized regression coefficients reveal the impact of various factors on the intention to use a mobile application for waste classification at the source The most influential factor is Perceived Usefulness (PU), with a coefficient of 0.308, followed by Environmental Concerns (EC) at 0.256, and Perceived Playfulness (PP) at 0.193 Perceived Ease of Use (PEU) has the lowest impact on intention, with a coefficient of 0.159 Additionally, Perceived Ease of Use significantly influences Perceived Usefulness with a coefficient of 0.481, while Intention to Use (IN) affects Behavior of Using Applications (BE) at 0.485 Consequently, the hypotheses are confirmed.
Hl (+): Perceived Ease of Use has a positive influence on Perceived Usefulness when using mobile applications that support waste classification at source using artificial intelligence (AI).
H2 (+): Perceived Ease of Use has a positive influence on Behavioral Intention to use mobile applications that support waste classification at source using artificial intelligence (Al).
H3 (+): Perceived Usefulness has a positive impact on Behavioral Intention to use mobile applications that support waste classification at source using artificial intelligence (AI).
H4 (+): Perceived Playfulness has a positive impact on Behavioral Intention to use mobile applications that support waste classification at source using artificial intelligence (AI).
H5 (+): Environmental Concern has a positive impact on Behavioral Intention to use mobile applications that support waste classification at source using artificial intelligence (AI).
H6 (+): Intention to use mobile applications that support waste classification using AI has a positive impact on people’s Behavior in using these applications in waste classification.
Evaluate the model
BE EC IN PEU pp PU
(Source: Data processing results from SmartPLS4)
The model does not have multicollinearity when the VIF index is less than 5 (Hair et al., 2017).
4.6.2 Evaluate the coefficient of determination R2
The R² value ranges from 0 to 1, indicating the accuracy of a model's predictions, with higher values signifying greater precision Establishing an acceptable R² value can be challenging, as it is influenced by the model's complexity and the specific research context R² results are utilized to assess the extent to which structural models explain the variance of endogenous variables.
Table 4.8 Evaluation table of determination coefficient R2
(Source: Data processing results from SmartPLS4)
The coefficient of determination (R²) for the dependent variable BE is 0.236, indicating that it accounts for 23.6% of the variation in this variable In contrast, the R² for the dependent variable IN is 0.32, which explains 32% of its variation.
4.6.3 Evaluate the impact of f-square coefficient
Table 4.9 Table for evaluating f-square coefficient
BE EC IN PEU PP PƯ
(Source: Data processing results from SmartPLS 4)
In addition to assessing the R2 value of dependent variables, researchers also analyze the change in R2 when an independent variable is excluded from the model to determine its impact on the dependent variable According to Cohen (1988), f-square values of 0.02, 0.15, and 0.35 indicate small, medium, and significant effects, respectively An f-square value below 0.02 suggests that the independent variable does not influence the dependent variable The analysis confirms that all f-square coefficients meet the necessary criteria.
Structural model
The comprehensive model comprises six interconnected factors: Perceived Usefulness, Perceived Ease of Use, Perceived Playfulness, Environmental Concern, Behavioral Intention, and Behavior These factors interact and influence one another, providing a foundation for developing effective strategies to enhance the adoption of mobile applications that utilize artificial intelligence (AI) for waste classification at the source.
Figure 4.1 Post-test research model
(Source: The research team synthesized from SmartPLS4)
Discussion of Research Results
Table 4.10 Summary of Hypotheses Testing
Hypothesis Content Sig Test Outcome
Perceived Ease of Use positively influences Perceived Usefulness in using mobile applications supporting waste classification at source using AI technology.
(Source: Data analysis synthesized using SmartPLS4)
Perceived Ease of Use positively impacts the Behavioral Intention to use mobile applications supporting waste classification at source using Al technology.
Perceived Usefulness positively influences the Behavioral Intention to use mobile applications supporting waste classification at source using AI technology.
Perceived Playfulness positively impacts the Behavioral Intention to use mobile applications supporting waste classification at source using AI technology.
Environmental Concern positively influences the Behavioral Intention to use mobile applications supporting waste classification at source using AI technology.
H6 (+) rhe Behavioral Intention to use mobile applications supporting waste classification at source using Al technology positively impacts the Actual Usage Behavior of these applications.
The analysis confirms that the theoretical model is consistent with the research data, leading to the acceptance of six research hypotheses (H1, H2, H3, H4, H5, and H6) This indicates that the identified factors significantly influence the behavioral intention to use mobile applications for waste classification at source, utilizing AI technology in Ho Chi Minh City Among the four key factors affecting this intention, the impact is ranked as follows: (1) Perceived Usefulness, (2) Environmental Concerns, (3) Perceived Playfulness, and (4) Perceived Ease of Use.
The study reveals that "Perceived Usefulness" is a key factor influencing the Behavioral Intention to use mobile applications for waste classification at the source, particularly in Ho Chi Minh City With a significance value below 0.05 and a standardized Beta coefficient of 0.308, there is a clear positive correlation between Perceived Usefulness and the intention to use these applications This indicates that for every one-unit increase in Perceived Usefulness, there is a corresponding 0.308-unit increase in the intention to use the application, establishing it as the most significant influencing factor.
Environmental Concern plays a crucial role in influencing the Behavioral Intention to use AI-powered mobile applications for waste classification at the source in Ho Chi Minh City The findings indicate a significant relationship, with a standardized Beta coefficient of 0.256, suggesting that a one-unit increase in Environmental Concern results in a 0.256-unit increase in the intention to use the application This positions Environmental Concern as the second strongest factor affecting user behavior.
In Ho Chi Minh City, "Perceived Playfulness" plays a crucial role in influencing the Behavioral Intention to use mobile applications designed for waste classification with AI technology The hypothesis supporting this relationship was validated, as indicated by a significance value below 0.05 and a standardized Beta coefficient of 0.193 This suggests a positive correlation, where an increase of one unit in Perceived Enjoyment leads to a 0.193-unit rise in the intention to use the application, positioning it as the third most significant factor in this context.
The "Perceived Ease of Use" significantly influences the Behavioral Intention to utilize mobile applications for waste classification using AI technology in Ho Chi Minh City This relationship is supported by a Sig value of less than 0.05 and a standardized Beta coefficient of 0.159, indicating that an increase in Perceived Ease of Use correlates with a 0.159-unit rise in the intention to use the application Despite its positive impact, it is identified as the weakest influencing factor in this context.
Additionally, with a p-value less than 0.05:
The factor impacting "Perceived Usefulness" is "Perceived Ease of Use", with the strongest influence having an impact coefficient of 0.481, and the factor impacting
"Actual Usage Behavior" is "Behavioral Intention to Use" with an impact coefficient of 0.485.
These research findings are essential for administrators looking to enhance public awareness of waste segregation at the source and for the advancement of management organizations They also provide valuable reference materials for teaching and research in Development Economics and Environmental Management, as well as for exploring solutions for circular economic development in Ho Chi Minh City.
New contributions of research
This research article significantly advances the field of life technology by integrating multiple factors influencing the intention to use waste classification applications, utilizing the Theory of Planned Behavior (TPB) and Technology Acceptance Model (TAM) Unlike previous studies, such as those by Sha Lou (2023) and Shen Lin (2019), which focused solely on technology-related aspects, this study expands the scope by incorporating Environmental Concerns, thereby highlighting the impact of environmental awareness on users' attitudes towards waste classification This comprehensive approach sets it apart from earlier works by Ruiwei Li et al (2022) and F Behnamifard (2021), which primarily examined user and technology factors.
This research enhances understanding of the interaction between people and technology, specifically in the context of waste classification, by applying the Technology Acceptance Model (TAM) and the Theory of Planned Behavior (TPB) It serves as a valuable reference for future studies on the factors influencing the intention to use mobile applications that facilitate waste classification through AI technology Additionally, the findings can inform educational programs and research in Development Economics, Environmental Management, and strategies for promoting circular economic development in Ho Chi Minh City.
This study's findings have significant implications for research on technology addressing social and environmental issues, aimed at achieving sustainable development Additionally, it enriches the existing knowledge on utilizing technology to tackle environmental challenges, serves as a valuable reference for administrators, and offers new developmental pathways for organizations, managers, researchers, and application developers.
In Chapter 4, "Research Results," the team details the characteristics of the research sample and evaluates the measurement scale through Cronbach's Alpha reliability analysis They also assess scale factors using composite reliability (CR) and convergent validity (AVE) Following this, the survey data is analyzed to test the hypotheses and emphasize the research's new contributions.