How the COVID-19 Pandemic Affected the Sustainable Adoption of Digital Signature: An Integrated Factors Analysis Model

Abstract

Digital signatures have been widely and primarily used for document approval activities during the Coronavirus pandemic in Indonesia. This is the digital equivalent of a handwritten signature or stamped seal, although it provides more inherent security, such as validating the authenticity and integrity of a message, software, or virtual document. Therefore, this study aims to determine factors affecting consumer intention in using digital signatures based on (i) the unified theory of acceptance and use of technology 2, (ii) the theory of planned behavior, and (iii) the information acceptance model. A total of 358 respondents answered the online questionnaire containing 69 question items, with the data analyzed using the structural equation modeling technique to examine the hypotheses. The results showed that the relationship between consumers’ attitudes, perceived behavioral control, subjective norms, and information adoption had the highest and lowest effects on consumers’ behavioral intention. Moreover, the consumers’ attitudes had the most significant effect on their attitudinal intention to use digital signatures. The significant positive impact of these consumers’ attitudes had relevant implications for the sustainable adoption of the signature system. This indicated that an integrating model with the potential of extending to consumers’ intention analysis was established for digital signature adoption in other countries after the post-Coronavirus period.

1. Introduction

The acceptance and adoption of digital technology have reportedly elevated in the last few decades due to the consumers’ increasing intention to change conventional behavior into a daily digital attitude. This indicates that technological developments and globalization have obtained communication systems in the modern era, specifically in the digitization field within ICT (information and communication technology) [1]. Based on this development, many fraud issues have reportedly been observed in the IT world related to data originality, such as the spread of fake emails involving person’s identity. With the occurrence of these issues, message authentication techniques are found to be indispensable. This led to the development of a new societal model, namely, digital technology, which became a necessity in the change of habits while also having a positive impact on social activities [2].

Countries such as Russia, the United States, and Canada are considered to be developed in a large area while, subsequently, having a high number of internet users. In Russia, the penetration of these users showed that 114.78 million (76.9%) of the total population was connected to the internet [3]. Furthermore, 237.1 thousand fraud issues involving the use of the internet, mobile devices, computers, or other information and telecommunication technologies were recorded in 2020 within Russia [4]. The penetration of internet users in the United States also showed that 313 million (90.8%) of the total population was digitally connected to the internet [5], where the occurrence of 156 million data exposures attacked an organization’s database in 2020. However, these cybercrimes’ issues have reportedly decreased compared to 2019 [6]. In Canada, the penetration of these users indicated that 36.4 million (95.6%) of the total population were connected to the internet [7], while 48% of the 36.4 million Canadians were extremely concerned about personal identity theft, with 28% of organizations reporting the occurrence of cyber-attacks [8].

In contrast to developed countries, Indonesia reportedly has a rapid growth of internet users as a developing nation. This indicates that more than 202.6 million (74.7%) of the total population is connected to the internet [9]. The country also has a low level of security system compared to Russia, the United States, and Canada. This is due to the occurrence of approximately 495.3 million global attacks in 2020 (a 41.4% increase from 2019) [10], leading to Indonesia being ranked in the first position as a worldwide cyber-attack destination. These showed that the distribution of provocative content, online fraud, and identity data theft were the biggest issues in the country. In addition, potential crimes often occur in online transactions, such as fraud, personal data misuse in cyberspace (carding), hacking, spamming, and defacing, when the e-commerce infrastructure security system is still weak [11].

Since the crowd restriction announcement made by the Indonesian government due to the COVID-19 pandemic, every aspect of environmental activity has drastically changed [12]. This was because the citizens considered these viruses to be very dangerous to body defense, with the government’s response being slow in avoiding the outbreaks. Based on restricting public access, the government subsequently and strictly prohibited social gatherings, with violators encountering severe sanctions. Meanwhile, the worries about the COVID-19 infection prompted uncaring social behavior [12], with the Ministry of Communication and Information showing a slight increase in fraud in e-commerce transactions during this period. According to the signing process, some of the conventional barriers encountered were the similar location of the signer and the concerned documents, as well as the increased forgery through signature and personal identity duplication [13]. This shows that digital signatures are the right solution to prevent security issues such as document forgery and personal identity theft [14]. Besides, the importance of the method is to prevent online fraud in document transactions through electronic cryptography to authenticate the developed digital process. It also ensures the identity of the signer through encryption and decryption algorithms by using a private key to protect the documents against data theft and forgery. Using cryptographic techniques, a digital signature was developed to secure transmissible information. This indicates that cryptography contains two primary keys used for message encryption and decryption, respectively [15].

Based on the Indonesian Bank data, economic deals and physical payments have decreased since the beginning of 2019, with the nominal value of e-commerce transactions estimated at IDR 266.3 trillion, or an increase of around 29.6% in 2020 [16]. These values are in line with the rapidly increasing e-commerce transactions, as digital payment deals are also found to be highly elevated. This is observed in the increased use of electronic money (EM) within e-commerce transactions, which generally causes very high market utilization. From the first to the third quarter of 2019–2020, the utilization of electronic money (e-commerce transactions) increased, with 41.71% achieved in the 4th quarter of 2020. This highly exceeded the market through the bank transfer and cash deposit methods, which only gained approximately 20.23% and 19.01%, respectively. With the increasing need for e-commerce in Indonesia, the adoption of digital signatures is required to accelerate the online transaction process. Moreover, e-transactions need information confidentiality and document legality based on Government Law No. 11 of 2008 concerning Electronic Information Systems and Transaction Operation. This is to legalize an electronic transaction through the implementation of digital signatures [15]. For appropriate security, these transactions are protected in regulation PP 82/2012, which states that electronic documents and information should maintain confidentiality, authenticity, accessibility, and availability through the e-system administrator. This indicates that an electronic system administrator should implement the authenticity combination in the e-signature to verify the signer’s identity (Article 58 (2) PP 82/2012).

According to [17], the factors affecting the adoption of electronic signatures were evaluated based on the executives’ perspective of the hospital information department in Taiwan. This indicated that the technology-organization-environment (T-O-E) model was applied to identify the factors affecting hospitals in adopting e-signature. The results showed that 70% of Taiwan hospitals were delaying the adoption of this system due to procrastination in the computerized medical records development on the digital project. The study of [18] also analyzed the factors affecting the use of a digital signature system with the technology acceptance model in Turkey. This indicated that the technology acceptance model (TAM) was applied to identify the variables affecting user’s perceptions and behaviors in using the signature system, to determine and conceptualize the impact level of each factor in the structural method. The results showed that perceived usefulness and utilization ease positively and insignificantly influenced consumers’ attitudes. In addition, [19] evaluated the factors influencing customers’ interest in adopting digital signature services with cloud technology in South Korea. This indicated that the TAM and T-O-E models were applied to a specific security area. The results show that consumers’ expectations for service preparation decreased when the organization did not support the reliability of cloud-based technology in the signature system. Therefore, this present study aims to determine the factors affecting the sustainable adoption of the digital signature system during the COVID-19 pandemic in Indonesia. This is carried out by using (i) an integrated unified theory of acceptance and use of technology 2 (UTAUT2), (ii) a theory of planned behavior (TPB), and (iii) an information acceptance model (IACM) approach. These are applied because no previous study has adopted the combination of the three theories in investigating the customer’s intention of using digital signature technology. The results are expected to be utilized and extended to measure the factors affecting the adoption of digital signatures in other countries, even after the post-COVID-19 period.

The factors influencing consumers’ intention and adoption of digital signatures are expected to be identified for a document validation process. The following are also likely to be utilized in understanding the consumers’ intention to adopt the digital signature in Indonesia: (i) the integration of the unified theory of acceptance and use of technology 2 (UTAUT2), (ii) the theory of planned behavior (TPB), and (iii) the information acceptance model (IACM) framework. Based on this condition, attitude (AT) is reflected in privacy and security (PS), performance expectancy (PE), effort expectancy (EE), hedonic motivation (HM), price value (PV), and habit (HT). Furthermore, information usefulness (IU) is reflected in information quality (IQ), information credibility (IC), and the needs of information (NOI). Facilitating conditions (FC), perceived behavioral control (PBC) is expected to be influenced by facilitating conditions (FC). Subsequently, attitude (AT), subjective norms (SN), perceived behavioral control (PBC), and information adoption (IA) should be influential on the consumers’ behavioral intentions (BI) to use the digital signature system. Besides the objectives of this study, the remaining sections include Section 2, which explains the conceptual framework of the proposed model with the UTAUT2, TPB, and IACM techniques; Section 3, which evaluates the details of the methodology, questionnaire design, and respondent demographic characteristics; Section 4, which assesses and analyzes the survey results and conceptual framework; Section 5, where implications, limitations, and future suggestions are observed to enrich the utilized conceptual model.

2. Conceptual Framework

2.1. Theoretical Models and Hypotheses

A digital signature is an alphabetical sequence obtained by processing transmitted texts in a specific writing style created by cryptographic technology. This leads to the verification of the intended text source while confirming any subsequent alterations [20]. Digital signatures also have some advantages, such as (1) increasing the speed of transactions, (2) reducing costs, (3) increasing security, (4) official utilization, (5) non-repudiation, (6) preventing fraud, and (7) having time stamps. Meanwhile, several disadvantages are also observed, including (1) short life and expiration, (2) the necessity of obtaining a certificate and software, (3) no common law in some countries, (4) technological compatibility, (5) training and troubleshooting, as well as (6) security concerns [21].

This framework was proposed based on the identification and determination of the latent variables and developed model correlation (Figure 1). The researchers have proposed the second-order model. The form of a second-order factor can embody the meaning of many first-order latent variables. Every first-order latent variable is applied as a reflective indicator of the second-order model.

The consumers’ technology adoption was also explained by three critical psychological constructs, namely, AT, SN, and PBC in the TPB model [22]. SN is known to directly influence BI in the TPB model, although it has similar interpretations to the social influence within UTAUT2. To use a digital signature, the direct effects of other important people (i.e., family, friends, and colleagues) were analyzed in this study. The methods by which the IACM IA influenced consumers’ BI were also evaluated during the COVID-19 pandemic in Indonesia

2.1.1. Determinants of Users’ Attitude towards Digital Signature System Adoption

This evaluates the hypothesis incorporating UTAUT2, TPB, and IACM models, as shown in Figure 1, where six factors of UTAUT2 are reflected in consumers’ attitudes towards digital signature adoption. This indicated that AT in the proposed framework depended on the user’s perception of digital signature utility. In this framework, the initial determinant of the reflected factor from consumers’ attitude is privacy and security (PS), with the Ministry of Communication and Information stating that the occurrence of cybercrimes in 2020 increased by 79% compared to 2019 [23]. This indicated that the effects of cybercrime were felt in the internet network, with the cloud computing system providing various advantages in the security of the data [24]. To ensure data confidentiality, integrity, and availability (CIA), the sender and the receiver of the documents encrypt and decrypt the digital text using public and private keys to verify the signature. This encryption aims to protect some privacy against the risk of secret key exposure in the digital signature system [25].

The second determinant is performance expectancy (PE), which is the advantages acquired by consumers when using the innovation system or services [26]. According to [27], PE was the most critical factor determining users’ BI to use new technologies. According to [21], it is also stated that digital signatures provide several advantages and utilities to consumers, such as (1) increasing the speed of transactions, (2) reducing costs, (3) increasing security, (4) official utilization, (5) non-repudiation, (6) preventing fraud, and (7) having time stamps. This indicated that the advantages affected consumers’ AT towards digital signature adoption.

The third determinant is effort expectancy (EE), which is the level of simplicity and flexibility related to utilizing an innovation system [26]. This concept was similar to the perceived ease of use (PEOU) factor in TAM theory [28], which supported that the level of consumers’ expectations should be risk-free without difficulty in using a system [29]. In the digital signature system, the available features still need to be more developed by adding new attributes to reduce consumers’ effort. This new feature automatically links a signature authentication between one utilized device. It also indicates that consumers do not need to bother re-creating signature authentication on each device. These were observed to save time and improve consumers’ AT towards digital signatures.

The fourth determinant is hedonic motivation (HM), which is the fun or enjoyment obtained by consumers while using a digital signature. Several previous studies showed that this factor was significant in determining consumers’ BI towards new technologies [25,26]. This indicated that users seeking sensation and adventure were more inclined to use new technology, such as a digital signature. However, those primarily seeking novelty felt bored after the adoption time became obsolete while trying to acquire efficiency from the innovation system [25]. This indicates the level of HM greatly contributed to the pleasure and enjoyment of technology users by increasing the utility and capability of the system.

The fifth determinant is price value (PV), which is the trade-off benefit obtained by consumers while using digital signatures towards the monetary costs [30]. Based on the customer, price issues were critical and received specific interest during the acceptance and rejection of the technology innovations [26]. The required facilities and resources (i.e., 4G internet services, Wi-Fi, and personal and other necessary devices) of the digital signature system were also critically adding to the expense costs of customers, leading to the critical enhancement of PV in the conceptual model [26,29]. In addition, the impact of the pandemic caused the inequality of opportunities to adopt the technology due to being constrained by income [31]. With a decrease in income, customers experienced difficulties adopting technology, leading to an increased risk of contracting COVID-19. When the expenses of digital signature services were very low, the implementation of economic transactions was found to be one of the positive effects on the technological adoption of the system [32].

The sixth determinant is a habit (HT), which is the method by which people automatically perform behaviors due to previous learning [33]. HT perspective is also defined as a strong predictor in influencing consumers’ BI towards the same subsequent technology system [34]. Furthermore, several previous studies determined that past behaviors, reflexes, and individual experience were measurable indicators of HT [25,35]. This indicated that previous use produced HT to enhance the interactions and experiences of adopting an innovation system as well as influence consumers to futuristically utilize the technology [26]. When consumers utilized technologies for a long time, the HT of using a digital signature was automatically produced and stored in their conscious minds. However, consumers without previous experience of using a digital signature required more time and effort to learn and familiarize themselves with the technology [36]. This caused discomfort and rejection from the user, which led to a negative AT towards using digital signatures.

These six determinants are reflected in consumers’ attitudes. The initial determinant of a user’s BI towards a digital signature is AT, which is the consumers’ convictions causing proper behavior to obtain the benefit of technological adoption [37]. This has a significant positive effect in influencing consumers’ BI towards digital signatures, which was increased by 43.4% [18], subsequently indicating the proposal of the following hypothesis:

Hypothesis 1 (H1).

Attitude has a positive effect on influencing consumers’ intention to use digital signature systems.

2.1.2. Determinants of Attitude towards Usefulness of Digital Signature Information

The TPB theory is the most prevalent model used in verifying consumers’ adoption of technological innovation. This indicates that BI is controlled by three main constructs, namely, AT, SN, and PBC [38]. These three main constructs of TPB, incorporated with facilitating conditions in the UTAUT2 theory and information usefulness in the IACM model, were utilized to examine consumers’ adoption of digital signatures.

Besides the consumers’ behavior in using the digital signature system, AT towards information should also be considered, due to being a positive or negative individual feeling influenced by psychological factors and present situations. Based on technological adoption, this AT was found to influence consumers’ BI [39]. This indicated that attention and right AT led to perceived information [40], subsequently causing the proposal of the following hypotheses:

Hypothesis 2 (H2).

Attitude has a positive effect on influencing the usefulness of digital signature information.

2.1.3. Determinants of Information Acceptance to Use Digital Signature System

The first determinant of information acceptance towards using digital signatures is information quality (IQ), which is the quality of content reviewed by consumers from the perspective of data characteristics [41]. This indicates that the assessment of the IQ determines the level at which many consumers learn and adopt the data obtained to make suitable decisions [40]. Moreover, higher IQ was obtained from any related sources of reviewed information, with the perceived data usefulness also found to be more helpful. This indicated that IQ had a positive impact on the usefulness of technical data, leading to the prediction of consumers’ BI to adopt new technology [42].

The second determinant is information credibility (IC). This indicates that the information obtained by consumers becomes useful when acquired from reliable sources, when customers make an informed decision based on the derived data. Subsequently, IC is defined as reliable knowledge for the recipient of the data [43], due to its being a convincing initial factor [44]. This is positively related to IU towards data adoption and consumers’ BI in using technology [45].

The third determinant is the needs of information (NOI), which are based on the theories of “seeking advice” [46] and “opinion” [47], as well as consumers’ data requirements used to make very interesting decisions [40]. Based on previous studies, the addition of NOI was used as the dependent variable due to being useful in ensuring that the obtained consumer information was sufficiently insightful. It also influences consumers’ BI to use technology [45].

These three determinants are reflected by information usefulness (IU), which is the individual’s perception of related data on the internet [48]. Information usefulness variable is a positive effect in influencing information adoption by 63.5% [49], indicating that the IU in social media was significantly related to the utilization of the technology, as proposed in the following hypothesis:

Hypothesis 3 (H3).

Information usefulness has a positive effect on influencing the adoption of digital signature information.

2.1.4. Determinants of Information Adoption to Use Digital Signature System

Information adoption (IA) is directly influenced by IU, where social media users showed that an enormous amount of technical data influenced the consumers’ BI to use a technological system [45]. This indicates that IA has a positive effect in influencing consumers’ BI by 70.3% [49], showing that IA in social media subsequently had a positive relationship as proposed in the following hypothesis:

Hypothesis 4 (H4).

Information adoption has a positive effect on influencing consumers’ intention to use digital signature systems.

2.1.5. Determinants of Consumers’ Influence to Use Digital Signature System

Subjective norms (SN) are proposed to have a direct effect on digital signature adoption. It is defined as the level at which an individual perceives that critical communities need them to perform or abandon a specific behavior [37]. This indicates that the more individuals perceive that influential persons or groups (i.e., families, friends, and colleagues) should be involved in a specific behavior, the more they adhere to the attitudinal adoption. SN also has significant positive effects in influencing consumers’ BI towards using the digital signature service, due to the communication content factors. This indicated that increased consumers’ BI led to more utilization of the digital signature by 19.7% [18]. The result was similar to a previous study, which stated that SN had a positive correlation with the purchase intention of utilizing renewable technology [50]. Based on this condition, the following hypothesis was proposed:

Hypothesis 5 (H5).

Subjective norms have a positive effect on influencing consumers’ intention to use digital signature systems.

2.1.6. Determinants of Perceived Behavioral Control to Use Digital Signature System

This present study states that FC has influenced PBC due to the degree of the individuals’ beliefs. These are based on the available support and consumers’ perceptions of the existing technical infrastructure to gain convenience in using a technology system [51]. FC is also the level of an individual’s comfort while using a system supported by technical and organizational infrastructure [52]. According to the communication content factors, these conditions have a positive effect on the adoption of digital signatures. This indicates that increased consumers’ BI in FC leads to an elevated PBC by 19.3% [18], subsequently causing the proposal of the following hypothesis:

Hypothesis 6 (H6).

Facilitating conditions has a positive effect on influencing consumers’ perceived behavioral control to use digital signature systems.

The last determinant is perceived behavioral control (PBC), which is the individuals’ perception of ease or difficulty in performing a preferred consumer behavior [37]. This is the consumer’s perception of the resources and opportunities needed to control behavior. When individuals have adequate resources (i.e., time, money, technology) to highly ensure the usefulness of technology, they also have an increased level of PBC. Based on the communication content factors on intention, PBC had a positive effect on the adoption of digital signatures. This indicated that increasing consumers’ BI in self-efficacy elevated PBC by 80.3% [18], leading to the proposal of the following hypothesis:

Hypothesis 7 (H7).

Perceived behavioral control has a positive effect on influencing consumers’ intention to use digital signature systems.

3. Methodology

3.1. Determine Measurement Items

Based on using the digital signature system, the assessment of numerous latent variables was observed, such as privacy and security, performance expectancy, effort expectancy, habit, price value, hedonic motivation, attitude, subjective norms, facilitating conditions, perceived behavioral control, information quality, information credibility, needs of information, information usefulness, information adoption, and behavioral intention to use the digital signature system. These latent variables were directly unmeasured, leading to the collection of the measurement items from the analysis of each factor.

3.2. Questionnaire Design

The utilized questionnaire contained three sections as follows: (1) The first section briefly describes the study objectives and defines the digital signature system. It also describes the measurement scale guidelines used in this survey, i.e., the respondents were directed to answer each question through a five-point Likert scale. A scale of 1 represents ‘strong disagree’ until a scale of 5 represents ‘strong agree’. (2) The second section contains 12 questions to determine the demographic characteristics of respondents, such as gender, age, marital status, domicile area, education, occupation, and monthly income, as well as digital signature utilization frequency and platform, and (3) The third section contains 69 questions, based on assessing the factors involved in integrating UTAUT2, TPB, and IACM models, which are shown in the Appendix A.

To improve the reliability and consistency of the distributed questionnaires, the elimination of negative questions is found to be very useful. This was due to the negative questions leading to a misunderstanding of the survey contents [53]. This indicated that the negative questions were reduced in the questionnaire design, with the positive items containing unambiguous words and comprehensive sentences, to ease the understanding of the respondents.

3.3. Demographic of Respondents

The snowball sampling technique was also used in this study due to being a method for identifying, selecting, and obtaining participants in a continuous network or chain of relationships. In this technique, initial identification begins with criteria that should be adequately met by the potential respondents [54]. Based on the sampling outputs, most of the respondents were aged from 17–45 years old. The descriptive statistics of the 358 respondents who participated in answering the questionnaire. According to the outputs, a total of 358 respondents participated in this survey, with 53.91% and 46.09% being males and females, respectively. For the respondent’s age, the highest ratio was 40.22% (17–25 years), accompanied by 39.66% (26–35 years), 16.20% (36–45 years), and 3.91% (more than 45 years). Other characteristics were also analyzed, such as domicile area, education background, occupation, monthly income, and frequency of using a digital signature (see

Table 1. Descriptive statistics of respondents.

Characteristics	Category	Frequency (n = 358)	Proportion (%)
Gender	Male	193	53.91%
	Female	165	46.09%
Marital Status	Single	182	50.84%
	Married	176	49.16%
Age	17–25	144	40.22%
	26–35	142	39.66%
	36–45	58	16.20%
	Over 46	14	3.91%
Domicile Area	Sumatera	185	51.68%
	Java	141	39.39%
	Kalimantan	9	2.51%
	Sulawesi	19	5.31%
	Papua	4	1.12%
Education Background	High School	61	17.04%
	Diploma	60	16.76%
	Undergraduate	205	57.26%
	Postgraduate	32	8.94%
Occupation	Student	85	23.74%
	Teacher/Lecturer	22	6.15%
	Government Staff	19	5.31%
	Private Employee	206	57.54%
	Others	26	7.26%
Monthly Income(In IDR)	<5 million	54	15.08%
	5–15 million	89	24.86%
	16–25 million	186	51.96%
	>25 million	29	8.10%
Frequency of Using Digital Signature (per day)	1–5 times	35	9.78%
	6–10 times	40	11.17%
	11–15 times	151	42.18%
	More than 15 times	132	36.87%

).

Based on Table 1, the majority of respondents reported a daily frequency of using digital signatures during the COVID-19 pandemic in frequency 11–15 times (42.18%). The remaining reported daily frequency of using digital signatures during COVID-19 for frequency 1 until 5 times (9.78%), 6 until 10 times (11.17%), and more than 15 times (36.87%).

3.4. Structural Equation Modeling

A statistical modeling approach providing a comprehensive process to analyze study questions and variables is often used to evaluate social behavior sciences. This indicated that structural equation modeling (SEM) was used to test the hypothesis on the strength and direction of the relationship between the predictor and outcome variables [55]. To examine the proposed measurement tool, analyzing the validity of the SEM was necessary [18]. In this study, the goodness of fit index was also used to determine whether the conceptual model was accepted or rejected.

4. Results and Discussion

4.1. Measurement Model Analysis

A confirmatory factor analysis, including reliability and validity tests were conducted to evaluate the goodness of fit of the measurement model. Since no single index identified a good and weak model [56], the utilization of various indices was suggested, to confirm an acceptable fit [57,58].

Table 2. Confirmatory factor analysis results.

Constructs	Items	Initial Model			Final Model
		Factor Loadings	CR	AVE	Factor Loadings	CR	AVE
Privacy and Security (PS)	PS1	0.649	0.826	0.543	Deletion 1	0.808	0.584
	PS2	0.758			0.767
	PS3	0.778			0.799
	PS4	0.756			0.816
Performance Expectancy (PE)	PE1	0.808	0.879	0.644	0.810	0.879	0.644
	PE2	0.824			0.823
	PE3	0.755			0.756
	PE4	0.820			0.820
Effort Expectancy (EE)	EE1	0.704	0.892	0.623	0.703	0.892	0.623
	EE2	0.790			0.790
	EE3	0.818			0.818
	EE4	0.824			0.824
	EE5	0.805			0.806
Hedonic Motivation (HM)	HM1	0.824	0.893	0.675	0.824	0.893	0.675
	HM2	0.840			0.840
	HM3	0.853			0.852
	HM4	0.768			0.768
Price Value (PV)	PV1	0.751	0.859	0.603	0.749	0.859	0.603
	PV2	0.790			0.790
	PV3	0.813			0.813
	PV4	0.752			0.753
Habit (HT)	HT1	0.857	0.904	0.701	0.855	0.904	0.701
	HT2	0.842			0.841
	HT3	0.849			0.850
	HT4	0.801			0.802
Facilitating Conditions (FC)	FC1	0.736	0.852	0.591	0.736	0.852	0.591
	FC2	0.809			0.809
	FC3	0.803			0.803
	FC4	0.723			0.723
Attitude (AT)	AT1	0.785	0.901	0.533	0.784	0.880	0.595
	AT2	0.660			Deletion 3
	AT3	0.728			0.727
	AT4	0.799			0.791
	AT5	0.781			0.780
	AT6	0.765			0.766
	AT7	0.683			Deletion 4
	AT8	0.617			Deletion 2
Subjective Norms (SN)	SN1	0.860	0.895	0.682	0.860	0.895	0.682
	SN2	0.852			0.852
	SN3	0.834			0.834
	SN4	0.751			0.751
Perceived Behavioral Control (PBC)	PBC1	0.792	0.897	0.685	0.792	0.897	0.685
	PBC2	0.876			0.876
	PBC3	0.801			0.801
	PBC4	0.838			0.838
Information Quality (IQ)	IQ1	0.837	0.885	0.658	0.837	0.885	0.658
	IQ2	0.853			0.853
	IQ3	0.802			0.802
	IQ4	0.747			0.747
Information Credibility (IC)	IC1	0.872	0.919	0.738	0.872	0.919	0.738
	IC2	0.868			0.868
	IC3	0.843			0.843
	IC4	0.853			0.853
Needs of Information (NOI)	NOI1	0.807	0.866	0.619	0.807	0.866	0.619
	NOI2	0.796			0.796
	NOI3	0.817			0.817
	NOI4	0.724			0.724
Information Usefulness (IU)	IU1	0.801	0.853	0.592	0.802	0.840	0.636
	IU2	0.680			Deletion 5
	IU3	0.793			0.792
	IU4	0.798			0.798
Information Adoption (IA)	IA1	0.840	0.895	0.682	0.840	0.895	0.682
	IA2	0.850			0.850
	IA3	0.819			0.819
	IA4	0.791			0.791
Behavioral Intention to Use(BI)	BI1	0.817	0.905	0.703	0.817	0.905	0.703
	BI2	0.824			0.824
	BI3	0.864			0.864
	BI4	0.848			0.848

presents the standardized factor loadings, the average variance extracted (AVE), and composite reliability (CR) for each construct. Therefore, the measurement model was considered a good fit when the minimum standard value in each category was achieved.

As shown in Table 2, the final model shows that all standardized factor loadings were greater than or equal to 0.5, CR values were greater than or equal to 0.7, and the AVE values were greater than or equal to 0.5 [36,59]. These results indicate that all the observed variables have high reliability as latent variables.

Based on Table 2, the data deletion process was carried out with 5 factor loading values that were considered to be lower than the predetermined value. Furthermore, the deletion of data was in the range of values that meet the standards as in PS1, AT2, AT7, AT8, and IU2, and was conducted to obtain a good model fit by eliminating the lowest factor loadings, which changes the CR and AVE values.

4.2. Structural Model Analysis

Using the initial SEM, the results are presented in Figure 2, where the standard regression coefficients are observed in the path analysis, with interrelationships also found between the variables in the model. This indicated that the seven hypotheses were included in the proposed model.

Based on the initial conceptual model results, consumers’ attitude (AT) has the largest effect in influencing consumers’ intention to use digital signatures. However, IA had the smallest effect on consumers’ intention to use digital signatures (see Figure 3).

Based on the form of the proposed conceptual model, which is captured as a second-order model. The form of a second-order factor can embody the meaning of many first-order latent variables. Every first-order latent variable is applied as a reflective indicator of the second-order model. Therefore, the number of variables that need to be examined in the structural model can be reduced significantly.

The hypothesis outputs are presented in

Table 3. Hypothesis results.

Research Hypothesis	Paths	Path Coefficient	Standard Deviation	p-Values	Significance
H1	AT → BI	0.429	0.074	0.000 ***	Significant
H2	AT → IU	0.833	0.018	0.000 ***	Significant
H3	IU → IA	0.781	0.028	0.000 ***	Significant
H4	IA → BI	0.098	0.065	0.043 **	Significant
H5	SN → BI	0.116	0.046	0.000 ***	Significant
H6	FC → PBC	0.689	0.033	0.000 ***	Significant
H7	PBC → BI	0.364	0.063	0.000 ***	Significant

Note: *** Correlation is significant at ≤0.001 level (two-tailed); ** Significant level at ≤0.05 level (two-tailed).

, where the final conceptual model was obtained by eliminating the five factor loading values. According to Figure 3, the final result of the structural model was observed, where the factor loading values that were lower than 0.7 were deleted.

When the standard regression weights were observed, the p-values (p < 0.05) of the predicted hypotheses were examined. This was based on determining whether the relationships in the model were significant and in the desired direction. Therefore, p < 0.05 was considered to be significant. All relationships were examined to determine the p-values influenced by each construct. Furthermore, the results of the structural model also allow for mediating relationships between several constructs. To test for mediating effects, the researchers followed a two-tailed bootstrapping procedure (see

Table 4. Estimates of indirect paths.

Indirect Paths	Indirect Effect	95% Confidence Intervals a
		Lower Bound	Upper Bound
Attitude → Information Usefulness → Information Adoption → Behavioral Intention	0.041	0.074	0.176
Facilitating Conditions → Perceived Behavioral Control → Behavioral Intention	0.047	0.160	0.248

Note: ^a 95% confidence intervals obtained through Bootstrapping (1000 resamplings).

).

Based on the results presented in Table 4, the significance of the indirect effect was examined using the confidence intervals (CI) provided by the bootstrap resampling (1000 resamplings). The results indicated that Attitude was indirectly influences Behavioral Intention through Information Usefulness and Information Adoption (CI = 0.074–0.176). In addition, Facilitating Conditions were indirectly influences Behavioral Intention through Perceived Behavioral Control (CI = 0.160–0.248).

4.3. Discussion on the Results of Hypothesis Testing

In this study, the theoretical integration approach of UTAUT2, TPB, and IACM aimed to determine the constructs’ effects on consumers’ interest in digital signature adoption. Based on Figure 3, the TPB model was found to adequately predict the interest in using this system within Indonesia. Subsequently, these interests were estimated and influenced by AT, which is reflected in the UTAUT2 model, respectively. SN and PBC were also the other TPB-based factors that influenced users’ interest in using digital signatures. This indicated that the influence of PBC was due to the underlying beliefs of FC. Furthermore, the integrating UTAUT2, TPB, and IACM constructs predicted the consumers’ interest in using this system AT was found to have the largest significant effect in influencing consumers’ interest. However, the IACM-based IA was found to have the smallest effect in influencing consumers’ perspectives on digital signature adoption.

4.3.1. The Results of Hypothesis Testing Predictors of Attitude to Use Digital Signature

This study examined an integrated structural model of consumers’ attitudes, attitudes towards information, information usefulness, information adoption, subjective norms, facilitating conditions, and perceived behavioral control. The first hypothesis (H1) predicted that the consumers’ attitude (AT) had a positive effect (β = 0.429, p < 0.001) and a significant relationship (as a second-order factor) on influencing consumers’ intention to adopt a digital signature. This result is in line with [60] found that the attitude factor as a mediator has a positive and significant effect on influencing consumers’ intention to adopt a technology. In addition, privacy and security (PS) should still be considered due to the loss of control towards personal and organizational information as well as stolen documents without permission [60]. Furthermore, performance expectancy (PE) had a positive correlation reflected by AT, subsequently indicating its usefulness in assisting customers to quickly accomplish their work [61]. Effort expectancy (EE) had a positive correlation reflected by AT, indicating the perspective of most respondents, based on the effortless utilization of a digital signature. The conveniences and features provided by the digital signature services also influenced consumers’ interest in technology adoption [61]. For hedonic motivation (HM) were observed, where users were primarily determining the novelty to acquire effectiveness from the utilized innovation technology products [61]. According to price value (PV), we need to consider that reflected from AT, where consumers perceived that subscribing price had very high benefits, although most of them often searched for cheap deals such as discounts while planning to activate the digital signature services [61]. For habit (HT) showed that consumers automatically used a digital signature when they need to approve or validate the documents during work from home [61].

4.3.2. The Results of Hypothesis Testing Predictors of Information Usefulness

The second hypothesis (H2) predicted that the consumers’ attitude (AT) had a positive effect (β = 0.833, p < 0.001) and a significant relationship towards information usefulness (IU) in influencing consumers’ behaviors towards using a digital signature. This result is contradicted with [45], which showed that the AT towards information was not influential on IU. Furthermore, the third hypothesis (H3) predicted that the information usefulness (IU) had a positive effect (β = 0.781, p < 0.001) and significant relationship (as a second-order factor) was observed with influencing information adoption (IA). This result is in line with [45], which found that IU had a positive and significant relationship on influencing consumers’ intention towards using a digital signature. Furthermore, the fourth hypothesis (H4) predicted that the information adoption (IA) had a positive effect (β = 0.098, p < 0.05) but a significant relationship in influencing consumers’ intention towards using a digital signature. This result is in line with [45], which showed that IA had a positive and a significant effect in influencing consumers’ intention to adopt a technology. Moreover, people often obtain information directly from their friends and references, or through social media. This was due to the assumption that the obtained information was helpful towards adopting digital signatures.

As for information quality (IQ), it had a positive correlation reflected by information usefulness (IU). These results indicated that consumers always performed information reviews online, in newspapers, or in technology magazines, to determine qualified information before adopting a digital signature [41]. Information credibility (IC) also needs to be considered for all received information related to digital signature information. Due to all online comments not being considered as credible information, the accuracy of online or offline data reviews did not affect the IU of digital signatures [45]. Furthermore, as for needs of information (NOI) from consumers’ perspectives, there was a positive correlation of reflected from IU, where interactions of digital signature providers’ fan pages on social media and legal platforms encouraged consumers to meet the NOI by using a digital signature [40].

4.3.3. The Results of Hypothesis Testing Predictors of SN to Use Digital Signature

The fifth hypothesis (H5) predicted that the subjective norms (SN) had a positive effect (β = 0.116, p < 0.001) and a significant relationship was observed in influencing consumers’ behavioral intention towards using a digital signature. This result is in line with [40], which showed that the opinions of salient references, such as family, friends, and advisors, increased consumers’ beliefs on technological adoption. In addition, information sharing on digital signature benefits greatly influenced consumers’ behavioral intentions towards using a digital signature.

4.3.4. The Results of Hypothesis Testing Predictors of PBC to Use Digital Signature

The sixth hypothesis (H6) predicted that the underlying belief structure known as the facilitating condition (FC) had a significant positive relationship in influencing PBC (β = 0.689, p < 0.001). This result is in line with [62], which showed that the availability of facility resources (i.e., customer service, technical support, and others) increased the adaptability of consumers using technology. Technical support, such as education and problem-solving techniques, also improved the belief that the system was more reliable and trustworthy.

The seventh hypothesis (H7) predicted that the perceived behavioral control (PBC) has a positive effect (β = 0.364, p < 0.001) and a significant relationship in influencing consumers’ behavioral intention towards using a digital signature. This result is in line with [36], which showed that individuals perceived extensive resources containing internal (i.e., skills, knowledge) and external (i.e., period, chance) causes to ensure the increased value of the digital signature. Consumers also had a more elevated standard of PBC, indicating that the concept affected the attitudinal intention to use digital signatures.

5. Conclusions

This study aimed to investigate the acceptance of the digital signature system during the COVID-19 pandemic in Indonesia. By integrating the UTAUT2 with the additional factor, the conceptual model presented and examined was privacy and security (PS). Furthermore, the expanded model provided a more exhaustive capture of digital signature adoption factors, with the results showing that the UTAUT2 constructs consisted of PS, PE, EE, HM, PV, and HT, which were reflected by AT as a second-order factor. This subsequently indicated that the IACM constructs also influenced digital signature adoption, where IQ, IC, and NOI were reflected by IU as second-order factors. Furthermore, the TPB constructs influenced digital signature adoption (i.e., SN, AT, and PBC, with the underlying construct being FC), with AT, SN, and PBC playing important roles in sustainable digital signature adaptability. However, IA had the smallest impact on the sustainable adoption of digital signatures.

The result reveals that FC-based PBC plays an important factor in primarily facilitating consumers during the current COVID-19 pandemic. This showed the importance of fully supporting the facilities used by consumers, such as technical support and customer service, and continuously adopting digital signatures during the COVID-19 pandemic.

5.1. Implications for Theory

This study provided new insight into the development model of the theoretical contribution to the social analysis towards consumers’ BI to use digital signatures during the COVID-19 pandemic in Indonesia. This was carried out by applying the following three integrated conceptual models: UTAUT2, TPB, and IACM. The development model was also adopted from [36] and integrated with the IACM technique developed by [45]. This was to measure the information acceptance of individual consumers towards using digital signatures. The integration of three theories was also developed for several reasons. Firstly, important aspects relevant to consumer interest were theoretically captured in using digital signatures. This indicates that technological providers should consider several significant factors affecting consumer interest, based on adopting digital signatures. Secondly, the models were theoretically extended by integrating three theories to understand whether individual analysis and IA significantly affected consumers’ interest in using digital signatures.

Several contributions were also provided to the theoretical development of the novel and innovative digital signature. In addition, the range of factors influencing the actual use of this technology was expanded by integrating three theories, namely, the UTAUT2, TPB, and IACM, due to the nonexistence of this combination in the previous studies, especially the form of second-order factor that can embody the meaning of many first-order latent variables of these integrated three theories. According to the results, this theoretical model provided a clear analysis of the factors affecting the sustainable adoption of digital signatures.

5.2. Implications for Practice

A digital signature is a technology mostly used in society during the COVID-19 pandemic in Indonesia. This study provided the technology providers with adequate knowledge on the influential factors of consumers’ interest towards a sustainable digital signature adoption. Based on the results, avail value was the most influential factor in digital signature adoption. These should be used to increase the number of people futuristically adopting digital signatures, although several aspects still need to be considered. Digital signature providers should seriously consider the user’s experience while designing and developing the systems. They should also instill a mindset as part of the users’ lifestyle, specifically during the pandemic, to cultivate a new habit for safe document approval and validation through online transactions and marketing techniques. To promote engagement towards sustainable adoption, a digital signature should be updating and providing educational information on their platform and fan page as sources of advertisements to potential users.

The significant positive impact of AT also had important implications for improving the benefits of digital signature utilization. Moreover, complementing the existing facilities in the development and marketing of digital signature platforms improved the consumers’ interest, with the significant effect of AT highlighting the importance of building a simple system without operational difficulties. This should be created through a technology development that effectively updates the digital signature system used by consumers.

5.3. Limitations and Future Research Directions

Besides the advantages, this study also has the following few limitations: (1) the results are only applicable in Indonesia, as they are likely not be implemented in different nations or societies due to political, social, environmental, technological, legal, and economic backgrounds. Therefore, future studies should conduct a correlation analysis to validate the outcomes towards consumers’ interest in continuously adopting digital signatures, (2) the selected factors did not capture all the variables influencing the consumers’ BI to use a digital signature in the country. This indicates that subsequent future studies should examine other factors such as trust and risk perspectives towards technological adoption, and (3) this is based on the analysis of human behavior factors, which were not thoroughly evaluated towards the adoption of digital signatures. Therefore, subsequent future studies should analyze the human behavior variables affecting customer’s interests within specified institutions or company agencies (i.e., banking, manufacturing and non-manufacturing industries, government institutions, and others) towards using a digital signature. This study should be futuristically conducted by using T-O-E (technology-organization-environment), UTAUT (the unified theory of acceptance and use of technology), TAM (technology acceptance model), or a combination of all these theories, to obtain valid and accurate results in analyzing human behavior towards the sustainable adoption of digital signatures.