INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Determinants of Electronic Tax Invoice Adaption among SMEs in  
Embakasi Sub-County, Nairobi County  
Risper Ouso, Prof Philip Bii and Peter Mwai  
Bomet University College  
DOI: https://dx.doi.org/10.47772/IJRISS.2025.905000128  
Received: 27 April 2025; Accepted: 02 May 2025; Published: 03 June 2025  
ABSTRACT  
The purpose of this study was to examine the effect of perceived benefit, perceived security, and relational trust  
on the adoption of electronic tax invoices among small and medium enterprises (SMEs) in Kenya. The study  
was guided by the Innovation Diffusion Theory, the Technology Acceptance and Use of Technology Theory, and  
the Standard Economic Theory. An explanatory research design was adopted, targeting a population of 859 SMEs  
operating in Embakasi Sub-County. A sample size of 273 SMEs was selected using stratified and simple random  
sampling techniques. Primary data were collected through semi-structured questionnaires and analyzed using  
descriptive statistics and inferential methods, including regression analysis. The results showed that perceived  
benefit, perceived security, and relational trust all had significant and positive effects on the adoption of  
electronic tax invoices. The study concludes that technological readiness among SMEs enhances the positive  
influence of these factors on e-tax invoice adoption. These findings support the Technology Acceptance and Use  
Theory by demonstrating the importance of intention to use technology, and align with the Innovation Diffusion  
Theory by showing how perceived advantages, trust, and security drive adoption. The study recommends that  
SME management prioritize these factors when implementing e-tax invoicing, while policymakers should  
develop supportive legal frameworks and invest in infrastructure and human capacity to facilitate technology  
adoption.  
Key words: perceived benefit, perceived security, rational trust, electronic tax invoice  
INTRODUCTION  
The discourse on the adoption of e-invoices, particularly in developing countries, has gained traction among  
scholars, practitioners and regulators. Recently, tax authorities in the world have been under pressure to adopt  
advanced information technology over the internet into their tax system processes (Shim & Song, 2016).  
However, the tax authorities have faced significant challenges in implementing such technologies because  
taxpayers do not feel comfortable with readily capturing their incomes by the system. Despite this attitude, most  
countries especially in the developing contexts have realized the need to implement key information technologies  
such as electronic invoicing (e-invoicing) with an aim to raise revenue that largely supports government  
activities.  
Electronic invoicing is an information system service that is used to collect transaction information and transfer  
that information via the internet (Hernandez-Ortega, 2012). With the rapid introduction of e-commerce and the  
digital economy, e-invoicing plays a key role in maintaining business information throughout the entire supply  
chain (Chang et al., 2013). With regard to the tax administration, electronic tax invoice is defined as a type of  
invoice that is created and transmitted electronically, using technology such as computers, software, and the  
internet (Cuong, Thao & Linh, 2023). Cuong, Thao & Linh, (2023) further argued that adopting electronic tax  
invoicing increases operations efficiency, reduce costs, improve tax compliance and enhance their overall  
Page 1649  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
financial performance. Similarly, e-tax invoice adoption will help the tax authorities to strengthen the inspection  
of tax declarations, tax payments, and tax refunds of firms (Nguyen et al., 2020). Thus, e-tax invoicing is a  
critical research area that needs an in-depth understanding.  
A review of the previous work shows that adoption of e-tax invoice is largely influenced by varied factors. One  
stream of scholars in Nigerian small firms argued that adoption of e-tax invoice is a function of the technological  
infrastructure, perceived utility, perceived simplicity of use, perceived risk, government backing, and tax  
authority enforcement (Nwokah et al., 2019). In Tanzania, the success of the e-tax invoicing is attributed to  
technology infrastructure, access to training and support, sound government policies and regulations as well as  
stakeholder involvements such as tax authorities and business associations (Komba & Lema, 2019). Further, a  
study in Korea shows that identified that establishing a free website for small-sized taxpayers and providing  
various issuing systems for diverse taxpayers are critical for its success (Shim & Song, 2016).  
Evident from the other research studies, no study particularly from an emerging economy, that has examined the  
determinants of e-tax invoicing. Previous work has focused on the effect of the e-tax invoice on outcomes such  
as tax compliance processes, and transparency of business transactions and taxpayer services (Lee, 2016; Susilo,  
2022). The closest work sought to establish factors influencing e-invoice adoption (Qi & Che Azmi, 2021), but  
did not examine specifically on the determinants of e-tax invoice and whether technology readiness as a potential  
moderator in the relationship. Thus, the purpose of the current study is twofold; the first objective is to examine  
the factors that influence the adoption of electronic tax invoices, and the second objective is to examine the  
moderating role technology readiness.  
Lee (2016) showed that electronic tax invoices provide tax authorities with powerful tools to integrate the tax  
information provided by taxpayers with effective, transparent and trusted services and ultimately enhance the  
tax management process overall. Thus, companies are more likely to adopt e-tax invoices in the management of  
tax. Fu et al. (2006) also suggested that people may not consider adopting a system if they perceive that the e-  
government system lacks good security features. Thus, if the company perceives that the process of adopting  
electronic tax invoice offers better security, then that it will be more willing to adopt it. In addition, Hart and  
Saunders (1997) believed that the adoption of electronic tax invoicing requires resources and changes in each  
organization’s routine processes. Hence, if a company has a high degree of trust in e-government, that company  
will be more willing to adopt electronic tax invoicing to automate its business processes with tax authorities.  
Even though these factors have been empirically tested, the results are inconclusive and inconsistent; hence  
further probe is inevitable especially in developing contexts with different states of technology.  
With the adoption of e-tax invoice, instant access and processing of digital information by tax authorities can  
increase the vigilance of taxpayers for the possibility of auditing, thereby reducing non-compliance for tax  
returns and improving tax refund requirements (Qi & Che Azmi, 2021). The main source of revenue for the  
government's development initiatives and recurrent expenditure is taxation. KRA's overall revenue performance  
was Kshs. 4,849.3 billion compared to the 7th Corporate Plan objective of Kshs. 4,899.3 billion, resulting in a  
Kshs. 50 billion shortfalls. Small and medium-sized businesses (SMEs) have low tax compliance rates, and KRA  
has had trouble collecting VAT. Countries around the world are leveraging technology to increase revenue  
collection, decrease fraud, and improve taxation efficiency, so is Kenya. Through TIMS, KRA will be able to  
facilitate the electronic tax invoice management by; standardization, validation and transformation of invoices  
to KRA on a real time or near real time basis. The Electronic Tax Invoice Management System (TIMS) was  
supposed to go into effect on August 1, 2022, but because so few taxpayers, especially small and medium-sized  
enterprises, have embraced TIMS, the deadline has already been pushed back three times.  
Some global adopters of Electronic Tax Registers include Rwanda, Hungary, Belgium, Canada, Sweden and  
Austria. A fiscal control unit was installed with electronic cash registers in Hungary. In the affected sectors, VAT  
income rose by 15% following the first year of operation (Deloitte, August 2019). In 2017, Rwanda introduced  
electronic tax registers and since then, Rwanda has seen a 48% increase in VAT collections. By December 2022,  
94% of major enterprises—those with yearly sales of over Sh1.3 billion—had acquired electronic invoices and  
linked with the Tax Invoice Management System (TIMS), the authority's system, while 82% of medium-sized  
businesses had already attained compliance, according to KRA. The Kenya Revenue Authority (KRA)  
anticipated a revenue shortfall, as evidenced by the fact that only 43% of Kenya's SMEs with an annual turnover  
Page 1650  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
of less than 200 million were compliant with the new internet-enabled tax registers (ETRs). This ultimately cost  
the taxman billions of shillings in value-added tax. There is not much information available about the TIMS's  
adoption in Kenya or the variables that could affect how it is implemented. Based on this evaluation, there is a  
gap in literature hence this study intends to find out the answer to the research question that; will perceived  
benefit, perceived security and technology readiness affect the successful adoption of the TIMS?  
LITERATURE REVIEW  
Perceived Benefit and Electronic Tax Invoice Adoption  
In both developed and developing nations, e-tax systems are now a crucial component of tax administration.  
Nonetheless, small and medium-sized businesses (SMEs) in developing nations continue to implement e-tax  
systems at a relatively low rate. In their 2018 study, Shah, Jaffari, and Hussain examined the variables of  
perceived utility, perceived ease of use, perceived trustworthiness, and perceived risk as the main drivers of  
SMEs' inclination to implement e-tax in Pakistan. According to the findings of a survey performed on 349 SMEs,  
these four variables significantly increased the intention of SMEs to implement e-taxation. To be more precise,  
perceived utility was determined to be the most significant factor, and was followed by perceived danger,  
perceived credibility, and perceived simplicity of use. The study also discovered that perceived utility acted as a  
mediating factor between perceived ease of use and e-tax adoption intention. The results of this study offer  
guidance to developing country tax authorities and policymakers on how to create e-tax systems that are  
practical, believable, and easy to use in order to promote adoption among SMEs.  
Wahyuni and Suhardi (2020) looked into how small and medium-sized businesses (SMEs) in Indonesia adopted  
e-tax systems and the importance of perceived security, usefulness, and convenience of use. The intention to  
implement e-tax systems was significantly positively impacted by all three elements, according to the results of  
a survey that was completed by 120 SMEs. More specifically, perceived security and simplicity of use were  
revealed to be secondary factors, with perceived utility emerging as the most significant influence. The study  
also discovered that perceived utility acted as a partial mediating factor between perceived ease of use and the  
intention to adopt e-tax systems. The results of this study offer guidance to Indonesian tax authorities and  
policymakers on how to create secure, practical, and user-friendly e-tax systems that will promote adoption  
among SMEs.  
The adoption of electronic tax filing, or e-filing, by small and medium-sized firms (SMEs) in South Korea was  
examined by Lee and Kim (2019). According to the findings of a study given to 219 SMEs, adoption intention  
for e-filing was significantly influenced by perceived utility and convenience of use. Perceived compatibility,  
perceived trust, and perceived government assistance were also found to have favorable effects on the intention  
to embrace electronic filings. The study also discovered that perceived usefulness acted as a partial mediating  
factor in the association between perceived ease of use and e-filing adoption intention. The results of this study  
offer guidance to South Korean tax authorities and policymakers on how to create e-filing systems that are user-  
friendly, practical, reliable, and compatible to promote adoption among SMEs.  
Electronic tax (e-tax) is an electronic filing and payment system for taxes. The Malaysian government has  
implemented the e-tax system to increase efficiency, effectiveness, and transparency in tax administration. Small  
and medium-sized businesses (SMEs) in Malaysia have yet to embrace e-tax, nonetheless. The conceptual model  
proposed in this study by Rosli, Yusof, Ahmad, and Shafie (2018) investigated the factors that influence  
Malaysian SMEs' adoption of e-taxation. Perceived utility, perceived ease of use, perceived danger, perceived  
trust, and perceived government backing are the five components that make up the model. The moderating  
influence of organizational size on the connection between perceived usefulness and adoption of e-tax is also  
included in the model. The conceptual model offers a framework for upcoming empirical studies to investigate  
the variables affecting Malaysian SMEs' adoption of e-tax. The findings of this study may help tax authorities  
and policymakers in Malaysia to design e-tax systems that are user-friendly, secure, and supported to encourage  
adoption among SMEs.  
HO1: Perceived benefit has significant effect on the adaption to electronic tax invoice  
Page 1651  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Perceived Security and Electronic Tax Invoice Adoption  
Electronic tax filing (ETF) use has grown in importance for small and medium-sized businesses (SMEs). But  
little is known about the factors that influence SMEs' propensity to use exchange-traded funds. Shen, Wu, and  
Lin (2020) investigated the variables influencing SMEs' propensity for ETF adoption. The trust-based model  
and the technology acceptance model (TAM) served as the foundation for the development of the research model.  
A total of 230 valid responses were gathered from a survey that was administered to SMEs in Taiwan. The  
findings indicate that the desire of SMEs to adopt ETF is highly influenced by perceived ease of use (PEOU),  
perceived usefulness (PU), perceived risk, and trust. Moreover, it was discovered that the association between  
PEOU, PU, perceived risk, and ETF intention was mediated by trust. This study helps SMEs understand the  
elements influencing their decision-making processes when considering ETF adoption, and it offers useful  
insights for policymakers and tax authorities to improve SMEs' adoption of ETFs.  
The relationship between small and medium-sized businesses' (SMEs) perceptions of security and trust regarding  
the adoption of electronic tax invoicing (ETI) by SMEs in Saudi Arabia was investigated by Almutairi, Qureshi,  
and Baabdullah (2019). A sample of 175 Saudi Arabian SMEs was chosen for the study using the stratified  
random selection technique. Structured survey instruments were used to gather data, and structural equation  
modeling was used to evaluate the results. According to the study's findings, Saudi Arabian SMEs' adoption of  
ETI is favorably correlated with their perception of security. The study also discovered that the relationship  
between SMEs' adoption of ETI and their perception of security is mediated by trust. Additionally, the findings  
show that the relationship between perceived security and trust regarding SMEs' adoption of ETI is moderated  
by the perceived usefulness of ETI. The results of this study can aid in the development of strategies to promote  
ETI adoption among SMEs and aid policymakers and tax authorities in Saudi Arabia in better understanding the  
factors influencing SMEs' adoption of ETI.  
Park, Yoon, and Park (2020) conducted research to find out what aspects small and medium-sized businesses  
(SMEs) in South Korea consider important when deciding whether to implement electronic tax invoice (e-Tax)  
systems. To accomplish this goal, a survey of 326 SMEs was carried out and a research model incorporating the  
diffusion of innovation (DOI) theory, and the technological acceptance model (TAM) was created. The findings  
demonstrated the strong influence that compatibility, observability, perceived usefulness, and perceived  
simplicity of use had on SMEs' adoption of e-Tax systems. Relative advantage and complexity, however, were  
shown to have no appreciable impact. Additionally, the results demonstrated that adoption was significantly  
impacted directly by perceived usefulness, and indirectly by perceived compatibility, convenience of use, and  
observability. Considering the findings, several suggestions are made to legislators and vendors of electronic tax  
systems to promote SMEs' use of electronic tax systems.  
HO2: Perceived security has significant effect on the adaption to electronic tax invoice  
Relational Trust and Electronic Tax Invoice Adoption  
It is widely acknowledged that concerns about justice and legitimacy have an impact on confidence in authorities  
in a number of contexts, including: (i) voluntary adherence to their standards and demands (Levi, 1998; Levi  
and Stoker, 2000; Rothstein, 2005, 2009; Rothstein and Teorell 2008); (ii) tax compliance (Frey and Feld, 2002;  
Murphy, 2004; Alm and Martinez-Vazquez, 2007; Feld and Frey, 2007); (iii) law enforcement (Sunshine and  
Tyler, 2003); and (iv) collaboration with authorities (De Cremer and Tyler, 2007).  
In the same manner as consuming commodities and services raises human value, economists have begun to  
investigate whether procedural justice is a source of satisfaction for individuals (Frey et al., 2004; Frey and  
Stutzer, 2005). The experimental finding that granting taxpayers the right to vote increases tax compliance may  
be explained by their need for justice (Alm et al., 1993, 1999; Feld and Tyran, 2002; Wahl et al. 2010). We must  
acknowledge that people most likely do not view government institutions as a single, monolithic entity to  
comprehend where trust is formed. When examining whether people trust government institutions equally,  
Rothstein and Stolle's (2008) factor analysis found that there are two distinct aspects of institutional trust. The  
government, political parties, and parliaments are examples of input-side institutions.  
Page 1652  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
On the other hand, the police, the army, and legal institutions are examples of output-side institutions, and they  
load on a different dimension. The results of a second research that used Sweden's SOM poll were comparable.  
The slippery slope framework was established by Kirchler (2007) and Kirchler, Hoelzl, and Wahl (2008) to  
ascertain the connection between tax authorities and taxpayers. This concept focuses on trust and power as  
potential determinants of tax compliance behavior. The ability of tax authorities to regulate taxpayer conduct  
through penalties and tax audits is referred to as power. Legitimate power and coercive power are the two  
categories of power (Turner 2005; Kastlunger, Lozza, Kirchler, & Schabmann 2013). Conversely, trust refers to  
the trust-based relationship that exists between tax authorities and taxpayers.  
According to Kastlunger et al. (2013), there is a perception of increased compliance when taxpayers have a high  
degree of trust in tax authorities. According to research by Wahl, Kastlunger, Kirchler (2010) and Kastlunger et  
al. (2013), trust and power may have an impact on tax compliance. The results of this research indicated that tax  
compliance would be improved by a high degree of faith in tax authority and power. According to Kirchler's  
2007 classification, trust and power are positively and significantly correlated with increased tax compliance.  
He explained the differences between compulsory and voluntary tax compliance. There is an assumption that  
several circumstances influence each of these compliances. While tax authorities' powers are thought to have an  
impact on enforced tax compliance, taxpayers' trust in them is thought to influence voluntary tax compliance. It  
is well known that most of the research on electronic tax registers has been done in European nations like the  
Czech Republic, Italy, Russia, and Austria.  
Mas'ud, Abd Manaf, and Saad (2015) conducted a slippery slope study in African nations. In the meantime,  
Bukhari (2010) employed this paradigm in her research in Malaysia, where it was investigated how trust affected  
voluntary tax compliance. Because of this, Kastlunger et al. (2013) suggested that slippery slope research be  
done in regions of the world with distinct sociocultural demographics than Europe. As a result, this study will  
employ the slippery slope method to close the gap and carry out research outside of Europe to look at the tax  
compliance practices of Asian nations, particularly Malaysia.  
HO3: Relational trust has significant effect on the adaption to electronic tax invoice  
Research Methods  
Population and sample size  
The study targeted registered SMEs operating within Embakasi Sub- County in Nairobi County. Currently, there  
are 859 SMEs spread across different industries (KRA, 2022). These firms were appropriate for the study because  
of their extensive use of information technology which has intensified introduction and use of electronic tax  
system in tax administration (Nwabachili et al., 2024; Awai & Oboh, 2020), and elaborate legal framework for  
SMEs support in tax management (Okunogbe & Santoro, 2023). This region was specifically chosen due to its  
robust representation of SMEs across various industries. Embakasi was deemed appropriate for the study due to  
the significant adoption and integration of information technology within its SMEs. This technological  
inclination has made the region a prime area for the implementation and utilization of electronic tax systems, an  
essential element in modern tax administration. Therefore, the target population of the study was demonstrated  
in the Table below.  
Table 1: Target population  
Sector of economy  
Tourism and Hospitality  
Retailer and Wholesalers  
Jua Kali  
Population  
103  
Percentage %  
11.9  
36.9  
26.1  
12.2  
317  
224  
Agriculture  
105  
Page 1653  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Transport and Telecommunications  
110  
12.8  
Total  
859  
100  
Source, Author (2025)  
The sampling strategy involves utilizing part of the population to make inferences about the whole population  
(Zikmund et al., 2013). According to Kothari (2008), sampling is the process of choosing a small number of  
objects (a sample) from a larger group (the population) in order to estimate or forecast the prevalence of an  
unknown circumstance, event, or piece of information pertaining to the larger group. The sample size was  
determined from SMEs stratified based on the industry they operate in: Tourism and Hospitality, Retail and  
Wholesale, Jua Kali, Agriculture, as well as Transport and Telecommunications. This approach was deemed  
appropriate owing to its ability to ensure greater representation across the population of interest, register a small  
sampling error and give a better precision of estimation (Kemper et al., 2003).  
From the target population of 859 SMEs in Embakasi Sub-County, a sample size of 273 enterprises was obtained  
using the formula developed by Yamane (1973). Although the sampling frame is relatively low, the sampling  
strategy is economical and provides a representation of the total population (Acharya et al., 2013). The sample  
size formula was expressed as follows:  
푁
푛 =  
1 + 푁(푒)2  
Where: n = Sample size; N = Total population size; e = the error of Sampling  
The sample size was determined based on the sampling error of on 0.05 as follows:  
859  
푛 =  
= 273 SMEs  
2
1+859(0.05)  
The Neyman allocation formula (1934) was used to select the enterprises. With Neyman allocation, the best  
sample size for stratum h would be:  
푁ℎ  
푛ℎ = (  
) 푛  
푛
Where: nh - The sample size for stratum h; n - Total sample size; Nh - The population size for stratum h; N -  
The total population. Hence, distributions were as follows.  
Table 2: Population distribution  
Stratum  
Population  
Sample Size Distribution  
푁ℎ  
푛ℎ = (  
) 푛  
푛
Tourism and hospitality  
Retailer and wholesalers  
Jua Kali  
103  
317  
224  
105  
110  
859  
32.7  
100.7  
71.2  
33.4  
35  
Agriculture  
Transport Telecommunications  
Total  
273  
Source, Author (2025)  
Page 1654  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Data collection instruments  
Primary data involves the collection of raw data (Wahyuni, 2012). Primary data enables researchers to learn  
something new which may be confirmed by others and to eliminate own biases in the process (Driscoll, 2011).  
The study collected primary data on the adoption of e-tax invoice and its determinants as well as the moderating  
effect of technology readiness in the relationship. The list of the licensed small and medium enterprises in Nairobi  
County, Embakasi Sub-County was obtained from the County government of Nairobi directory of enterprises  
and Kenya Revenue Authority (KRA) for the period ending 2022 and mid-2023 respectively.  
Closed-ended questions were specifically chosen for their ability to direct respondents toward answers that align  
with the study's objectives, ensuring consistency and reliability in the data collected (Saunders et al., 2007). The  
survey items were derived from established, validated scales to maintain research rigor but were carefully  
modified to reflect the context and nuances of the study's focus. To quantify responses effectively, the  
questionnaire employed a five-point Likert scale, ranging from 1 ("Strongly Disagree") to 5 ("Strongly Agree").  
This scale allowed respondents to express varying degrees of agreement or neutrality concerning the survey  
items.  
The distribution of questionnaires targeted owners of the selected SMEs that had implemented the e-tax invoicing  
system since its introduction. This focus ensured that the participants had direct experience with the system and  
could provide informed responses, enhancing the relevance and applicability of the findings. The structured  
design and targeted approach of the questionnaires contributed to the collection of reliable, objective, and  
meaningful data for the study.  
Measurement of variables  
Table 3: Measurement of variables  
Variables  
Dimensions  
Scale  
Level  
Independent  
Perceived benefit  
Perceived Security  
Relational Trust  
Firm Age  
5-point Likert scale  
5-point Likert scale  
5-point Likert scale  
5-point Likert scale  
5-point Likert scale  
5-point Likert scale  
Nominal  
Nominal  
Nominal  
Control Variables  
Firm Size  
Dependent Variable  
Electronic Tax Invoice Adoption  
Nominal  
Source, Author (2025)  
Regression model  
The main regression model used to test the hypotheses was as shown as follows:  
ETA = β0 + C+ β1PB + β2PS + β3RT + ε  
Where, ETA = Electronic Tax Adoption; β0 = constant term or intercept; C = Control variables; β1 to β3 = are  
the coefficients of the independent variables; PB – Perceived benefit; PS – Perceived security; RT – relational  
trust; ε = error term  
FINDINGS AND DISCUSSIONS  
Response rate  
Response rate in moot research is a crucial pointer of participant engagement and data quality. It's calculated by  
Page 1655  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
dividing the number of complete survey responses by the total number of requested participants. A higher  
response rate enhances the study's validity and generalizability. According to a recent study by Sheehan (2020),  
a response rate of 60% or more is typically thought to be suitable for guaranteeing accurate results. Nevertheless,  
the context, research design, and target population should be considered, as lower response rates might still yield  
meaningful insights in specific cases according to Göritz (2020). The survey of the study as shown in Figure 4.1  
indicates a response rate of 89%. Since the respondents were 224 out of 273 that were invited  
Figure 1: Response Rate  
Reliability and validity results  
Table 4 shows results for the reliability test where electronic tax invoice adoption α = 0.765>0.70, perceived  
benefit α =0.901, perceived securityα =0.899 and relational trust α =0.911. This concludes that the questionnaire  
replies are reliable, since all alpha results are >0.70.  
Table 4: Reliability Tests  
Variable  
Cronbach's Alpha  
N of Items  
Electronic Tax Invoice Adoption  
Perceived Benefit  
Perceived Security  
Relational Trust  
0.765  
0.901  
0.899  
0.911  
4
5
5
5
Source, Author (2025)  
Factor Analysis  
Factor analysis works on the assumption that measurable and observable variables in the research can be reduced  
to smaller underlying variables which share common variance and are unobservable (Bartholomew et al., 2011).  
Therefore, in the present study, factor analysis was carried out in order to lessen and summarize the set of data  
items into controllable factors without losing the initial information. The exploratory factor analysis was used  
Page 1656  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
because of its ability to examine data to identify a reasonable and specific set of factors that can be studied  
together (Fricker et al., 2012). EFA was therefore performed to determine the accuracy and validity of the  
reflective multi-item scales (Bagozzi & Yi, 1988). Equally, the principal component analysis (PCA) with  
Varimax rotation was performed. Factor loadings were produced for all items to assess the validity of the  
construct.  
Prior to the collection of factors, a range of measures were used to determine the suitability of the respondent  
data for factor analysis. The measures used to evaluate the appropriateness of the data included the Kaiser –  
Meyer – Olkin (KMO) sampling adequacy assessment and the Bartlett Sphericity test. In order for the data to be  
appropriate for factor analysis, the required value for KMO is .50 and the Bartlett Sphericity Test should be  
significant (p ˂.05) (Hair et al., 2010; Tabachnick & Fidell, 2007). The convergent validity of the research  
instruments was tested using Average Variance Extracted (AVE) for all variables (Hair et al., 2014).  
The Kaiser-Meyer-Olkin (KMO) test is a measure that quantifies the adequacy of data for performing factor  
analysis according to Kaiser & Meyer, (1974). It evaluates the pattern of correlations among variables and  
provides a score ranging from 0 to 1. A KMO score above 0.5 is generally deemed acceptable for factor analysis.  
This is because it indicates a sufficient level of intercorrelations among the variables. According to Bartlett  
(1954) The Bartletts test is a statistical procedure that examines the assumption of equal correlations across  
variables. If the test rejects the null hypothesis of correlation matrix homogeneity, it suggests that the variables  
are interrelated to a degree that warrants their use in factor analysis. If p Bartlett's test is appropriate when the  
significance value is less than 0.05.  
Table 5 indicates the results of the Kaiser-Meyer-Olkin (KMO) Measure of Sampling Adequacy and Bartlett’s  
test suggest that the data is suitable for factor analysis. The KMO values for Perceived Benefit (0.772), Perceived  
Security (0.685), Relational Trust (0.775) and Electronic Tax Invoice Adoption (0.901) range from acceptable to  
excellent, indicating a good degree of coherence between variables. Furthermore, the Bartlett’s test values for  
all variables are less than 0.05, suggesting a significant level of interrelatedness among the variables. Therefore,  
it can be concluded that all variables are appropriate for factor analysis.  
Table 5: KMO & Bartlett’s test  
perceived benefit  
0.772  
perceived security relational trust  
e-tax invoice adoption  
KMO  
0.685  
0.022  
0.775  
0.035  
0.901  
0.000  
0.000  
Bartlett’s test for  
sphericity  
Source, Author (2025)  
Factor analysis is a statistical method employed to uncover underlying patterns and associations within a dataset  
of observed variables. It seeks to identify latent factors contributing to the observed variability. EFA is  
particularly valuable for simplifying data by grouping correlated variables, abetting in reducing complexity. By  
assessing shared variances among variables, EFA exposes inherent structures and offers insights into potential  
underlying constructs.  
As highlighted by Costello and Osborne (2005), EFA assists researchers in comprehending intricate datasets by  
identifying variables that cluster together and contribute to the same underlying factor. Widely used in  
psychology, social sciences, and market research, EFA reveals essential dimensions influencing observed  
behaviors or attitudes a weak factor has low factor loadings and fails to explain a significant proportion of the  
variance. A strong factor exhibits high factor loadings, indicating it well represents observed variables. Loadings  
above 0.7 are often considered strong.  
In the provided factor loading table, which pertains to the results of the exploratory factor analysis (EFA), strong  
factor loadings are typically identified based on a threshold of 0.7 or higher in absolute value. Following this  
Page 1657  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
criterion, several factor loadings can be categorized as strong, signifying robust associations between latent  
factors and observed variables.  
Table 6: Factor Analysis (EFA)  
perceived benefit  
perceived security  
relational trust  
e-tax invoice adoption  
PB1  
PB2  
PB3  
PB4  
PB5  
PS1  
0.752  
0.906  
0.356  
0.566  
0.456  
0.922  
0.898  
0.901  
0.917  
0.897  
PS2  
PS3  
PS4  
PS5  
RT1  
RT2  
RT3  
RT4  
RT5  
ETA1  
ETA2  
ETA3  
ETA4  
0.887  
0.906  
0.916  
0.916  
0.889  
0.550  
0.895  
0.456  
0.716  
Source, Author (2025)  
Test of Regression Assumptions  
In academic research, verifying statistical assumptions is vital to ensure result accuracy. Assumptions like normal  
distribution, variance homogeneity, and independence underlie statistical methods and ignoring them can yield  
erroneous conclusions. Recent research by Smith, (2019) and Johnson, (2020) underscores the significance of  
checking assumptions to enhance result credibility. This practice safeguards against biased inferences, bolstering  
findings' trustworthiness in disciplines such as psychology, economics, and biology.  
Test for Normality  
Statistical assumption testing for normality, often using the Shapiro-Wilk test, assesses if data follows a normal  
distribution. In this test, a p-value is generated. The null hypothesis is that the residuals follow a normal  
distribution. A p-value suggests<0.05 departure from normality. Table 7 indicates the p value for perceived  
Page 1658  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
benefit is 0.075>0.05, for perceived security is0.065>0.05, for rational trust is 0.063>0.05, Lastly, the control  
variables Firm age and firm size have p value of 0.055 and 0.355 respectively p<0.05. The results indicates that  
all the independent variables and control variables are normally distributed, p>0.05  
Table 7: Normality Test  
Shapiro-Wilk  
Statistic  
.902  
df  
Sig.  
.075  
.065  
.063  
.055  
.355  
perceived benefit  
perceived security  
relational trust  
Firm Age  
224  
224  
224  
224  
224  
.933  
.928  
.882  
Firm size  
.756  
Source, Author (2025)  
Homoscedasticity Test  
Statistical supposition testing, like the Breusch-Pagan test, inspects homoscedasticity, which ensures consistent  
variability of errors across predictor values. Breusch-Pagan test assesses whether variance in residuals is  
constant. The null hypothesis was that there is constant variance. A lower p-value indicates significant  
heteroscedasticity, implying unequal variability, A significant Breusch-Pagan test (p < 0.05) highlights potential  
issues with linear regression, aiding scholars in refining models, considering transformations, or employing  
robust regression techniques to improve the validity and reliability of their analyses. Recent studies, like Smith  
(2017) and Johnson (2020), underline the importance of addressing this assumption. Table 8 Breusch-Pagan test  
is 0.462, which is greater than the commonly used significance level of 0.05. This non-significant p-value  
indicates that there is no significant evidence of heteroscedasticity in the model's residuals. Therefore, the  
assumption of homoscedasticity is likely met in this case.  
Table 8: Homoscedasticity Test  
Breusch-Pagan / Cook-Weisberg test for heteroskedasticity  
Ho: Constant variance  
Variables: residuals  
chi2(1)  
0.998  
Prob.>  
0.462  
Source, Author (2025)  
Multicollinearity Test  
Multicollinearity is vital for robust analysis. The Variance Inflation Factor (VIF), introduced by Hair et al. (2019),  
is a key metric in this context. VIF quantifies the extent of multicollinearity among predictor variables. A VIF  
value exceeding 10, as suggested by Hair et al., indicates problematic multicollinearity requiring attention. High  
VIF values signify that a predictor can be linearly predicted from others. Managing multicollinearity through  
VIF assessment enhances the reliability of regression analysis, ensuring accurate model interpretation. Table 9  
presents the results of the multicollinearity test, assessing collinearity statistics for the model. For the predictor  
variables: Perceived Benefit: The VIF is 4.950, and the Tolerance is 0.202. Perceived Security: The VIF is 8.772,  
and the Tolerance is 0.114. Relational Trust: The VIF is 5.495, and the Tolerance is 0.182. The control variables  
firm age has a VIF of 8.696, tolerance of 0.125, and firm size has a VIF of 2.817, and tolerance of 0.355. The  
Page 1659  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
assumption for multicollinearity has not been violated since VIFs<10.  
Table 9: Multicollinearity Test  
Collinearity Statistics  
Model  
Tolerance  
.202  
VIF  
Perceived Benefit  
4.950  
8.772  
5.495  
8.696  
2.817  
Perceived Security  
Relational Trust  
Firm Age  
.114  
1
.182  
.115  
Firm Size  
.355  
Source, Author (2025)  
Autocorrelation Test  
Testing for autocorrelation assesses whether successive observations in a dataset exhibit intra correlations. For  
the study of panel data or time series, it is essential. The Durbin-Watson statistic is used to quantify residual  
autocorrelation. There are four possible values: 0 to 4. Values that are substantially below 2 indicate positive  
autocorrelation, values that are over 2 indicate negative autocorrelation, and values that are near 2 indicate no  
autocorrelation. The null hypothesis was that the residuals are independent (no autocorrelation). Applying similar  
tests in the social sciences and econometrics is explained in recent literature, such as Wooldridge (2019). A  
statistic of 2.035 >2 in Table 4.15 suggests that there is no autocorrelation.  
Table 10: Autocorrelation Test  
D-Watson Statistic  
2.035  
Source, Author (2025)  
Linearity Test  
Chen (2005). In this study, linearity tests were conducted individually for each variable to assess the likelihood  
of the data originating from a linearly structured population. If Deviation test is not significant p-value >0.05,  
this means that there was a linear relationship between predictor variables and the outcome variable. According  
to Table 11, showed F statistic of 1.154, and p-value of 0.075>0.05, these results indicated that the assumption  
the there was a linear relationship between the predictors and the outcome variables was not violated.  
Table 11: Linearity Test  
H0: No significant Deviation from Linearity  
F
1.154  
0.075  
Sig  
Source, Author (2025)  
Correlation Analysis  
Correlation analysis, a crucial statistical tool in research, examines relationships between variables. It quantifies  
the degree and direction of associations, helping researchers understand patterns and make predictions Shi et al.  
(2019). The significance value (p-value), commonly set at 0.05, assesses whether correlations observed in a  
Page 1660  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
sample are statistically significant. Significance indicates findings are unlikely due to chance alone. Low p-  
values (<0.05) suggest robust relationships, reinforcing the legitimacy of research conclusions, enhancing  
confidence, and aiding evidence-based decision-making Vassar Stats (2020).  
The correlation analysis reveals notable relationships between the variables under consideration and Electronic  
Tax Invoice Adoption. Perceived Benefit exhibits a strong positive correlation of 0.780 with Electronic Tax  
Invoice Adoption (p<0.05), indicating that perceived benefits are closely linked to the adoption of electronic tax  
invoices. Similarly, Perceived Security demonstrates a substantial positive association of 0.615 (p<0.05),  
suggesting that a higher sense of security contributes to a greater likelihood of adopting electronic tax invoices.  
The control variables indicated a positive and significant correlation with electronic tax invoice adoption firm  
age r= 0.152 p<0.05 and firm size r = 0.12 p<0.05. These findings underscore the interrelated nature of these  
factors and their influence on the adoption of electronic tax invoices.  
Table 12: Correlation Analysis  
ETA  
1
PB  
PS  
RT  
TR  
FA  
FS  
E-Tax Invoice Adoption  
Perceived Benefit  
Perceived Security  
Relational Trust  
Technology Readiness  
Firm Age  
0.780**  
0.615**  
0.660**  
0.805**  
0.152**  
0.120**  
224  
1
0.015  
0.265**  
0.552  
0.336**  
0.665**  
224  
1
0.031  
1
0.003  
0.665  
0.550  
224  
0.231  
0.458  
0.115  
224  
1
0.665  
0.556  
224  
1
Firm Size  
0.112  
224  
1
N
224  
Correlation is significant at 0.05 level **(2 tailed)  
Source, Author (2025)  
Regression Analysis  
Effect of control variables  
The effects of the control variables firm age and firm size on electronic tax invoice adoption was determined by  
a regression model.  
The regression equation was as below.  
Y = 0.127 + 0.012fa + 0.067fs  
The model summary table 4.18 concluded that firm age and firm size collectively have a strong correlation with  
electronic tax invoice adoption R= 0.311, the firm age and firm size also account for a 9.7 % variance caused on  
electronic tax invoice adoption, the remaining 90.3% of the variation was caused by factors not included in the  
model. This variance was significant since the f statistic of 11.870 and p-value <0.05 indicates that the model  
significantly explains the variance The model further showed that a unit change in firm age caused a significant  
unit improvement in adoption of electronic tax invoice of 0.012. Further a unit change in firm size caused a  
significant increase of 0.067 in electronic tax invoice adoption. The findings showed that the control variable  
positively impacts the adoption of electronic tax invoice adoption.  
Page 1661  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Table 13: Regression Model effects of Control variables  
Model  
1
Standardized Coefficients β  
Std. Error  
0.049  
t
Sig.  
(Constant)  
Firm Age  
Firm Size  
0.127  
0.012  
0.067  
2.592  
2.050  
3.511  
0.021  
0.043  
0.001  
0.006  
0.019  
Model summary  
R
0.311  
0.097  
0.089  
R-Square  
Adjusted R-Square  
R-Square Change  
F-Statistic  
11.870  
0.000  
p-value  
Dependent variable: electronic tax invoice adoption  
Source, Author (2025)  
Direct effects  
Multiple Linear regression analysis was used to determine the effects of perceived benefit, perceived security,  
relational trust, firm age and firm size on electronic tax invoice adoption and its determinants among SMEs in  
Kenya. The combined estimate of all factors accounted for approximately 50% of the overall variance in financial  
innovation (R2 = .500, Adjusted R2 = .449). The ANOVA method showed that the combined estimation of all  
independent variables as shown in Table 14 below was statistically significant (F = 43.574, ρ< .05). The model  
was thus designed to predict electronic tax adoption through perceived benefit, security and relational trust. The  
following are the hypotheses for the direct relationship. The standardized coefficients showed that a unit change  
in a unit change in the perceived benefit of electronic tax invoices leads to a significant increase in electronic tax  
invoice adoption, with a standardized coefficient of 0.045. This suggests that if taxpayers believe they are gaining  
some benefit from using electronic tax invoices, their willingness to adopt the system will increase significantly.  
A unit change in the perception of security causes a significant increase in their electronic tax invoice adoption  
by 0.023. This implies that ensuring the security of electronic tax invoice systems can significantly improve  
taxpayer compliance. A unit change in relational trust, the trust that taxpayers have in the tax authorities or the  
system results in a significant increase in electronic tax invoice adoption by 0.025, indicating that building trust  
between taxpayers and tax authorities plays an important role in promoting adoption of electronic tax invoice  
systems. A unit change in firm age corresponds to an increase in electronic tax invoice adoption by 0.012,  
suggesting that older firms are slightly more likely to adopt electronic tax invoices than newer firms. A unit  
transformation in firm size leads to a significant increase in electronic tax invoice adoption by 0.009,  
demonstrating that firms with more employees are more likely to adopt the electronic tax invoice systems.  
Table 14: Regression Results for direct effect  
Model  
2
Standardized Coefficients β  
Std. Error  
0.049  
t
Sig  
(Constant)  
0.127  
2.592  
0.021  
Page 1662  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
Perceived Benefit  
Perceived Security  
Relational Trust  
Firm Age  
0.045  
0.023  
0.025  
0.012  
0.009  
0.021  
0.003  
0.010  
0.006  
0.003  
2.143  
7.667  
2.500  
2.050  
2.967  
0.035  
0.000  
0.039  
0.043  
0.001  
Firm Size  
Model summary  
R
0.707  
0.500  
0.449  
0.360  
43.574  
0.000  
R-Square  
Adjusted R-Square  
R-Square Change  
F-Statistic  
p-value  
a. Dependent Variable: ETA  
b. Predictors: (Constant), PB, PS, RT FA, FA  
Source, Author (2025)  
CONCLUSION AND RECOMMENDATION  
In conclusion, the study revealed that perceived benefit, perceived security, and relational trust all have positive  
and statistically significant effects on the adoption of electronic tax invoices among SMEs in Kenya. The findings  
underscore the importance of practical advantages, trust, and security perceptions in influencing technology  
adoption decisions. Consistent with prior studies from various countries, the results highlight that SMEs are  
more likely to embrace electronic tax systems when they perceive them as beneficial, secure, and supported by  
trustworthy relationships with tax authorities. These insights offer valuable guidance for policymakers and tax  
administrators seeking to enhance e-tax adoption through targeted interventions that build trust, ensure system  
security, and communicate the tangible benefits of digital tax solutions.  
Based on the findings, policymakers should enhance the adoption of electronic tax invoices by promoting  
awareness campaigns that clearly communicate the benefits of the system, strengthening system security  
protocols to build user confidence, and fostering trust through transparent engagement with SMEs. Investment  
in user-friendly digital platforms and responsive customer support can also improve perceived utility and  
security. Additionally, tax authorities should establish consistent, trust-based relationships with SMEs by  
involving them in system design and feedback processes. For managers, it's essential to train staff on the use and  
advantages of electronic tax systems, integrate secure IT infrastructure, and actively engage in trust-building  
initiatives with regulatory bodies to encourage smooth and sustained adoption.  
REFERENCES  
1. Acharya, S. (2013). A note on stable limit theory for the OLSE with non usual rates and the robust  
Wald test. Communications in Statistics - Theory and Methods, 47(1),28-41.  
2. AlAwadhi, S., Morris, A., & Wynne, A. (2009). The effect of attitudes on the behavioral intention  
to adopt e-government services among Kuwaiti citizens. Journal of Enterprise Information  
Management, 22(6), 682-703.  
Page 1663  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
3. Almarri, K., Alomari, A., & Almajali, D. (2021). SMEs' intention to adopt electronic invoicing in  
the UAE: The role of perceived benefits, trust, and perceived security. Journal of Business  
Research, 123, 690-701.  
4. Al-Mudimigh, A. S., Zairi, M., & Al-Mashari, M. (2012). The adoption of electronic invoicing by  
SMEs in Saudi Arabia: An exploratory study. International Journal of Electronic Business  
Management, 10(3), 182-195.  
5. Almutairi, A., Qureshi, M. A., & Baabdullah, A. M. (2019). Perceived Security and Trust towards  
Electronic Tax Invoicing (ETI) Adoption: A Study of Saudi Arabian SMEs. Sustainability, 11(11),  
3065.  
6. Awai, E. S., & Oboh, T. (2020). Ease of paying taxes: The electronic tax system in  
Nigeria. Accounting and taxation review, 4(1), 63-73.  
7. Baron, R. M., & Kenny, D. A. (1986). The Moderator-Mediator Variable Distinction in Social  
8. Belanger, L. C. (2005). The utilization of e-government services: Citizen trust, innovation and  
acceptance factors. Information Systems Journal, Vol. 15 No. 1, pp. 5-25.  
9. C. Kun, K. L. (2008). User evaluation of tax filing websites. Journal of online information Review,  
32(6), 842-859.  
10. Chang M. S., & Younus, S. (2013). Determinants of tax-compliance behaviour explored by  
slippery: An evidence from small business owner. Journal of Management Sciences, 6(2), 33-47.  
11. Chang, Y. C., Chang, K. H., & Chang, T. K. (2013). Applied column generation-based approach to  
solve supply chain scheduling problems. International Journal of Production Research, 51(13),  
4070-4086.  
12. Chen, L., Hong, W., & Li, S. (2020). Examining the determinants of Chinese SMEs' electronic tax  
filing adoption: The moderating role of trust. Journal of Business Research, 112, 60-70.  
13. Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four  
recommendations for getting the most from your analysis. Practical Assessment, Research &  
Evaluation, 10(7), 1-9.  
14. Cuong, T. T., Thao, T. P., & Linh, B. P. (2023). The Impact of Job Satisfaction on Job Performance  
of Tax Officials: An Empirical Study in Vietnam. International Journal of Professional Business  
Review: Int. J. Prof. Bus. Rev., 8(10), 4.  
15. E. Moyi and E. Ronge. (2006). Taxation and tax modernization in Kenya and options for further  
reforms. Nairobi, Kenya: Institute of Economic Affairs.  
16. Fidler, F., & Wilcox, J. (2019). Null hypothesis significance testing: A review of an old and  
continuing controversy. Psychological Methods, 24(4), 411-427.  
17. Fu, J. F. (2006). Acceptance of electronic tax filing: a study of taxpayer intentions. Information and  
Management, Vol. 43 No. 1, pp. 109-126.  
18. Göritz, A. S. (2020). Recommendations for improving response rates in student and other  
populations. International Journal of Social Research Methodology, 23(2), 189-199.  
19. Hair, J. F., Hult, G. T. M., Ringle, C. M., & Sarstedt, M. (2019). A Primer on Partial Least Squares  
Structural Equation Modeling (PLS-SEM) (2nd ed.). Sage.  
20. Hart and Saunders T.L. & Tran, M. D. (1997). Determinants influencing tax compliance: The case  
of Vietnam. The Journal of Asian Finance, Economics, and Business, 7(2), 65-73.  
21. Hart, P. a. (1997). Power and trust: Critical factors in the adoption and use of electronic data  
interchange. Organization Science, Vol. 8 No. 1, pp. 23-42.  
22. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A  
regression-based approach. Guilford Press.  
23. Hernandez-Ortega & Puteh Salin, A. S. A. (2012). Improving tax compliance via tax Education  
experience. Management & Accounting Review (MAR), 15(2), 1-20.  
24. Hernandez-Ortega, B. (2012). Key factors for the adoption and subsequent use of electronic  
invoice. Vol. 50, pp. 15-30.  
25. Hernandez-Ortega, B., & Jimenez-Martinez, J. (2013). Performance of e-invoicing in Spanish  
firms. Information Systems and E-Business Management, 11, 457-480.  
26. Hsu, C. L., & Yen, C. H. (2018). Understanding the adoption of electronic tax filing systems:  
Technology readiness, privacy concerns, and perceived risk. Journal of Enterprise Information  
Management, 31(6), 830-847.  
Page 1664  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
27. Huang, Y., & Fang, Y. (2017). Perceived benefits, costs and barriers to the adoption of electronic  
tax invoicing: A study of Chinese small and medium-sized enterprises. Information Technology &  
People, 30(1), 179-202.  
28. Hwang, J., & Kim, Y. (2012). Understanding customer acceptance of e-banking services: An  
extension of the technology acceptance model. Journal of Business Research, 65(6), 789-800.  
29. Hyz, U., & Bougie, R. (2010). Research methods for business—A skill building approach. 6th  
edition. West Sussex: John Wiley and Sons.  
30. Iacovou, C. B. (1995). Electronic data interchange and small organizations: adoption and impact  
of technology. MIS Quarterly, Vol. 19 No. 4, pp. 465-485.  
31. Johnson, A. B., Smith, C. D., & Williams, E. F. (2018). The Role of Relational Trust in Electronic  
Tax Invoice Adoption. Journal of Business Technology, 42(3), 156-170.  
32. Johnson, J. B. (2020). Assumption testing in the analysis of variance. The Quantitative Methods  
for Psychology, 16(2), 110-119.  
33. Kemper, A. (2003). Reliability, validity and true values in surveys. Social Indicators Research,  
78(1), 85-110.  
34. Kim, H. W., Chan, H. C., & Gupta, S. (2020). Ensuring the quality of survey data: The role of  
reliability and validity. MIS Quarterly Executive, 19(1), 1-16.  
35. Komba, G. P., & Lema, E. M. (2019). Factors influencing electronic tax (e-tax) invoice adoption  
among small and medium enterprises in Tanzania: A case of Dar es Salaam region. Journal of  
Economics and Business, 2(2), 30-47.  
36. Kothari, C. R. (2002). Research Methodology; Methods and Technique, 2nd Edition: New Age  
International P Ltd.: New Delhi.  
37. Krishnan, R. M. (2006). When does trust matter to alliance performance? Acad. Management J,  
Vol. 49 No. 45, pp. 894-917.  
38. Lee, H. C. (2016). Can electronic tax invoicing improve tax compliance? A case study of the  
Republic of Korea's electronic tax invoicing for value-added tax. A Case Study of the Republic of  
Korea's Electronic Tax Invoicing for Value-Added Tax (March 7, 2016). World Bank Policy  
Research Working Paper, (7592).  
39. Lee, J. H., & Park, S. K. (2020). Reassessing the Role of Relational Trust in Electronic Tax Invoice  
Adoption: A Comparative Analysis. Contemporary Business Research, 46(5), 678-692.  
40. Lee, S. M., & Kim, H. J. (2019). Factors affecting the adoption of electronic tax filing: Evidence  
from South Korea. International Journal of Electronic Commerce, 23(1), 98-123.  
41. Lin, C.-H., & Lin, I.-C. (2013). The adoption of electronic tax invoicing among SMEs: A study of  
perceived benefits. Journal of Global Information Management, 21(4), 17-36.  
42. Ma, X., & Sun, Y. (2021). Understanding the determinants of taxpayers' adoption of electronic tax  
filing in China: The role of trust. Technological Forecasting and Social Change, 166, 120618.  
43. Manly, D. T. (2005). Attracting non-filers through amnesty programmes. Journal of American  
Taxation Association, 27, 75-95.  
44. Mwacha, J. M., & Kashasha, L. K. (2021). Determinants of electronic tax (e-tax) invoice adoption  
among small and medium enterprises in Tanzania. International Journal of Scientific and Research  
Publications, 11(1), 300-311.  
45. Nguyen Charity, M., Chan, W. (2020). Literature review on the impact of tax knowledge on tax  
compliance among small medium enterprises in a developing country. International Journal of  
Entrepreneurship, 22(4), 1-15.  
46. Nguyen, M., & Nguyen, J. H. (2020). Economic policy uncertainty and firm tax  
avoidance. Accounting & Finance, 60(4), 3935-3978.  
47. Nwabachili, C. O., Nwankwo, C. B., & Ugwu, M. O. (2024). Legal assessment of electronic tax  
payment systems: Nigeria, Kenya and tanzania and in perspectives. Nnamdi Azikiwe University  
Journal of International Law and Jurisprudence, 15(1), 129-140.  
48. Nwokah, N. G., Okorie, G. C., & Okafor, C. E. (2019). Determinants of e-invoicing adoption in  
Nigeria: A study of selected small and medium enterprises in Aba metropolis. International Journal  
of Economics, Commerce and Management, 7(5), 33-46.  
49. OECD, O. (2004). The OECD principles of corporate governance. Contaduríay Administración, ,  
(216).  
Page 1665  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
50. OECD. (2015). Addressing the tax challenges of the Digital Economy. OECD.  
51. Ogunleye, O. M., & Akinbode, O. R. (20182020). An empirical analysis of electronic tax (e-tax)  
adoption in Nigeria: A study of small and medium enterprises (SMEs). Academy of Accounting  
and Financial Studies Journal, 22(3), 1-14.  
52. Okunogbe, O., & Santoro, F. (2023). Increasing tax collection in African countries: The role of  
information technology. Journal of African Economies, 32(Supplement_1), i57-i83.  
53. Oso, J., & Open, E. (2008). Multiple Regression that Researchers should always test. Practical  
Assessment, Research, and Evaluation, 8(2).  
54. Psychological Research. Conceptual, Strategic, and Statistical Considerations. Journal of  
Personality  
and  
Social  
Psychology,  
51(6),  
1173–1182.  
https://doi.org/10.1037/0022-  
3514.51.6.1173  
55. Qi, Y., & Che Azmi, A. (2021). Factors affecting electronic invoice adoption and tax compliance  
process efficiency. Transforming Government: People, Process and Policy, 15(1), 150-168.  
56. Razali, N. M., & Wah, Y. B. (2011). Power comparisons of Shapiro-Wilk, Kolmogorov-Smirnov,  
Lilliefors, and Anderson-Darling tests. Journal of Statistical Modeling and Analytics, 2(1), 21-33.  
57. Rosli, A. H., Yusof, R. M., Ahmad, S. B., & Shafie, R. A. (2018). Determinants of e-tax adoption  
among Malaysian SMEs: A conceptual model. Journal of Advances in Information Technology,  
9(1), 1-8.  
58. Saunders, P. (2007). Social theory and the urban question. Routledge.  
59. Shah, S. Z. A., Jaffari, A. R., & Hussain, J. (2018). Understanding the determinants of e-tax  
adoption intention in developing countries: Evidence from Pakistan. Journal of Business Research,  
91, 238-251.  
60. Sheehan, K. B. (2020). E-mail survey response rates: A review. Journal of Computer-Mediated  
Communication, 25(2), 113-133.  
61. Shen, C. C., Wu, M. L., & Lin, S. S. (2020). Understanding the determinants of electronic tax filing  
intention: Evidence from small and medium-sized enterprises in Taiwan. Journal of Enterprise  
Information Management, 33(3), 593-611.  
62. Shi, D., Lee, S. Y., & Maydeu-Olivares, A. (2019). Understanding and Using Factor Scores:  
Considerations for the Applied Researcher. Practical Assessment, Research & Evaluation, 24(1), 1-  
13.  
63. Shim & SongSubekti, I., & Baridwan, Z. (2016). Trust and uncertainty orientation: An effort to  
create tax compliance in social psychology framework. Procedia- Social and Behavioral Sciences,  
211, 938-944.  
64. Shim, T. S., & Song, I. K. (2016). Critical success factors of the web-based tax invoice system in  
Korea. KSII Transactions on Internet and Information Systems (TIIS), 10(9), 4487-4507.  
65. Shim, T. S., & Song, I. K. (2016). Critical success factors of the web-based tax invoice system in  
Korea. KSII Transactions on Internet & Information Systems, 10(9).  
66. Smith, J. K., & Smith, L. F. (2019). Validity and reliability in research instruments. Journal of the  
American Academy of Physician Assistants, 32(12), 1-3.  
67. Smith, P. (2019). The importance of statistical assumptions. Journal of Research Practice, 15(2),  
Article V.  
68. Susilo, D. J. (2022). Effect of Electronic Tax Invoice Adoption on Transparency of Indonesia Stock  
Exchange Main Board Companies. Journal of Indonesian Applied Economics, 10(2), 153-179.  
69. Sweeten, G. (2016). Lecture 8: Heteroskedasticity Assumption MLR5 : Homoskedasticity. Class  
Lecture.  
70. Tan, J., Li, C., Chen, Y., & Wang, S. (2019). Factors affecting the adoption of electronic tax filing  
in China: A trust-based perspective. Information Systems Frontiers, 21(3), 577-593.  
71. Tilahun, M. (2019). Determinants of tax compliance: a systematic review. Economics, 8(1), 1-7.  
72. VassarStats. (2020). Understanding p-values. Retrieved from http://vassarstats.net/.  
73. Wahyuni, D., & Suhardi, B. (2020). The role of perceived ease of use, perceived usefulness, and  
perceived security in the adoption of e-tax. Journal of Accounting and Investment, 21(1), 1-11.  
74. Walsh, H. O., & Roespinoedji, D. S. (2021). The role of tax education in supply chain management  
a case of Indonesian supply chain companies. Polish Journal of Management Studies, 18.  
75. Waris, M. K. (Enhancing revenue capacity to enhance human welfare). Taxation and state building  
Page 1666  
www.rsisinternational.org  
INTERNATIONAL JOURNAL OF RESEARCH AND INNOVATION IN SOCIAL SCIENCE (IJRISS)  
ISSN No. 2454-6186 | DOI: 10.47772/IJRISS | Volume IX Issue V May 2025  
in Kenya: Kenya Report: Tax Justice Network-Africa (TJN-A).  
76. Wooldridge, J. M. (2019). Introductory econometrics: A modern approach. Nelson Education  
77. Yang, C. C., & Hsu, C. L. (2012). The effect of website quality on customer satisfaction and  
purchase intention: Implications for e-businesses. Procedia-Social and Behavioral Sciences, 57,  
217-224.  
78. Yu, T. H., & Fang, K. (2019). Perceived technology readiness and e-invoicing adoption among  
SMEs: An empirical study in Taiwan. Sustainability, 11(23), 6613.  
79. Zhang, J., Hu, Q., & Zhu, K. (2018). Exploring the determinants of e-tax adoption intention:  
Evidence from Chinese SMEs. Electronic Commerce Research and Applications, 29, 117-129.  
80. Zhang, Z., & Zhang, H. (2020). Assumption Checking for Parametric Models. Springer.  
Page 1667  
www.rsisinternational.org