Content area
Purpose
Today, e-government (electronic government) applications have extended to the frontiers of health-care delivery. E-Nabız contains personal health records of health services received, whether public or private. The use of the application by patients and physicians has provided efficiency and cost advantages. The success of e-Nabız depends on the level of technology acceptance of health-care service providers and recipients. While there is a large research literature on the technology acceptance of service recipients in health-care services, there is a limited number of studies on physicians providing services. This study aims to determine the level of influence of trust and privacy variables in addition to performance expectancy, effort expectancy, social influence and facilitating factors in the unified theory of acceptance and use of technology (UTAUT) model on the intention and behavior of using e-Nabız application.
Design/methodology/approach
The population of the study consisted of general practitioners and specialist physicians actively working in any health facility in Turkey. Data were collected cross-sectionally from 236 physicians on a voluntary basis through a questionnaire. The response rate of data collection was calculated as 47.20%. Data were collected cross-sectionally from 236 physicians through a questionnaire. Descriptive statistics, correlation analysis and structural equation modeling were used to analyze the data.
Findings
The study found that performance expectancy, effort expectancy, trust and perceived privacy had a significant effect on physicians’ behavioral intentions to adopt the e-Nabız system. In addition, facilitating conditions and behavioral intention were determinants of usage behavior (p < 0.05). However, no significant relationship was found between social influence and behavioral intention (p > 0.05).
Originality/value
This study confirms that the UTAUT model provides an appropriate framework for predicting factors influencing physicians’ behaviors and intention to use e-Nabız. In addition, the empirical findings show that trust and perceived privacy, which are additionally considered in the model, are also influential.
1. Introduction
To facilitate the successful adoption of technologies, it is important for all stakeholders in the health sector to be informed about the various factors that influence acceptance (Almarzouqi et al., 2022; AlQudah et al., 2021). The use of information technologies in health-care reduces costs, improves quality and efficiency, and allows for more health-care services in a limited time (Cheng et al., 2022; Hsu et al., 2013; Voeffray et al., 2006). However, investments in such technologies are quite expensive and carry the risk of not meeting expectations. In particular, users’ unwillingness to use the applications constitutes the most important reason for not achieving the goals (Alaiad and Zhou, 2014).
In the studies on technology acceptance models in health services delivery, studies aiming to determine the opinions of the beneficiaries have found a place in the field (Dwivedi et al., 2016; Kijsanayotin et al., 2009; Ong et al., 2022; Quaosar et al., 2018). The number of studies examining the factors that facilitate the acceptance of technologies open to the use of patients and their relatives from the perspective of physicians has been very limited compared to the volume of literature (Aljarboa and Miah, 2021; Dabliz et al., 2021; Falana et al., 2023; Proulx et al., 2021). A significant portion of the few studies aiming to reveal physicians’ perspectives have also focused on physician trainees (Bazelais et al., 2022; Guo, 2022; Wang et al., 2022). However, determining physicians’ attitudes toward technologies is important for achieving the targeted success. For example, many remote care technologies have not been integrated into clinical routines in the health-care sector because they are not accepted by health-care professionals (Rouidi et al., 2022). Therefore, it is necessary to explain in detail the behavioral intention of physicians regarding information technologies and the factors affecting their usage behavior for the success and sustainability of the technology.
This research extends the unified theory of acceptance and use of technology (UTAUT) model to determine the acceptance of a specific health information system, e-Nabiz, by physicians. It provides a unique contribution to the studies on health information systems in Turkey, a developing country application. Furthermore, this study explores the role of additional variables, such as trust and privacy, in the acceptance of health information systems. This adds significant depth to the existing literature and can guide broader research on the acceptance of health information systems. The examination of active physicians’ experiences and perspectives within the UTAUT framework offers a distinct contribution to the field. Considering the limited studies targeting this specific audience using the UTAUT model in the literature, this research deepens our understanding of the acceptance of health information systems and presents new perspectives. Particularly in the context of a developing country like Turkey, it provides valuable insights into how a health information system like e-Nabız is perceived and accepted by physicians. In this regard, the study adds an important and innovative perspective to the current literature.
Many studies have been conducted to explain the variables affecting information technologies in health services. Studies have been conducted on remote health services (telemedicine, online self-management, teledentistry, mobile health, telepsychology, telehomecare, telehealth and eHealth) (Rouidi et al., 2022) and electronic record systems health information technologies (Kijsanayotin et al., 2009; Yıldız and Dinçer, 2021), electronic prescription (Falana et al., 2023), medication management systems (Dabliz et al., 2021), artificial intelligence assisted diagnosis and treatment systems (Cheng et al., 2022), decision support systems and nursing information systems (AlQudah et al., 2021). There is a limited number of studies on e-Nabız, which is a personal electronic health information system (Birinci, 2023).
In the health-care sector, the maintenance of various electronic records is a key aspect. Prominent among these are electronic health records (EHR), electronic medical records (EMR) and personal health records (PHR), along with others such as records pertaining to genomic care, oncology patients and pharmaceutical networks (Kara and Kurutkan, 2021). EHRs, EMRs and PHRs are the most commonly researched types of records (Symons et al., 2019). The influence of EHRs and e-health technologies on health-care services in terms of quality, efficiency and cost is extensively analyzed. The concept of “Meaningful Use” plays a vital role in maximizing the advantages and efficiency of these technologies (Blumenthal and Tavenner, 2010). The beneficial impact of these technologies on health-care quality and efficiency is supported by systematic studies (Chaudhry et al., 2006; Black et al., 2011). Factors like time efficiency, enhanced quality and cost reduction indicate that adopting EHRs and e-health technologies broadly could boost health-care services’ overall effectiveness (Jha et al., 2009; Hillestad et al., 2005; Poissant et al., 2005).
EHRs, as institutional systems, EHRs electronically store medical data related to patients and facilitate the sharing of this information among various health-care providers. EHRs encompass diverse data categories, including allergies, vital signs, appointments, lab results, medical imaging and diagnoses (Heart et al., 2017; Roehrs et al., 2017). EMRs, these are systems within health-care institutions for storing electronic records about patient health. EMRs generally include information such as patient demographics, medication, vital statistics, medical history, lab findings, vaccinations and radiology reports and are specific to the health-care organization’s internal medical field (Heart et al., 2017; Häyrinen et al., 2008). PHRs, defined as online systems that use health informatics standards to manage and store patient health care and medical data collections, PHRs and personal health information management (PHIM) enable individuals to oversee their health information. The use and effectiveness of PHRs and PHIM can vary based on demographic and health conditions (Kim et al., 2018; Taylor et al., 2018). These applications are instrumental in enhancing communication between patients and health-care providers (Detmer et al., 2008; Ancker et al., 2015). Their design and application should be tailored to meet diverse user needs and demographic features (Agarwal and Khuntia, 2009).
E-Nabız, a digital health record system in Turkey, exhibits characteristics that align with EMRs, EHRs and PHRs. In Turkey, with a population of approximately 85 million, E-Nabız collects health information from all citizens at various health facilities, similar to EMRs. Additionally, it allows health-care providers and facilities across the country to access patient information with the patient’s consent, akin to EHRs. E-Nabiz enables certain restrictions on films and analyses within the framework of evidence-based practices defined by health authorities, thus preventing unnecessary repeated tests. Since previous medical information of patients is accessible through the system resources can be used more efficiently, and patient safety is enhanced by avoiding unnecessary radiation and repeated imaging. Moreover, the system allows individuals to compile their own specific information, enhancing the completeness of the records and personalizing them in line with the concept of PHRs.
Customization is especially important for specific groups, such as those with chronic illnesses or the elderly (Ancker et al., 2015; Turner et al., 2021). The role of family and friends in managing PHRs should also be considered, as they significantly influence the social dynamics involved (Taylor et al., 2018). Factors like health literacy can affect the effective management of PHRs, necessitating enhancements in system accessibility and understandability (Kim et al., 2018). The successful implementation of PHR and PHIM involves considering a range of factors, including demographic variations, individual preferences and social dynamics. This approach promotes more effective communication between patients and healthcare providers, thereby improving the quality of health-care services overall.
E-Nabız is a nationwide electronic PHR system in which the Ministry of Health integrates health data collected electronically from all health institutions and which can be accessed by individuals and health professionals via personal computers, tablets and mobile devices. Health service users can access laboratory results, medical images, prescription and medication information, emergency contact information, diagnosis information, reports and health records containing all details of their examinations. E-Nabız is also one of the e-government applications (Birinci, 2023; Republic of Türkiye Ministry of Health, 2023b). Physicians access patient data in the e-Nabız system through the “E-nabız Doctor Access Service” via the Hospital Information Management System and Family Medicine Information System (Republic of Türkiye Ministry of Health, 2023a). The number of users of the application is approximately 57 million as of December 12, 2021. The eGovernment Benchmark 2022 Background Report of the European Union is among the examples of good practices (The European Commission, 2022).
E-Nabız’s data-sharing feature aims to improve the quality and speed of the diagnosis and treatment process, eliminating the duplication of procedures such as medical imaging, thereby saving health-care expenditures and reducing working hours for health-care staff. For example, the e-Nabız system eliminated the cost of printing around 47.3 million computed tomography images (for 2020 and 2021), saving around US$11.8m, half of the investment cost of e-Nabız (The European Commission, 2022). Due to the expected benefits of the e-Nabız system, it is expected to be used by governments.
Although the use of the E-Nabız system started in 2015, the system is constantly being updated; new features are being added according to the current and proactive needs of individuals and institutions and work on transition and change management continues (The European Commission, 2022). A user’s acceptance of technology remains important at any time, not just during the design phase or immediately after implementation (AlQudah et al., 2021). For example, during the COVID-19 pandemic another innovation was added to e-Nabız, and doctors serving in hospitals affiliated to the Ministry of Health started to be able to make video calls (Tele Health) through the e-Nabız system (Republic of Türkiye Ministry of Health, 2021).
The main question of this research is:
Thus, this research will contribute to the literature to determine the behavioral intention among physicians regarding e-Nabız and to reveal the factors affecting usage behavior. For this purpose, this research aims to validate the integrated technology acceptance and utilization model (UTAUT), which has previously been used to predict the adoption of different information systems for health for e-Nabız. It also extends the UTAUT model by integrating privacy and trustworthiness constructs. On the other hand, little is known about the determinants of the adoption of e-health innovations by physicians (Dünnebeil et al., 2012). This research provides empirical evidence on the determinants of physicians’ acceptance of e-Nabız as a health information system. Finally, it provides important information for application developers, government agencies and administrators to further increase the intention and behavior of e-Nabız use by physicians.
The second part of the study discusses the research model and hypothesis development. The methodology is then analysed in the third section, the findings in the fourth section, the discussion and conclusions in the fifth section, and the limitations and future research topics in the sixth section.
2. Research model and development of hypotheses
Different technology acceptance models and theories have been used to explain the use of health technologies. These theories and models aim to better understand users’ behavior toward a particular technology or service and the determinants of this behavior (Al-Maroof et al., 2022). Models and theories enable users’ awareness of their readiness to use health technologies by revealing the technical, social and cultural factors related to the use of technology. In this way, it is stated that the effectiveness of technologies will be increased (AlQudah et al., 2021).
Among many theories and models, UTAUT offers a new and integrative research theoretical perspective and is widely used, valid and reliable (Alaiad and Zhou, 2014; Rouidi et al., 2022; Venkatesh et al., 2023). UTAUT is recognized as one of the most common models to explain what influences the acceptance of various health technologies across different user groups, settings and countries (AlQudah et al., 2021). Therefore, the UTAUT model was adopted as the basis for this research.
Technology acceptance is defined as an individual’s psychological state regarding the voluntary or purposeful use of a particular technology (AlQudah et al., 2021; Alaiad and Zhou, 2014). It helps to understand the drivers of acceptance to proactively identify interventions (including education and marketing) targeting user populations that may be less inclined to adopt and use technologies (Venkatesh et al., 2003). Venkatesh et al. (2003) proposed the UTAUT model based on the principles of the reasoned action model, technology acceptance model, motivation model, theory of planned behavior, model combining technology acceptance model and theory of planned behavior, personal computer use model, diffusion theory of innovation and social cognitive theory. UTAUT is used to explain the acceptance of various technologies by professionals in health care (Arfi et al., 2021).
In the UTAUT model, behavioral intention is determined by performance expectancy, effort expectancy and social influence, while usage behavior is directly determined by facilitating conditions and intention to use. Gender, age, experience and volunteerism are considered as mediating variables in the model (Venkatesh et al., 2003). In this study, the main variables were addressed within the scope of the model and mediating variables were not included. In addition, within the scope of the research, not only the UTUAT model was used, but also security and privacy, which are two important variables based on the literature, were included in the model.
Performance expectation
Performance expectancy is defined as the degree to which an individual believes that using the information system will help him/her achieve gains in job performance (Venkatesh et al., 2003). Many studies have demonstrated the impact of performance expectancy on usage intention (Alaiad and Zhou, 2014; Ong et al., 2022; Hoque and Sorwar, 2017; Hoque et al., 2016). In fact, performance expectancy is the most important determinant of behavioral intention (Raman and Don, 2013). According to the results of a study conducted on the use of EMR in a tertiary hospital in Indonesia, it was revealed that performance expectancy explained a significant portion of behavioral intention (Faida et al., 2022). In a study conducted to measure e-health adoption behavior in developing countries, performance expectancy and behavioral intention were found to be positively related (Sahoo et al., 2023). Physicians’ expectation that the e-Nabız system will help access patient information quickly and efficiently and improve the quality of treatment and diagnosis will influence the intention to use the application. Therefore, we propose the following hypothesis:
Effort expectancy
Effort expectancy is defined as the degree of ease associated with the use of the system (Venkatesh et al., 2023). If it is felt that the information system is easily grasped and used, the willingness to adapt will increase. Previous studies have also shown that effort expectancy is a determinant of behavioral intention (Arfi et al., 2021; Hoque and Sorwar, 2017; Kijsanayotin et al., 2009). In addition to performance expectancy, studies also find a significant relationship between effort expectancy and behavioral intention to use e-health services (Mensah et al., 2022; Sahoo et al., 2023). It is expected that physicians’ finding the application easy, effortless and easy to learn will increase their intention to use it. Therefore, the following hypothesis is proposed:
Social impact
Social influence is defined as the degree to which an individual perceives that people they care about believe that they should use the new system (Venkatesh et al., 2003). Employees socially model the behavior accepted by others who are important to them (Alaiad and Zhou, 2014). Social influence plays an important role in shaping behavioral use in various contexts. Many studies have examined the impact of social influence on behavioral intention and usage behavior using different theoretical frameworks and methodologies (Salathé et al., 2013) and pointed out that social influence is one of the determinants of usage intention (Arfi et al., 2021; Chang et al., 2007; Hossain et al., 2019). In a study conducted to predict and explain the adoption of EMR system in the health system, it was determined that social influence was effective on behavior (Almarzouqi et al., 2022). Considering that the people who are important to them may be influential in physicians’ use of e-Nabız, the following hypothesis was put forward:
Facilitating influence
Facilitating conditions are defined as the degree to which an individual believes that an organizational and technical infrastructure exists to support the use of the information system (Venkatesh et al., 2003). If the necessary tools and organizational enablers for the technology are in place, the level of acceptance will be high (Rouidi et al., 2022). Infrastructure support has an important role in facilitating the use of health information systems (Hossain et al., 2019). In previous studies on physicians’ acceptance of technology, it was found that the facilitating condition contributed to the usage behavior (Chang et al., 2007). It is expected that resources related to e-Nabız, physicians’ knowledge of how to use the technology and the presence of technical support will increase implementation behavior. Therefore, the following hypothesis is proposed:
Behavioral intention
Another relationship in the UTUAT model is that behavioral intention positively influences usage behavior. Intention to use is known to play an important role in influencing the actual use and implementation of new technologies. Behavioral intention refers to the extent to which the technology is intended to be used (Venkatesh et al., 2003). An attitude toward the use of a system is considered as an attitude toward the behavior of that system (Hartwick and Barki, 1994). It has been determined in previous research that intention to use plays an important role in influencing the actual use and implementation of new technologies (Dash and Sahoo, 2022; Hossain et al., 2019; Venkatesh et al., 2023). The usage behavior of the e-Nabız system is voluntary for both physicians and patients, so physicians’ intention to use the system is important for the emergence of behavior. Physicians’ intention to use the application is thought to be determinant in terms of usage behavior, and this relationship is defined within the scope of the UTAUT model. For this, the following hypothesis is proposed:
In this study, the UTAUT model was taken as the basis for determining the variables of the research, but the variables of trust and privacy, which have an important place in the field of health, were also included in the model. Doctors need access to patient information for accurate diagnosis and treatment. However, privacy and trust, which form the basis of the patient-doctor relationship, also have an important place in information sharing and their importance is increasing with the developing information technologies (Gu et al., 2021). Users’ intention and attitude toward privacy can affect the adoption of health information systems. Another fact is that security issues are a deep-rooted concern for many technology users. Information security and privacy concerns are consistently cited as one of the main challenges to the success of health information systems (Falana et al., 2023; Hsu et al., 2013). Criticism of health information systems, especially in the health field, is centered on the fact that they harm privacy and do not ensure the security of information resources (Belanger and Crossler, 2011). Health care is a sensitive area regarding the privacy of personal information and the reliability of systems. For this reason, researchers also consider reliability (Alaiad and Zhou, 2014; Quaosar et al., 2018) and privacy (Chopdar, 2022) among the factors affecting the acceptance and intention to use information systems in health care. Therefore, there is a need to further examine the importance of these two variables in physicians’ intention to use health information systems.
Perceived privacy
The use of information technologies in health services has some disadvantages as well as advantages. Perhaps the biggest of these disadvantages is the inability to ensure privacy on behalf of patients (Belanger and Crossler, 2011). Privacy is basically defined into four types. These are information privacy, bodily privacy, communication privacy and personal space (Alaiad and Zhou, 2014). Communication privacy and information privacy come to the fore in the use of information technologies in health services. Health-related information of individuals is highly sensitive. Moreover, the potential misuse of personal health information raises privacy concerns. When users perceive a greater privacy risk/loss compared to the benefits, they are less likely to use information systems (Chopdar, 2022). Privacy in IT refers to application features that restrict the possibility of data or network breach by malicious sources, unwanted data deletion, disclosure or modification, denial of service and/or economic loss through fraud with the user of the application (Yousaf et al., 2021).
Research shows that privacy is effective in the acceptance of health information technologies (Gu et al., 2021; Li et al., 2016). Studies on the adoption of COVID-19 contact tracing application (Chopdar, 2022), the use of home health robots (Alaiad and Zhou, 2014), the use of mobile applications in the follow-up of patients with diabetes (Shareef et al., 2014) and the use of health information systems by undergraduate students (Hsu et al., 2013) have determined that perceived privacy is determinant in use. In a study conducted to measure the adoption behavior of e-health in developing countries, it was determined that the perceived risk of privacy was effective in technology acceptance (Sahoo et al., 2023). Research on privacy in the acceptance of health technologies focuses more on health service recipients. However, privacy is also an important component for physicians in the patient-physician relationship. Although there are studies that have previously revealed the importance of privacy and reliability in the acceptance of health information systems for physicians (Aljarboa and Miah, 2021), there are also findings that high privacy concerns in physicians do not constitute a significant obstacle to the acceptance of e-health (Dünnebeil et al., 2012). Based on the general literature, the following hypothesis was proposed considering that perceived privacy for physicians is a factor that increases the use of e-Nabız among physicians:
Trust
Trust can be defined as the belief that the provider of information systems applications will act responsibly and fulfill the expectations of the trusting party without exploiting its weaknesses (Gefen, 2000; Pavlou, 2003). Trust is seen as a prerequisite for behavior in decisions (Gefen, 2000). The Ministry of Health fulfills the security measures for the e-Nabız system. However, the adequacy, accuracy and completeness of the information in the system is the responsibility of the institutions that send data to e-Nabız, the relevant personnel in these institutions, the people who add their own data to the system and the service providers of the drug information system from which drug information is received (Republic of Türkiye Ministry of Health, 2023b).
Trust is among the factors that affect acceptance and preference in information systems (Alaiad and Zhou, 2014; Chircu et al., 2000). Within the scope of the UTAUT model, trust has been theoretically and empirically examined, and it is suggested that trust removes barriers to usage intention and behaviors (Cody-Allen and Kishore, 2006). Empirical studies also suggest that trust is effective on usage intention. In a study, when the factors affecting the adoption of e-health were examined, trust affected the intention to use along with perceived risk (Arfi et al., 2021). In other studies related to information systems in the delivery of health services, trust was found to be an important variable in technology acceptance (Alaiad and Zhou, 2014; AlHadid et al., 2022; Zhou, 2012). In a study conducted on Palestinian doctors, intention to use is significantly influenced by trust (Falana et al., 2023). Based on the theoretical perspective of the UTAUT model and the results obtained from several studies given as examples, it is expected that physicians' trust in the e-Nabız system will increase their intention to use it. Therefore, the following hypothesis is proposed:
Theoretical model
The aim of this study is to reveal the factors affecting the acceptance and use of the e-Nabız system by physicians who are among the users of this system. In the study, the UTAUT model was taken as the main model, and in this direction, performance expectation, effort expectation, social impact and facilitating conditions that directly affect usage were taken into consideration within the scope of the model, while the UTAUT model was expanded by including trust and data privacy in the model due to their importance in health information systems (Figure 1).
3. Method
Ethical approval
This study was conducted in accordance with the decision of Düzce University Scientific Research and Publication Ethics Committee, numbered 2022/299. The decision of the ethics committee was delivered to each participant, and the questionnaires were administered on a voluntary basis after being informed.
Sampling and procedures
The population of the study consisted of general practitioners and specialist physicians actively working in any health facility in Turkey, accepting patients. The sample pool consisted of physicians working in foundation, state, education and research, city hospital, university hospital and family health center. First, face-to-face interviews were conducted with 30 physicians in the province where the researchers were located for the pilot study. In the interviews, the comprehensibility of the questions in the measurement tool and the average application time were tested, and then the actual data collection process was started. The data were collected through face-to-face interviews in the province where the researchers resided and online through Google Forms from physicians working outside the provinces where the researchers resided. Online data were obtained by delivering research links to physicians who are members of the platform with special permission from the representatives of the Turkish Private Hospitals Platform. The number of questionnaires distributed was 600. In total, 300 of the 600 questionnaires received a response. As a result of the preliminary examination, a total of 64 scales with incomplete data entry, incomplete and randomly filled in were excluded from the data set. The valuable response rate of data collection was calculated as 47.20%.
Data are produced using survey language, and surveys are susceptible to common method bias (CMB), also known as pervasive technique bias (Podsakoff et al., 2003). The single-factor test created by Harman was used in this experiment to identify the existence of CMB. Data from the analysis using the SPSS 23 version showed that no factor could explain the majority of the variance. The variance interpretation for the first factor was 16.297% of the total variance and under 40%. As a result, according to Lindell and Whitney (2001), we do not believe that CMB had a substantial impact on this study.
Measurement scale
Personal information form, unified technology acceptance and use scale, data privacy scale and trust scale were used to collect research data.
Personal information form.
A form consisting of seven questions, which was created by the researchers and finalized after expert opinion, was used to determine the demographic and job status of the participants (age, gender, working year […].
Unified technology acceptance and use scale.
In the study, the Unified Theory of Acceptance and Use of Technology (UTAUT) developed by Venkatesh et al. (2003) and adapted to Turkish culture by Oktal (2013) was used. Venkatesh et al. (2003) conducted the validity and reliability study of the scale with measurements at three different times. The item loadings of the original scale ranged between 0.75 and 0.94. In the UTAUT scale, 70% of the variance can be explained by the independent variables (performance expectancy, effort expectancy, social influence and facilitating influence) and four moderators (gender, age, experience and willingness to use). The Cronbach’s alpha (Cα) coefficient of the scale adapted to Turkish culture was found to be 0.89. Confirmatory factor analysis (CFA) results indicate that the scale is at an acceptable level (χ2 = 230.89 df = 108; χ2/df = 2.14; root mean square error approximation (RMSEA) = 0.062; goodness of fit index (GFI) = 0.92; adjusted goodness of fit index (AGFI) = 0.88). Within the scope of this study, the scoring of the scale and the instructions were adapted to collect data on the e-Nabız personal data system. To measure usage, the scale was made into a 10-point Likert scale. The scale ranges from 0 = “I never use the e-Nabız system” to 9 = “I use it for almost every patient.”
Trust scale.
Trust was measured with three items adapted to the e-Nabız personal data system by Pavlou (2003) and Jarvenpaa et al. (2000). As the score increases, trust increases. Pavlou (2003) calculated the average variance extracted (AVE) value for the items as 0.95. In the results of the explanatory factor analysis, the factor loading values of the items were found to be above 0.70.
Perceived privacy.
Perceived privacy was measured with three items developed by Casaló et al. (2007) adapted by Yousaf et al. (2021) and adapted by us to the e-Nabız personal data system. As the score increases, privacy increases. Yousaf et al. (2021) determined the AVE value of data privacy as 0.758, the composite reliability (CR) value as 0.90 and Cα value as 0.84, and the loading values of the items were found to be above 0.85.
Data analysis
In the current work, data analysis was carried out using Analysis of a Moment Structure (AMOS®) 23, a very effective piece of software. It is a statistical program that performs structural equation modeling (SEM) using novel methods. It generates an explicit study model for the researchers, provides high-quality drawings for presentation in the paper and calculates the most accurate numerical values. It also features a graphical interface that is simple to use. In the analysis of the data in the study, the construct validity of the measurement model was first tested with CFA, CR and AVE were used to test convergent and discriminant validity, and correlation analysis between constructs was examined, Cα coefficient was calculated to determine reliability. Path analysis was then conducted to determine the relationships between the variables and to test the hypotheses. SEM was used to estimate a goodness-of-fit index between the conceptual model and the sample data. The relative Chi-square ratio over the degree of freedom (χ2/df), the GFI, the RMSEA and the standardized root mean square residual (SRMR) were used to evaluate the goodness of fit of the measurement model and the structural model, respectively.
4. Results
Descriptive analysis
When the demographic characteristics of the 236 physicians reached within the scope of the study are analyzed, it is seen that 96 (40.7%) of the participants were female and 140 (59.3%) were male. The mean age of the participants was 39.3 years (lowest, 25; highest, 67 years). Of the physicians participating in the study, 55 (23.3%) were assistant physicians, 124 (52.5%) were specialists and 57 (24.2%) were general practitioners. The professional experience of the physicians ranged between 1 and 43 years (x¯ = 13.6, SD = 9.9). Of the participants, 224 (94.9%) stated that they used the e-Nabız system, 9 (3.8%) stated that they did not use the system and 3 (1.3%) of the participants did not answer this question. Eighteen physicians (7.6%) reported receiving training on the e-Nabız system, while 218 (92.4%) reported not receiving any training. The results are given in Table 1.
Measurement model
A crucial technique that is frequently used to support a theoretical measurement model is CFA. It demonstrates a relationship between latent or unobserved variables and variables or indicators that are observed. Furthermore, CFA is a statistical method rooted in concepts that elucidate measurement errors and estimate the unidimensional model. As a result, it is recommended to conduct data analysis. Construct validity and reliability were used to evaluate the measurement model. Hence, CFA was performed to scrutinize the factor structure of the variables.
For CFA, the normality of the data, no systematic missing data, adequate sample size and correct model specification which are important assumptions in the application of SEM, were tested. Normality for the variables was assessed by skewness and kurtosis, and minimum, maximum and mean values were also calculated (Table 2). Similarly, the sample size of 236 used in this study is larger than the minimum threshold of 200 cases typically used in SEM studies suggested by Kline (2023). For this, the data were calculated using the maximum likelihood estimates model in AMOS. The measurement dimensions of the model consist of 29 variables, including performance expectancy (PE), effort expectancy (EE), behavioral intention (BI), trust (T), facilitating conditions (FC), social influence (SI) and perceived privacy (PP).
We used factor loading to estimate the reliability of the indicators. Constructs with high loadings indicate that the associated indicators have a lot in common exhibited by the construct. A factor loading greater than 0.50 was considered highly significant. It was observed that all items except Item 3 in the trust dimension contributed significantly to the model with and their loadings were higher than 0.5 (Table 3). Therefore, Item 3 in the trust dimension was removed from the analysis, and the analysis was repeated.
Fit in research refers to a model’s capacity to adequately describe the data. In CFA, a model fit specifically shows how close the observed data are to the theoretical model’s relationship predictions. When a model fits the data well, it is easy to determine if the model is substantially in line with the data or not. As a result, a number of tests were used to quantify the model’s goodness-of-fit in relation to the data. Therefore, the model is declared to be confirmed based on the goodness-of-fit indices.
To assess the model fit, we calculate the χ2/df, and if the ratio is ≤3, it indicates an acceptable fit. If the value is ≤5, it suggests a reasonable fit (Butt et al., 2022; Marsh and Hocevar, 1985). The model used in this study has a normed chi-square (χ2/Df) of 1.814 (<3.00), indicating a satisfactory fit. The GFI is a metric used to compare the fit between the proposed model and the measured covariance matrix. According to Tabachnick and Fidell (2013), GFI value ≤ 1 represents the proportion of variance accounted for by the determined covariance of the population. A value of 1 signifies a perfect fit. With an increasing sample size, the GFI value is expected to increase. A GFI value > 0.95 is considered a good fit, while a value < 0.65 indicates a sustainable fit. The GFI value obtained in this study is 0.851. Comparative fit index (CFI) values were used to measure the model’s fit in this study is 0.941. In covariance structure modeling, the RMSEA. RMSEA measures the difference between the observed covariance matrix per degree of freedom and the predicted covariance matrix representing the model. An RMSEA value < 0.05 is considered a good fit, a value between 0.08 and 0.10 is regarded as an average fit, and a value greater than 0.10 denotes a poor fit. SRMR < 0.09 indicates a good model fit (Hu and Bentler, 1999). In this study, the values obtained for RMSEA and SRMR are 0.059 and 0.062, respectively, indicating a reasonable unidimensionality of the constructs.
Table 4 shows the Cα, CR and AVE values of the items and constructs. Discriminant validity is achieved when the mean value of the square root of the two extracted variants is greater than the value of the correlation between all the variables.
Constructs with a CR value > 0.7 and AVE > 0.5 were considered acceptable (Fornell and Larcker, 1981). All AVE and CR values are above the acceptable level. Convergent validity was achieved for all constructs. Discriminant validity is defined as the degree to which items discriminate between constructs or evaluate individual concepts for the measurement model and is usually verified using Fornell–Larcker and heterotrait-monotrait ratio. This study uses the Fornell–Larcker criterion. According to the Fornell–Larcker method, the square roots of the AVE (located on the diagonal of Table 4) are higher than the relationship between the constructs (corresponding row and column values). Therefore, it indicates that the constructs are strongly related to their relative indicators, unlike other constructs in the model (Chin, 1998; Fornell and Larcker, 1981). Cα internal consistency coefficient was calculated between 0.79 and 0.92. According to these results, it can be concluded that the analyses will yield reliable results (Thomas et al., 2022).
Structural model assessment
SEM analysis is the second mandatory approach in SEM. It is tested after validating the measurement model by showing the relationship between the constructs. Thus, the structural model explains the relationship between variables, has specific relationships between exogenous variables and their corresponding endogenous variables and shows the link between constructs. Structural model results help to assess how closely the empirical data support the theory or to test whether the theory is empirically confirmed (Hair et al., 2010). The goodness of fit of the structural model corresponds to the goodness of fit of the CFA measurement model. The fit indices in the proposed structural model indicate a good fit between the predicted model and the observed data (χ2/df = 1.314, GFI = 0.992, CFI = 0.998, RMSEA = 0.037 and SRMR = 0.020).
Path analysis was used to examine exogenous variables and determine both their immediate and long-term impacts. The relationship between the constructs is shown in a path diagram in Figure 2 based on findings from earlier work. PE, EE, SI, FC, PP ve T are endogenous variables while BI ve user behavior are an exogenous variable.
According to the path model, performance expectancy (PE), effort expectancy (EE) and trust variables have a statistically positive effect on behavioral intention (BI) (p < 0.05). Contrary to the hypothesis, privacy (PP) variable has a statistically negative effect on BI (β = −0.117; SE = 0.058; p = 0.045). While all six hypotheses were accepted, one of them was not found to be statistically significant. In other words, social interaction did not have a statistically significant effect on behavioral impact. The information on the acceptance or rejection of the hypotheses is summarized in Table 5.
5. Discussion-Conclusion
The aim of this study is to reveal the determinants of physicians’ acceptance and use of the e-Nabız system. For this purpose, performance expectancy, effort expectancy, social influence and enabling conditions are taken within the scope of UTAUT, while the impact of trust and data privacy are also examined within the model due to their importance in health information systems. Previous research on eHealth has largely been conducted in a developed country context (Almarzouqi et al., 2022). However, in this study, it was concluded that the UTAUT model provides an appropriate framework to predict the factors affecting the intention to use and behavior of e-Nabız, one of the e-government applications in a developing country, and that trust and privacy variables are important in the acceptance of health information technologies.
Performance expectancy was the strongest determinant of intention to use in this study. Within the scope of the research, H1 was accepted. Physicians believe that they will achieve high performance and good results with the e-Nabız application. In many studies conducted within the framework of the UTAUT model, similar to this study, the relationship between performance expectancy and behavioral intention was found to be positive (Alaiad and Zhou, 2014; Öner Gücin and Sertel Berk, 2015; Raman and Don, 2013). Similarly, Faida et al. (2022) found that performance expectancy explained a significant portion of behavioral intention on the use of EMR. In a study involving physicians regarding the e-Nabız application, performance expectancy was found to be effective on behavioral intention (Bektaş Uçar, 2021). Again, in a study conducted on e-Nabız, 91.5% of physicians stated that e-Nabız facilitated their work, 94.5% stated that it accelerated their work and 92.5% stated that they found the application useful; the descriptive findings obtained indirectly contributed to the research findings that physicians meet the performance expectation. When individuals perceive that using a technology or system will lead to better performance or outcomes, they are more likely to intend to engage in the behavior associated with that technology or system. As physicians perceive that e-Nabız will lead to better performance or outcomes, they will intend to use the technology. This understanding creates an enabling environment for the design and implementation of interventions or strategies to promote the adoption and use of e-Nabız.
Another hypothesis accepted within the scope of the research is H2. Effort expectancy was identified as an important factor that positively affects behavioral intention. When physicians perceive the e-Nabız application as easy and effortless to use, they are more likely to adopt and intend to use it. Many studies are similar to the findings of this study. Mensah et al. (2022) found that performance expectancy affects behavioral intention in the context of mobile health services. Cheng et al. (2022) showed that effort expectancy is positively related to health-care workers’ intention to adopt AI-assisted diagnosis and treatment. Dash and Sahoo (2022), in a study on the factors affecting patients’ willingness to receive digital health advice in India, similarly revealed a positive relationship between effort expectancy and intention to use. In parallel with the findings of this study, the results in the literature show that effort expectancy is an important determinant of behavioral intention.
In this study, it has been revealed that social influence is not an important determinant in explaining physicians’ intention to use the e-Nabız application. According to this result, H3 could not be confirmed. The results obtained in previous studies are contradictory. In a group of studies, social influence has been found to contribute positively to usage intention (Cheng et al., 2022; Venkatesh et al., 2003). However, on the contrary, similar to these research findings, there are also studies that cannot confirm the effect of social influence on behavioral intention (Chang et al., 2007; Dash and Sahoo, 2022; Ong et al., 2022; Shiferaw et al., 2021). Rouidi et al. (2022), who examined the remote care technologies of health-care professionals with the systematic analysis method, identified social influence as the weakest predictor. There is also a theoretical perspective that considers that social influence on intention to use may decrease over time as experience with the system increases. It is interpreted that before a system is developed, users’ knowledge and beliefs about the system are “uncertain and ill-formed,” and therefore, they have to rely more on the opinions of others as the basis for their intentions. After using the system, the normative effect decreases as more is known about its strengths and weaknesses (Hartwick and Barki, 1994; Venkatesh and Davis, 2000). Hartwick and Barki (1994) found this effect after a three-month period. The e-Nabız application has been in existence since 2015, and it can be thought that the physicians’ own opinions were formed in this case, and the social influence on usage behavior disappeared. In general, although the literature shows that social influence can significantly affect behavioral intention and usage behavior in various contexts, the extent and direction of this influence may vary depending on factors such as the type of behavior considered. More research is needed to explore the mechanisms and dynamics of social influence on behavioral use in different domains.
Within the scope of the research, as in the study of Venkatesh et al. (2003), facilitating factors positively affected usage behavior, and H4 was accepted. The presence of institutional and technical infrastructure increases physicians’ e-Nabız usage behavior; e-Nabız usage depends on the support of information and communication technology infrastructure. Many studies examining the relationship between facilitating conditions and usage behavior provide insight into the determinant role of facilitating conditions (Rouidi et al., 2022). Hossain et al. (2019), in their research on the factors affecting the adoption of EHR by physicians in the Bangladesh health system within the framework of UTAUT and Bektaş Uçar (2021) in their research, explain the use of e-Nabız application by physicians found facilitating conditions as a determinant of usage behavior. There are also studies that cannot reveal the effect of facilitating factors against these results (Shiferaw et al., 2021). These research findings are parallel to the results of Hossain et al. (2019) and Bektaş Uçar (2021). Understanding the impact of facilitating conditions on usage behavior can provide valuable insights for the design and implementation of technology-based systems and services.
As expected, this research shows that behavioral intention plays an important role in influencing usage behavior, and H5 is accepted. However, the relationship between behavioral intention and usage behavior is complex and can be influenced by various moderators and mediators. Intention to use and facilitating factors were able to explain 19% of usage behavior (Figure 2). Understanding these factors can provide valuable information for designing interventions and strategies to promote the adoption and use of technologies and services. Similar results have been found in previous studies on health technologies (Almarzouqi et al., 2022; Rouidi et al., 2022). Chang et al. (2007) found that behavioral intention is an important determinant in the emergence of behavior in their research on pharmacokinetics-based clinical decision support on physicians.
Within the scope of this study, it was concluded that perceived privacy is an important variable affecting behavioral intention. However, contrary to expectations, perceived privacy was found to be inversely related to behavioral intention. Although physicians have low perceptions of privacy, their intention to use the e-Nabız application is high. However, there are measures to protect privacy in the e-Nabız system, which is one of the e-government applications. Users are authorized for the duration and limits of system use, and within the framework of this authorization, PHR can be made available to the relevant physician(s) (Birinci, 2023; T.C. Republic of Türkiye Ministry of Health, 2023b). In order for physicians to view the health data in the e-Nabız application, the patient must be authorized by the patient and the patient must have an appointment with the physician for that day; after these conditions are met, the physician can view the health data by logging in with e-Government or e-Signature. If the patient has not authorized access to his/her health records by the physician but wants to share his/her health records with the physician at that moment, the login process is initiated when the physician sends an short message service confirmation code to the patient’s phone and the patient gives this code to the physician. Data is not shared with third parties, institutions and organizations except for legal obligations (Republic of Türkiye Ministry of Health, 2023a). Users can also view the internet protocol address, the date and time of the last access (Birinci, 2023). Despite the measures taken, in other studies, only 48.1% of health-care professionals (Güngör Ketenci et al., 2021) and 34.7% of physicians (Karahisar, 2018) think that their personal health data are secure in the e-Nabız system. In a study conducted by Dünnebeil et al. (2012) for physicians, the perceived usefulness level of physicians with high privacy demands regarding e-health application was also found to be high. However, high privacy concerns did not constitute a significant obstacle for the acceptance of e-health. A similar result was encountered in this study. The high level of other determinants (performance, effort expectation and trust) related to the e-Nabız application may be considered to override privacy concerns. At the same time, the fact that the responsibility for the application lies with the Ministry of Health, and physicians do not feel responsible for this issue may have left privacy concerns in the background. Many studies have found that confidentiality (privacy) increases the intention to use. However, these research findings generally reflect the results of those who receive health services. For example, Chopdar (2022) found findings on the public’s contact tracing practices during the COVID-19 pandemic, Gu et al. (2021) on patients’ intentions to adopt electronic health technology, and Li et al. (2016) on the importance of privacy for wearable health technology users. Compared to patient groups, it can be thought that privacy may have a different effect on determining physicians’ behavioral intentions.
The results of this study, similar to the results of other studies, confirm that trust has a positive and significant effect on usage intention, and thus, the final hypothesis, H7, is accepted. These results support the findings in the literature. Chopdar (2022), in his research on the adoption of monitoring applications related to the COVID-19 virus theme, determined that in addition to the UTAUT model, perceived security is important in the adoption of the application. In addition, trust has been an important determinant in other studies (AlHadid et al., 2022; Dash and Sahoo, 2022).
Overall, this study confirms that the UTAU model provides an appropriate framework, excluding the social influence variable, for predicting the factors affecting physicians’ usage intention and behavior of the e-Nabız application. In addition, the empirical findings show that trust, which is considered as an additional variable in the model, is an important variable, indicating that recent concerns about the perceived privacy of EHR also apply to e-Nabız, but this does not pose a barrier to physicians’ intention to use.
6. Study contribution
Practical contribution
The practical contributions of this study offer key insights into enhancing the acceptance and usage of the e-Nabız system among physicians. The findings provide vital clues for the adoption and effective use of health technologies. These practical contributions can be examined under five main categories: system design and development, education and awareness, policy-making and management, importance of trust and privacy and user-centered approach.
In terms of system design and development, it has been identified that factors such as performance expectancy, effort expectancy, trust and perceived privacy have a significant impact on physicians’ behavioral intentions. This indicates that these factors should be prioritally considered in the design of health information systems like e-Nabız. The study underscores the importance of training programs and awareness campaigns in enhancing physicians’ awareness of the fundamental factors affecting their adoption of technology. Such programs can assist physicians in using the system more effectively. For health policymakers and administrators, this research provides vital information. It can contribute to the development of strategies necessary for the wider and more effective use of the e-Nabız system. The finding that trust and perceived privacy significantly influence physicians’ adoption of technology highlights the importance of these two elements in the development of health information systems. This could be a key driver in increasing technology usage in the health sector (Esmaeilzadeh, 2019; Iott et al., 2019).
The study emphasizes the importance of centering user needs and expectations in the development and implementation of health technologies. Feedback and experiences from physicians can play a critical role in making the system more user-friendly.
These findings can provide valuable contributions to the development and widespread implementation of the e-Nabız system and similar health technologies. Additionally, they could serve as a guide in the adoption and effective use of technology in the health sector.
Societal contributions
From a societal perspective, the insights provided by this study on the acceptance and usage of the e-Nabız system by physicians have a wide-ranging impact. This research enhances the quality of health-care services by improving the effective use of the e-Nabız system in diagnostic and treatment processes, thereby elevating patient safety and privacy, raising health awareness, improving public health, contributing to the formulation of health policies and encouraging the integration of technological innovations in the health sector. These extensive impacts play a significant role in the advancement of health-care services.
The effective use of the e-Nabız system by physicians can enhance the quality of health-care services. The system offers physicians easy access to patients’ health histories and treatment information, facilitating more accurate and rapid decision-making. This aids in the improvement of diagnosis and treatment processes.
The study’s emphasis on the importance of trust and privacy highlights the protection of patients’ personal health information. A secure and privacy-respecting system can increase patients’ trust in the health-care system, contributing to better health outcomes. Effective utilization of the e-Nabız system can facilitate patients’ access to their own health data, thereby increasing health consciousness and health literacy. This empowers patients to better manage their health and use health-care services more effectively.
The efficient use of the system can increase the efficiency and accessibility of health-care services across the community. This contributes to the improvement of overall public health and the more equitable distribution of health-care services (Diel et al., 2023). This research can assist health policymakers in developing strategies to increase the acceptance and use of systems like e-Nabız. Policymakers can use the findings of such research to develop policies that ensure the more effective and efficient delivery of health-care services. The successful implementation of health technologies like the e-Nabız system can promote wider acceptance and integration of technological innovations in the health sector. This contributes to the provision of more innovative and effective health-care services across society.
In conclusion, this study can play a significant role in enhancing the quality of health-care services, patient safety and overall health awareness. Research of this nature holds critical importance in the development of the health sector and in improving public health.
Theoretical contributions
This study, focusing on physicians’ adoption of health information systems like e-Nabız, offers valuable theoretical contributions in four distinct aspects: model extension and adaptation, addressing research gaps and implications for future research.
The study extends the theoretical background of the UTAUT by incorporating sector-specific variables such as trust and privacy, which have not been extensively discussed in the existing literature. These additions provide new perspectives and focus on the contributions of elements that have been overlooked in current theories. Critically, while some may argue that the study does not significantly contribute to the field, it actually adds substantial depth to the existing literature by exploring the role of additional variables like trust and privacy in the adoption of health information systems.
By examining the experiences and perspectives of active physicians within the UTAUT framework, this research makes a unique contribution to the field. Considering the limited number of studies targeting this specific audience with the UTAUT model (Diel et al., 2023), this research deepens our understanding of the adoption of health information systems and offers new perspectives (Bawack and Kamdjoug, 2018; Ranjeni et al., 2023). Particularly in the context of a developing country like Turkey, it provides valuable insights into how physicians perceive and accept health information systems like e-Nabız, thereby adding an important and innovative perspective to the existing literature.
The main findings of this study, which used the UTAUT model, indicate that factors such as performance expectancy, effort expectancy, trust and perceived privacy significantly influence physicians’ intentions to adopt e-Nabız. It was observed that social influence does not have a significant effect on behavioral intention. These findings could shape future research in various ways.
Additionally, future studies could explore factors specific to physicians’ acceptance of information technologies, such as the role of clinical utility or the impact of technology on doctor-patient communication. The findings suggest potential modifications in acceptance models within the health context, integrating aspects specific to health IT like clinical workflow integration (Kumar and Mostafa, 2020) or perceptions of data security (Yao et al., 2023), thereby enhancing the predictive power of the model.
These aspects contribute not only to academic knowledge but also offer practical implications for the development, implementation and promotion of health IT systems like e-Nabız. This is crucial for the successful integration of such technologies into clinical practice.
7. Limitations and future research
Although this study is a pioneering study in terms of determining the usage intention and behavior of physicians regarding the e-Nabız system, which is needed in the literature, it has a number of limitations.
The first one is methodological. In this study, data were collected cross-sectionally. Therefore, the results obtained cannot show the causal relationships between the variables examined. Conducting a longitudinal or qualitative examination of the causal relationships between variables in a separate study could yield more detailed results. Another limitation of the study is related to the groups from which data were collected. The majority of the physicians participating in the study (94.9%) were physicians using the e-Nabız system. The representation rate of non-user groups is quite low. Therefore, it can be concluded that the model is limited to explaining the usage intention and behavior of physicians using the application. Another limitation of the research group is that the research covers physicians working at all levels, and does not provide information on e-Nabız specific to the field or level. However, results may differ at different levels and levels of health-care services where patient privacy or trust is more important. While this is a limitation of the current research, it also highlights an issue that should be explored by researchers interested in the subject. Since this study mainly focuses on the UTAUT model, trust and privacy issues, some important issues affecting intention and behavior in social and psychological fields and personal characteristics (age, gender, education, etc.) were ignored as they focus on the main variables related to the model. This study was conducted only among physicians in Turkey, which limits the generalizability of the findings.
In addition to these, it may be suggested that privacy regarding information technologies in health-care services should be investigated in more detail in terms of physicians. In this study, it is seen that the perception of privacy continues to be a source of concern, although it is ignored when other variables of the practice are found to be high (performance expectation, effort expectation, facilitating conditions and trust). Privacy concerns can be influenced by personal concerns, laws, regulations, controls and even various components of government interventions such as publicity (Belanger and Crossler, 2011). Research at the individual, group and organizational levels is needed to explore these issues.
Conflicts of interest: None declared.
Funding: None. This research did not receive any specific grant from funding agencies in the public, commercial or not-for-profit sectors.
Data availability: Data will be made available on request.
Credit author statement: Dilek Şahin, conceptualization, literature review, methodology definition, discussion and conclusions and writing; Mehmet Nurullah Kurutkan, calculations, analysis of results and discussion; Tuba Arslan, field data collection, writing revision.
Figure 1.Theoretical framework for the current study
Figure 2.Standardized values of structural model
Table 1.
Respondents’ profile
| Variable | Description | Frequency | (%) |
|---|---|---|---|
| Gender | Female | 96 | 40.7 |
| Male | 140 | 59.3 | |
| Title | Assistant physician | 55 | 23.3 |
| Specialist | 124 | 52.5 | |
| General practitioner | 57 | 24.2 | |
| E-Nabız Use | Used the e-Nabız | 224 | 94.9 |
| Not use e-Nabız | 9 | 3.8 | |
| Not answer this question | 3 | 1.3 | |
| E-Nabız Training | Receiving training | 18 | 7.6 |
| Not receiving any training | 218 | 92.4 |
| Mean | SD | Min. | Max. | |
| Age | 39.3 | 9.9 | 25 | 67 |
| Seniority | 13.6 | 9.9 | 1 | 43 |
Source: Authors’ own work
Table 2.
Descriptive statistics
| Constructs | N | Minimum | Maximum | Mean | Std. deviation | Skewness | Kurtosis | ||
|---|---|---|---|---|---|---|---|---|---|
| Statistic | Std. error | Statistic | Std. error | ||||||
| PE | 236 | 1.33 | 5.00 | 4.4225 | 0.64901 | −1.393 | 0.158 | 2.504 | 0.316 |
| EE | 236 | 1.00 | 5.00 | 3.9133 | 0.82252 | −0.600 | 0.158 | 0.090 | 0.316 |
| SI | 236 | 1.00 | 5.00 | 3.5882 | 0.87657 | −0.386 | 0.158 | −0.233 | 0.316 |
| FC | 236 | 1.00 | 5.00 | 3.8036 | 0.82353 | −0.758 | 0.158 | 0.772 | 0.316 |
| BI | 236 | 1.00 | 5.00 | 4.2066 | 0.80860 | −1.033 | 0.158 | 1.039 | 0.316 |
| PP | 236 | 1.00 | 5.00 | 3.5826 | 1.06362 | −0.454 | 0.158 | −0.435 | 0.316 |
| T | 236 | 1.00 | 5.00 | 3.7929 | 0.85698 | −0.573 | 0.158 | 0.139 | 0.316 |
Notes:PE = performance expectancy; EE = effort expectancy; SI = social influence; FC = facilitating conditions; BI = behavioral intention; PP = perceived privacy; T = trust
Source: Authors’ own work
Table 3.
Confirmatory factor analysis
| Constructs | Items | Estimate | S.E. | C.R. | Stand. est. | P |
|---|---|---|---|---|---|---|
| PE | PE1 | 1.000 | 0.645 | |||
| PE2 | 1.743 | 0.139 | 12.549 | 0.777 | *** | |
| PE3 | 1.921 | 0.193 | 9.928 | 0.793 | *** | |
| PE4 | 2.029 | 0.197 | 10.286 | 0.830 | *** | |
| PE5 | 1.098 | 0.141 | 7.792 | 0.581 | *** | |
| PE6 | 1.787 | 0.172 | 10.368 | 0.835 | *** | |
| EE | EE1 | 1.000 | 0.674 | |||
| EE2 | 1.080 | 0.113 | 9.529 | 0.714 | *** | |
| EE3 | 1.133 | 0.106 | 10.730 | 0.920 | *** | |
| EE4 | 1.006 | 0.092 | 10.893 | 0.858 | *** | |
| SI | SI1 | 1.000 | 0.731 | |||
| SI2 | 0.933 | 0.085 | 10.951 | 0.798 | *** | |
| SI3 | 1.007 | 0.067 | 14.984 | 0.741 | *** | |
| SI4 | 0.788 | 0.091 | 8.678 | 0.617 | *** | |
| SI5 | 0.781 | 0.093 | 8.440 | 0.600 | *** | |
| FC | FC1 | 1.000 | 0.749 | |||
| FC2 | 1.004 | 0.084 | 11.944 | 0.816 | *** | |
| FC3 | 0.931 | 0.118 | 7.860 | 0.540 | *** | |
| FC4 | 1.041 | 0.092 | 11.259 | 0.766 | *** | |
| BI | BI1 | 1.000 | 0.753 | |||
| BI2 | 1.408 | 0.093 | 15.191 | 0.929 | *** | |
| BI3 | 1.415 | 0.093 | 15.289 | 0.937 | *** | |
| PP | PP1 | 1.000 | 0.890 | |||
| PP2 | 1.049 | 0.054 | 19.581 | 0.886 | *** | |
| PP3 | 1.011 | 0.051 | 19.915 | 0.894 | *** | |
| T | T1 | 1.000 | 0.849 | |||
| T2 | 0.862 | 0.060 | 14.417 | 0.774 | *** | |
| T4 | 0.760 | 0.074 | 10.309 | 0.609 | *** |
Note:***P < 0.001
Source: Authors’ own work
Table 4.
Discriminant validity and reliability
| Constructs | α | CR | AVE | EE | BI | T | FC | PE | SI | PP |
|---|---|---|---|---|---|---|---|---|---|---|
| EE | 0.855 | 0.855 | 0.597 | 0.773 | ||||||
| BI | 0.888 | 0.891 | 0.732 | 0.585 | 0.856 | |||||
| T | 0.791 | 0.797 | 0.576 | 0.652 | 0.646 | 0.759 | ||||
| FC | 0.824 | 0.843 | 0.664 | 0.726 | 0.514 | 0.574 | 0.815 | |||
| PE | 0.888 | 0.889 | 0.576 | 0.555 | 0.722 | 0.620 | 0.418 | 0.759 | ||
| SI | 0.839 | 0.837 | 0.517 | 0.485 | 0.501 | 0.672 | 0.512 | 0.614 | 0.719 | |
| PP | 0.920 | 0.920 | 0.793 | 0.489 | 0.394 | 0.889 | 0.409 | 0.417 | 0.559 | 0.891 |
Source: Authors’ own work
Table 5.
Path coefficients for the final model
| Hypotheses | Relations | Estimate | Std. estimate | S.E. | C.R. | P | Results |
|---|---|---|---|---|---|---|---|
| H1 | BI ← PE | 0.524 | 0.420 | 0.074 | 7.045 | *** | Supported |
| H2 | BI ← EE | 0.188 | 0.192 | 0.057 | 3.309 | *** | Supported |
| H3 | BI ← SI | 0.020 | 0.022 | 0.055 | 0.362 | 0.718 | Not Supported |
| H4 | UB ← FC | 0.310 | 0.127 | 0.156 | 1.985 | 0.047 | Supported |
| H5 | UB ← BI | 0.916 | 0.369 | 0.159 | 5.762 | *** | Supported |
| H6 | BI ← PP | −0.117 | −0.153 | 0.058 | −2.001 | 0.045 | Supported* |
| H7 | BI ← T | 0.308 | 0.327 | 0.081 | 3.830 | *** | Supported |
Notes:*A significant relationship was found, but the expected direction of the relationship was reversed; ***P < 0.001
Source: Authors’ own work
© Emerald Publishing Limited.