Digital Financial Capability Scale

Abstract

Financial digitization is an irreversible phenomenon. The objective of this study is to construct the Digital Financial Capability Scale (DFCS). Starting with the development of a definition, we created a multidimensional scale composed of digital financial knowledge, digital financial behavior, and digital financial confidence. The validation process involved a qualitative stage, consisting of focus groups, expert validation, and pre-testing, and a quantitative stage, with exploratory and confirmatory factor analyses and structural equation modeling. The DFCS assesses an individual’s perception of their ability to apply financial knowledge, adopt appropriate financial behaviors, and feel confident in making financial decisions in a digital environment. The final version of the DFCS consists of a set of 33 items divided into the three dimensions. The scale can be very useful for researchers who wish to study financial capability in the digital environment, for financial agents to evaluate clients, and for assessing the outcomes of public policies aimed at enhancing the financial capability of the population.

1. Introduction

Financial digitization is an irreversible phenomenon. Self-service terminals are falling out of use at the same rate that digital financial transactions via mobile banking are increasing. Physical branches are giving way to digital banks and fintechs on a massive scale worldwide. Central banks are creating and issuing their own digital currencies.

These changes redefine banking processes (Murinde et al. 2022) and intensify competition. Fintech has disrupted traditional banking models and challenged established financial institutions to embrace digital transformation (Mainardes and Freitas 2023), changing consumer behavior in the banking industry (Mogaji 2023). Fintech offers new and alternative ways for individuals and businesses to manage their finances, make payments, access credit, and invest their assets (Mogaji et al. 2022b), making banking services more accessible, convenient, and efficient (Mogaji and Nguyen 2022a). In this context, the question arises: what is the capacity of citizens to handle this digital financial transformation?

The importance of this question is linked to the fact that the application of new digital technologies in the financial sector can generate benefits such as increased access to financial services; the provision of faster, more secure, and uninterrupted transactions; and the expansion of supply and competition among financial companies (Elsinger et al. 2018; Koskelainen et al. 2023). From an economic perspective, studies indicate positive impacts on economic growth (Liu et al. 2021), corporate innovation (Ding et al. 2022), individual entrepreneurship (Sun and Xie 2024), household income generation (Cai et al. 2024), green development (Pang et al. 2024; Xin and Xie 2023), and the reduction of income inequality (Yan et al. 2024).

However, like other technological innovations, technological advances in the financial sector present potential risks (Yadav and Banerji 2023). These involve new risks such as the misuse of unknown products by uninformed consumers, new types of fraud, lack of security, privacy, and data confidentiality, excessive use of digital profiles to identify potential customers, exclusion of undesirable groups, and quick access to high-cost credit or essentially speculative products (Asif et al. 2024). Thus, to ensure that the advantages lead to greater financial inclusion and that these new risks are mitigated, it is essential that citizens are capable of handling all these emerging digital transformations in the financial market.

Another key point in addressing this question is understanding the meaning of digital financial capability. There is no consensus in the literature regarding its definition, but in this article, we propose that the term involves a combination of financial capability with digital literacy. Financial capability refers to the ability to apply appropriate financial knowledge and perform desirable financial behaviors to achieve financial well-being (Xiao et al. 2014). Digital literacy is the ability to understand and use information in multiple formats from a wide variety of sources (Gilster 1997).

Moreover, a necessary condition for addressing this question is the existence of a consolidated measure. However, as far as we know, there is no validated measure of digital financial capability in the literature. Thus, considering the relevance of digital financial capability in the current technological landscape and the lack of a consolidated definition and scale, the objective of this study is to construct the Digital Financial Capability Scale (DFCS). Starting with the development of a definition, we developed a multidimensional scale composed of digital financial knowledge, digital financial behavior, and digital financial confidence.

The existence of an adequate model for evaluating digital financial capability is crucial for countries to assess their progress in achieving citizens’ financial inclusion goals and financial well-being. As countries develop public financial literacy policies and begin to integrate digital competencies, a measure for assessing the population’s digital financial capability is indispensable for identifying the outcomes of these policies. Moreover, the effectiveness of policies such as the creation of digital accounts for emergency aid payments and income transfer policies depends on the population’s digital financial capability. Therefore, without understanding the beneficiary’s ability to handle technology, such policies may not achieve the expected results.

For academia, measurement models are essential for advancing research. For the financial system, understanding the client’s digital financial capability can assist in developing financial apps, guiding product development, and creating consumer risk mitigation programs.

2. Building the Digital Financial Capability Scale

The need to create a scale capable of assessing digital financial capability is linked to the rapid and intense growth of digital financial services (Lyons et al. 2022). Demirgüç-Kunt et al. (2022) report that 64.1% of adults worldwide have embraced digital payments. However, the digitalization of financial services requires new skills and an understanding of digital financial risks to allow customers to perform everyday financial actions (Koskelainen et al. 2023). In this context, for an individual to be effectively financially included, they must have digital literacy in addition to financial capability. Thus, the Digital Financial Capability Scale (DFCS) is built on two central concepts: financial capability and digital literacy.

Some researchers define financial capability and financial literacy as interchangeable concepts (Lusardi and Mitchell 2014; Xiao and Porto 2017; Xiao and Bialowolski 2023). Lusardi and Mitchell (2014) define financial capability as people’s ability to process economic information and make informed decisions about financial planning, wealth accumulation, debt, and pensions. For Lučić et al. (2023), financial capability is the capacity of consumers to undertake comprehensive financial activities that assure individual financial well-being. Similarly, Xiao and Bialowolski (2023) define financial capability as an individual’s ability to apply appropriate financial knowledge, perform desirable financial behaviors, and take available financial opportunities to achieve financial well-being.

These definitions are based on the theory of self-efficacy, in which self-efficacy is considered an important determinant of goal achievement (Bandura 1982). Self-efficacy relates to an individual’s sense of how well they can carry out the different behaviors required to manage diverse situations. It is similar to perceived behavioral control (Liu et al. 2020) and influences one’s mental processes, emotional reactions, and behavioral preparations, making it one of the most essential predictors of action (Bandura 1986). In the context of financial behavior, if the consumer has positive experiences and is motivated during financial decision making, and/or receives financial education and reduces stress levels, their self-efficacy will be higher, and they will be able to achieve financial capability as the desirable behavior (Lučić et al. 2023).

Regarding the dimensions of financial capability, there are different proposals. Some models associate the concept with financial knowledge, skills, confidence, and behaviors (Bialowolski et al. 2021; Xiao and O’Neill 2016, 2018). In some studies, researchers emphasize one dimension, such as financial behavior, financial knowledge, financial opportunity, or financial outcomes. (Xiao et al. 2024). Lučić et al. (2023) propose that financial capability comprises financial literacy and mental capacity to engage in thought processes, reasoning, and comprehension to behave in a financially responsible manner. Xiao et al. (2014) indicate that financial capability is built from the interactions between three components: knowledge, behavior, and opportunity.

However, when it comes to financial capability applied in the context of digital financial services, a new set of skills becomes important. Digital literacy is necessary for individuals to use digital financial services in an informed, responsible, and transformative manner. Digital literacy was initially conceived as the ability to understand and use information in multiple formats from a wide variety of sources (Gilster 1997). Applied in the context of services, digital literacy refers to the knowledge, skills, and abilities to effectively access and use digital financial services (Lyons and Kass-Hanna 2021).

Thus, based on the concepts of financial capability and digital literacy, we define digital financial capability as the ability to apply financial knowledge, adopt appropriate financial behaviors, and be confident in making financial decisions in a digital environment. Therefore, the Digital Financial Capability Scale (DFCS) assesses an individual’s perception of their digital financial capability. Perception is defined as the process, result, or behavior of people based on their interpretation of objects, relationships, and events (VandenBos 2007).

In this definition, three key dimensions are identified for exploration in the digital environment: knowledge, behavior, and confidence. The first two dimensions are present in both financial literacy (Organization for Economic Co-Operation and Development 2020) and financial capability (Xiao et al. 2014) definitions, making them essential for constructing the digital financial capability scale. Confidence, on the other hand, is a crucial factor for individuals to be willing to use digital financial technologies (Shankar and Datta 2018). Additionally, according to the Organization for Economic Co-Operation and Development (2014), financial literacy involves the confidence to apply knowledge in making decisions across various financial contexts. Thus, the DFCS will be a three-dimensional scale with digital financial knowledge, digital financial behavior, and digital financial confidence as its formative dimensions.

Financial knowledge is a method through which people enhance their understanding of financial concepts and risks, enabling them to develop the skills and confidence necessary to make fundamental and secure decisions (Organization for Economic Co-Operation and Development 2017). Meanwhile, knowledge of digital financial products and services captures the basic understanding of digital financial products and services (Morgan et al. 2019).

Table 1. Items of the digital financial knowledge dimension.

Item
Basic	1	Check account statements on digital channels.
	2	Pay bills on digital channels.
	3	Change withdrawal and transfer limits on digital channels.
	4	Activate an account on digital channels.
	5	Locate payment proof for a bill, transfer, or real-time payment on digital channels.
	6	Change the access password on digital channels.
	7	Activate/deactivate the card for domestic purchases on digital channels.
	8	Create a virtual credit card on digital channels.
	9	Open an account at a digital bank.
	10	Register a real-time payment key.
	11	Make online purchases via real-time payment.
	12	Unblock the app through self-service.
	13	Install a banking app on your mobile phone.
	14	Use digital channels without the help of another person.
	15	Identify the fees charged by the bank/brokerage on digital channels.
Advanced	16	Perform an investment or redemption in investment funds or agribusiness letter of credit on digital channels.
	17	Perform an investment or redemption in government bonds on digital channels.
	18	Perform an investment or redemption in bank deposit certificates on digital channels.
	19	Buy and sell stocks on digital channels.
	20	Buy and sell foreign currency on digital channels.
	21	Buy and sell derivatives (call/put options, futures market, etc.) on digital channels.
	22	Contract insurance on digital channels.
	23	Release access to open finance (sharing your data between financial institutions) on digital channels.

Note: alternatives: yes, no, and never tried.

presents the items proposed for assessing digital financial knowledge. These items were inspired by the studies of Abdallah et al. (2024), Ravikumar et al. (2022), Respati et al. (2023), and Kass-Hanna et al. (2022).

Financial behavior is human behavior in relation to financial management (Dew and Xiao 2011). Financial behavior and habits are the most significantly impacted by digitalization (Garai-Fodor et al. 2022). Thus, digital financial behavior refers to an individual’s behavior in financial management within a digital environment. This includes engaging in online purchasing behaviors, online money transfers, and other digital financial transactions (Chhillar et al. 2024). In the context of the scale, appropriate digital financial behavior refers to consumer behavior in money management in the digital context that seeks to achieve financial well-being. The proposed items for assessing digital financial behavior are presented in

Table 2. Items of the digital financial behavior dimension.

Item
24	I prefer using financial apps over going to a physical bank or ATM.
25	I try to solve my financial issues through the banking app before going to the bank or ATM.
26	I avoid carrying cash in my wallet (money, coins).
27	I buy financial products (life insurance, car insurance) through the financial app.
28	I save more frequently since I started using financial apps.
29	I consider the values of the fees that are charged before making a digital financial transaction.
30	I control my finances more since I started using financial apps.
31	I try to keep my access passwords to the digital apps secure.

Note: Alternatives: strongly disagree, disagree, indifferent, agree, and strongly agree.

. The construction of the items considered, in addition to the items traditionally used in financial behavior scales, are from the studies of Setiawan et al. (2022), Kumar et al. (2023), Ravikumar et al. (2022), and Lyons and Kass-Hanna (2021).

The third dimension of the scale is digital financial confidence. Financial confidence is the self-assurance needed to make sound financial decisions (Palameta et al. 2016). To sustain the impact of financial knowledge on financial decision making and behavior, financial confidence is essential (Morris et al. 2022). Although financial knowledge brings benefits, individuals who demonstrate greater financial confidence may be more capable of implementing healthy financial choices (Atlas et al. 2019) and more willing to engage in negotiation (Krische and Mislin 2020). Already in the digital context, trust is the individuals’ expectation that digital technologies and services—and the organizations providing them—will protect all stakeholders’ interests and uphold societal expectations and values (World Economic Forum 2022). Therefore, the digital financial confidence dimension assesses how confident individuals feel about using, and how safe and secure they feel while using, digital financial services. It was inspired by studies highlighting the importance of security issues when using the system, knowledge of the risks involved, consumer rights (Rahim et al. 2022; Kumar et al. 2023; Ravikumar et al. 2022; Lyons and Kass-Hanna 2021), and also financial confidence issues (Morris et al. 2022; Respati et al. 2023; Xiao and Meng 2023).

Table 3. Items of the digital financial confidence dimension.

Item
32	I can operate different digital payment methods without difficulty (e.g., PayPal, Amazon Pay, etc.).
33	I can identify the financial products I wish to analyze in financial apps.
34	I believe that the information on products and services provided in banking apps is sufficient for me to make the best financial decisions.
35	I can resolve errors that occur in digital financial transactions.
36	I am proud of knowing how to handle digital financial operations.
37	I am confident that I will have the necessary skills to continue using financial apps in the future.
38	Continuously using digital financial products and services makes me feel confident and error-free.
39	I am aware that I am exposed to various risks when making a digital financial transaction.
40	I am confident that I know how to protect myself against the risks involved in digital financial operations.
41	I am confident that my problems will be resolved if my account is attacked by any digital scam.
42	I trust that my complaints in the digital channels provided by the app will be quickly responded to.
43	I trust that the protection systems used in the banking app (password, biometrics, etc.) are secure.

Note: Alternatives: strongly disagree, disagree, indifferent, agree, and strongly agree.

presents the set of items.

Therefore, the DFCS is a scale capable of assessing an individual’s perception of their digital financial capability, consisting of three main dimensions: digital financial knowledge, digital financial behavior, and digital financial confidence. Figure 1 represents the proposed theoretical model for the DFCS.

Digital financial knowledge will be represented by two variables that assess the individual’s basic and advanced digital financial knowledge, as classified in Table 1. The knowledge variables are constructed from the 15 items that make up basic financial knowledge and the 8 that constitute advanced knowledge. Digital financial behavior will be formed from the 8 items presented in Table 2, and digital financial confidence will be constituted by the 12 items in Table 3.

3. Method

The Digital Financial Capability Scale (DFCS) was developed through a qualitative and a quantitative stage. The first stage involved conducting a focus group, expert validation, and a pre-test to define the measure and initially construct the items. The second stage involved exploratory and confirmatory factor analyses, as well as structural equation modeling to validate the DFCS.

The focus group is a structured discussion method in which a small group of people are asked about their perceptions, opinions, beliefs, and attitudes toward a product, service, concept, or idea (Krueger 2014). The focus group comprised five individuals selected for convenience: two bank employees with extensive customer service experience, two researchers in financial literacy, and one frequent user of financial apps. During the session, a presentation was given on the definition and dimensions of the DFCS, and participants were encouraged to discuss the definitions and items. Throughout the discussions, participants suggested modifications to some items and proposed new ones. They were also invited to categorize the digital financial knowledge items as basic or advanced.

For expert validation, following DeVellis and Thorpe (2021) recommendation, five experts were contacted. Three researchers have experience in scale development and two researchers have expertise in behavioral finance. At this stage, a document was sent containing an introduction, definitions of the scale and dimensions, and a presentation of the items to be evaluated for relevance and language adequacy.

For the pre-test, ten respondents were selected for convenience, as recommended by Boateng et al. (2018). The aim of the pre-test was to ensure the appropriateness of the scale items’ language for individuals with different socioeconomic and demographic profiles. The participants were interviewed by the researchers and encouraged to express their level of understanding of the items.

In the second stage, an online instrument was developed and administered to 775 participants in the second half of 2023. The study was approved by the ethics committee, and participants signed the Informed Consent Form.

The sample is predominantly female (51.7%), single (64.5%), and of white ethnicity (70%). The average age is 30 years, with a minimum of 16 and a maximum of 87 years. Regarding income, 8.6% have no income, and 56.3% earn up to BRL 3960.00 (approximately USD 792.00). As for educational attainment, 17.9% have completed at most high school, and 46.7% of the sample has either completed or is pursuing a college degree.

In the first phase of this stage, exploratory factor analyses were estimated to validate each dimension of the scale. A polychoric correlation matrix was used, with the Unweighted Least Squares (ULS) factor extraction method. Based on Timmerman and Lorenzo-Seva (2011), the optimal implementation of parallel analysis was performed to estimate the number of factors. It is noteworthy that factor analysis using polychoric correlations tends to more accurately identify the number of underlying factors in the data and also produces more consistent parametric estimates of factor loadings and correlations between factors (Asún et al. 2016). Internal consistency was also assessed through Cronbach’s Alpha (Cronbach 1951) and McDonald’s Omega (ω) (McDonald 1999), with values equal to or greater than 0.7 considered adequate (Hair et al. 2019).

In the second phase, convergent validity, unidimensionality, and discriminant validity of the constructs were verified. The models were estimated by maximum likelihood via a direct procedure. Convergent validity was analyzed by observing the magnitude and statistical significance of the standardized coefficients, using the following absolute fit indices: chi-square statistics (χ²), Residual Root Mean Square (RMRS), Root Mean Square Error of Approximation (RMSEA), Goodness of Fit Index (GFI), and the Comparative Fit Index (CFI). Values greater than 0.950 were considered for CFI, GFI, and NFI; and RMSEA and RMRS values lower than 0.060 and 0.080, respectively (Byrne 2016; Hair et al. 2019; Kline 2023).

The unidimensionality of the construct indicates the degree to which a set of items represents only one construct (Garver and Mentzer 1999). Hair et al. (2019) point out that constructs with standardized residuals smaller than 2.58 at a significance level of 5% are considered unidimensional. Discriminant validity was assessed following Fornell and Larcker (1981), where the square root of the Average Variance Extracted (AVE) should be greater than the correlation between the constructs. In the final phase of the quantitative stage, structural equation modeling was used to adjust the final DFCS model.

4. Analysis of Results

The validation of the DFCS involved a qualitative stage, consisting of a focus group, expert validation, and a pre-test, and a quantitative stage, divided into exploratory and confirmatory phases. Thus, the analysis of the proposed scale began with the focus group. Based on the literature review, the authors previously constructed the definition of the DFCS and the initial set of items. During the focus group, the definition of the scale and the items were presented. Participants were then questioned about the items in each of the three dimensions. The participants judged the dimensions as appropriate for assessing digital financial capability.

Regarding the items, the participants deemed the proposed items appropriate and suggested six new items (3, 4, 5, 10, 12, and 23 from Table 1) for digital financial knowledge, two items (28 and 30 from Table 2) for digital financial behavior, and two more items (37 and 42 from Table 3). The focus group participants were also responsible for classifying the items of the digital financial knowledge scale into basic and advanced categories.

Following the qualitative stage of the DFCS validation, two additional validations were conducted. The first involved a sample of five experts selected for convenience, who evaluated the scale based on item relevance and language appropriateness. In this evaluation, all five experts considered the items relevant for constructing each dimension. Regarding language, the following changes were suggested and implemented: (1) for item 15, the inclusion of the word “broker,” and (2) items 16 and 17, originally described as a single item, were separated.

The second validation, conducted with a sample of ten individuals from various profiles, assessed the face validity of the scale. All interviewees demonstrated an adequate understanding of the items, and no further changes were needed. Thus, after the focus group, expert validation, and pre-test, the scale was finalized with the 43 items presented in Section 2. In the subsequent stage, for the quantitative validation phases of the scale, the instruments were applied, and a sample of 775 cases was obtained. This stage began with exploratory factor analyses, aimed at the initial validation of the proposed measurement models.

The exploratory factor analysis of digital financial knowledge indicated through parallel analysis that the measurement model is bidimensional, with one dimension formed by basic knowledge items and another by advanced knowledge items (

Table 4. Results of the exploratory factor analysis for digital financial knowledge.

DFKS	Item	Factor Loading	M.S.A	Parallel Analysis (Percentage 95%)		Alpha	Omega
Basic	1	0.803	0.934	50.616 *	10.075	0.970	0.970
	2	0.834	0.935	12.862 *	9.023
	3	0.665	0.943	3836	8.179
	4	0.728	0.949	3.369	7.606
	5	0.703	0.946	3.253	7.103
	6	0.721	0.958	2.922	6.722
	7	0.634	0.941	2.678	6.316
	8	0.741	0.924	2.561	5.960
	9	0.732	0.926	2.447	5.619
	10	0.911	0.944	2.239	5.278
	11	0.668	0.942	2.178	4.931
	12	0.667	0.970	1.832	4.646
	13	0.872	0.910	1.675	4.358
	14	0.654	0.940	1.531	4.057
	15	0.591	0.955	1.413	3.756
Advanced	16	0.644	0.905	1.095	3.448	0.963	0.963
	17	0.903	0.922	1.032	3.180
	18	0.827	0.920	0.690	2.762
	19	0.908	0.927	0.578	2.449
	20	0.689	0.913	0.550	2.089
	21	0.640	0.901	0.392	1.696
	22	0.668	0.927	0.253	1.142
	23	0.543	0.931

Note: * Advised number of dimensions, based on optimal implementation of the parallel analysis.

).

It is observed that all Measures of Sampling Adequacy (MSA) are above 0.9, indicating that all items can be retained in the scale. Additionally, all items have adequate factor loadings, supporting their retention. The parallel analysis indicates that digital financial knowledge is composed of two dimensions. An analysis of the items within each dimension shows that the first dimension corresponds to items classified as basic, and the second to items classified as advanced by the focus group. Moreover, Cronbach’s Alpha and McDonald’s Omega are high, indicating excellent internal consistency for the dimensions. Thus, the results suggest that the dimension of digital financial knowledge can be assessed through two factors: basic digital financial knowledge, consisting of 15 items, and advanced digital financial knowledge, consisting of 8 items.

Next, the results of the factor analyses for the dimensions of behavior and confidence are presented (

Table 5. Results of the exploratory factor analyses for digital financial behavior and digital financial confidence.

Dimension	Item	Factor Loading	M.S.A	Parallel Analysis (Percentage 95%)		Alpha	Omega
Digital Financial Behavioral	24	0.814	0.77823	58.434 *	32.082	0.868	0.872
	25	0.842	0.77517	18.412	25.525
	26	0.713	0.93766	9.173	21.058
	27	0.355	0.79795	5.538	17.064
	28	0.581	0.76134	4.374	13.766
	29	0.651	0.91299	3.264	10.850
	30	0.708	0.81177	0.806	7.155
	31	0.708	0.90838
Digital Financial Confidence	32	0.732	0.91421	69.338 *	20.260	0.948	0.948
	33	0.824	0.92522	10.027	17.543
	34	0.794	0.95830	4.275	15.508
	35	0.821	0.97006	3.581	13.399
	36	0.791	0.95634	3.382	11.763
	37	0.847	0.90858	2.503	10.439
	38	0.858	0.94782	2.344	9.132
	39	0.660	0.93968	1.820	7.735
	40	0.831	0.94881	1.294	6.374
	41	0.700	0.89497	1.140	5.048
	42	0.708	0.90987	0.298	3.101
	43	0.754	0.93519

Note: * Advised number of dimensions, based on optimal implementation of the parallel analysis.

).

At the top of Table 5, the parallel analysis results indicate that the dimension of digital financial behavior is unidimensional, confirming what was proposed in the initial theoretical model. The values for Alpha and Omega suggest that the construct has good internal consistency. All items have M.S.A values above 0.5 and factor loadings above 0.3, indicating their retention.

The results at the bottom of Table 5 also indicate the formation of a digital financial confidence dimension with the initial 12 items, as all show adequate M.S.A values and factor loadings. Together, these items demonstrate excellent internal consistency.

Thus, after conducting the exploratory procedures, it is confirmed that the three dimensions initially proposed meet the evaluation criteria and should be retained in the digital financial capability scale. Subsequently, to perform confirmatory procedures, confirmatory factor analyses were conducted for the dimensions of digital financial behavior and digital financial confidence. The results from the maximum likelihood estimations are presented in

Table 6. Results of the confirmatory factor analyses for the dimensions of digital financial behavior and digital financial confidence.

Index	Bound 1	Digital Financial Behavior		Digital Financial Confidence
		Initial	Final	Initial	Final
x2 (value)	-	544.327	2.286	924.656	19.029
x2 (probability)	>0.050	0.000	0.319	0.000	0.008
GFI—Goodness of Fit	>0.950	0.826	0.999	0.810	0.992
CFI—Comparative Fit Index	>0.950	0.782	1.000	0.855	0.995
NFI—Normed Fit Index	>0.950	0.777	0.998	0.848	0.993
RMSR—Root Mean Square Residual	<0.080	0.233	0.013	0.112	0.021
RMSEA—R. M. S Error of Approximation	<0.060	0.184	0.014	0.144	0.047
Cronbach’s Alpha	>0.700		0.837		0.903
Composite Reliability	>0.700		0.847		0.901
Average Variance Extracted (AVE)	>0.500		0.588		0.602

Note: ¹ Appropriate levels for the adjustment statistics based on Hooper et al. (2008) and Hu and Bentler (1999).

.

The results indicate that the initial models did not meet the fit indices suggested by the literature. Therefore, a model refinement strategy was employed (Hair et al. 2019), which involved removing items with low factor loadings and incorporating correlations between errors.

To improve the digital financial behavior construct, items 27, 28, 29, and 30 were removed. For the digital financial confidence dimension, items 32, 33, 39, 41, 42, and 43 were removed. As a result, the final model for behavior retained four items, and the confidence model retained six items. As shown in Table 6, after these adjustments, the final measurement models demonstrated adequate fit indices (CFI, GFI, RMSR, RMSEA), indicating convergent validity of the constructs. Additionally, Cronbach’s Alpha and Composite Reliability were excellent, supporting the internal consistency of both constructs. All standardized residuals were below 2.58, indicating unidimensionality.

Regarding the Average Variance Extracted (AVE), values above 0.5 support the validity of each construct. Furthermore, the correlation between the two constructs (0.750) is lower than the square root of the AVEs, indicating discriminant validity. It is also noted that a correlation below 0.85 supports discriminant validity, as per Kline (2023). Therefore, after confirmatory analyses, evidence of convergent validity, discriminant validity, unidimensionality, and internal consistency for the constructs was obtained.

After validating the dimensions, the theoretical model of the scale was validated. In this estimation phase, variables for basic and advanced financial knowledge were created. For basic financial knowledge, the number of “yes” responses from the fifteen items in the scale was summed, and then the average was computed. Thus, basic digital financial knowledge is assessed on a scale from zero to one, with values closer to one indicating higher knowledge of the respondent. For advanced financial knowledge, the same procedure was used, i.e., the average of “yes” responses for the eight items.

Table 7. Fit indices for the DFCS.

Index	Bound *	DFCS
		Initial	Final
x2 (value)	-	369.009	211.165
x2 (probability)	>0.050	0.000	0.000
GFI—Goodness of Fit	>0.950	0.920	0.955
CFI—Comparative Fit Index	>0.950	0.938	0.967
NFI—Normed Fit Index	>0.950	0.929	0.959
RMSR—Root Mean Square Residual	<0.080	0.070	0.037
RMSEA—R. M. S Error of Approximation	<0.060	0.092	0.074

Note: * Appropriate levels for the adjustment statistics based on Hooper et al. (2008) and Hu and Bentler (1999).

presents the results of the maximum likelihood estimations.

It is observed that, in the initial model, most of the fit indices were slightly below the suggested threshold values. Therefore, the model was refined once again. At this stage, only correlations between items within the same construct that made theoretical sense were added. The relationships, values, and significance of the correlations between errors can be found in Appendix A. After these additions, the final model met the suggested thresholds, except for the Root Mean Square Error of Approximation (RMSEA), which remained slightly above the recommended value. Thus, the estimations confirm the theoretical model of the DFCS with the three dimensions, as shown in Figure 2.

Figure 2 confirms that the DFCS is a three-dimensional scale. It is observed that the constructs of digital financial behavior and digital financial confidence have coefficients significantly higher than that of digital financial knowledge. These results indicate that, for the formation of digital financial capability, knowledge is less relevant compared to behavior and confidence. These findings are similar to results obtained in financial literacy studies, where financial knowledge was shown to be less impactful (Ingale and Paluri 2022; Lusardi and Mitchell 2014; Potrich et al. 2018; Potrich et al. 2024).

5. Methodology for Applying the DFCS

Finally, we present a proposed methodology for applying the scale. It is important to note that the DFCS was designed with online administration in mind. Therefore, applications in interview or self-administered formats may require adaptations to the language of the items. Considering the validation evidence presented, we outline the methodology based only on the items that remained after confirmatory validation. However, users of the scale may choose to apply all the proposed items and adjust the calculations to incorporate them. Thus, we present the four steps below:

Step 1: With the respondents’ answers, code the items according to

Table 8. Dimensions, acronym, items and codes.

Dimension	Acronym	Items *	Codes
Basic digital financial knowledge	BDFK	Items 1 a 15	Yes = 1 No = 0 Never tried = 0
Advanced digital financial knowledge	ADFK	Items 16 a 23.
Digital financial behavior	DFB	Items 24, 25, 26 and 31	Strongly disagree = 1 Disagree = 2 Indifferent = 3 Agree = 4 Strongly agree = 5
Digital financial confidence	DFC	Items 34, 35, 36, 37, 38 and 40

Note: * The list of items can be found in Appendix B.

.

Step 2: Obtain the perceptions of each respondent (j) for each of the dimensions, based on the responses to the items belonging to each dimension. To calculate digital financial knowledge (DFK), first find the basic and advanced financial knowledge (BDFK and ADFK). Therefore, calculate the following equations:

$${BDFK}_j={\displaystyle \frac{{\sum }_{i=1}^{15}{item}_i}{15}}$$

$${ADFK}_j={\displaystyle \frac{{\sum }_{i=16}^{23}{item}_i}{8}}$$

$${DFK}_j={0.67\times BDFK}_j+{0.33\times ADFK}_j$$

$${DFB}_j={\displaystyle \frac{\left({item24}_j+{item25}_j+{item26}_j+{item31}_j\right)-4}{16}}$$

$${DFC}_j={\displaystyle \frac{\left({item34}_j+{item35}_j+{item36}_j+{item37}_j+{item38}_j+{item40}_j\right)-6}{24}}$$

Step 3: The respondent’s digital financial capability is the weighted average of the dimensions. The weighting was constructed based on the factor loadings from the final model of the scale. Thus, to obtain the DFCS of each individual, the following expression was used:

$${DFCS}_j=0.21\times {DFK}_j+0.41\times {DFB}_j+0.38\times {DFC}_j$$

Step 4: After obtaining the individual perception, the perception for the total sample can be calculated:

$${DFCS}_a={\displaystyle \frac{{\sum }_{j=1}^{n}D{FCS}_j}{n}}$$

Here, DFCS is the Digital Financial Capability Scale and n is the sample size. The values of DFCS_a and DFCS_j range from zero to one, with values closer to one indicating higher digital financial capability of the sample or individual.

6. Final Considerations

As the digitization of financial services grows rapidly, the need for individuals to have the necessary infrastructure, such as internet access and smartphones, as well as to develop new skills for making complex financial decisions in dynamic virtual environments, also increases. In this context, the potential for financial digitization to promote inclusion, rather than creating a new contingent of financially excluded citizens, will depend on the digital financial literacy of individuals.

Financial capability is considered part of human capital that helps improve wellbeing (Lusardi and Mitchell 2014; Xiao et al. 2024). Increasing consumer financial capability may help reduce consumer vulnerability (Hill and Sharma 2020), consumer anxiety (Xiao and Meng 2023), and financial fragility (Kim et al. 2024), as well as improve financial wellbeing (Fan and Henager 2022; Guo and Huang 2023) and the quality of life in many countries (Xiao and Bialowolski 2023). However, in the context of digitalization, achieving these benefits depends directly on individuals’ abilities to handle advancements in digital financial services. The implementation of extensive digitalization strategies may not yield financial benefits if people cannot adapt to these changes due to low levels of knowledge (Ferilli et al. 2024).

Thus, aiming to incorporate digital literacy into financial capability, we propose a new scale. The Digital Financial Capability Scale (DFCS) assesses an individual’s perception of their ability to apply financial knowledge, engage in appropriate financial behaviors, and feel confident in making financial decisions in a digital environment. The exploratory and confirmatory procedures applied indicated the validity of the proposed theoretical model. Therefore, the scale is composed of a combination of digital financial knowledge, digital financial behavior, and digital financial confidence.

The DFCS is a simple scale that can be very useful for researchers, governments, and financial system agents. Researchers may use it in cross-sectional and longitudinal studies to assess digital financial capability across different population groups. It can also be used in conjunction with other scales to evaluate antecedents and consequences of digital financial capability. For example, researchers may consider potential differentiated effects of cultural features on digital financial capability. In a country with high uncertainty avoidance, financial literacy interventions may encounter more barriers compared to a country low in this cultural dimension (Bialowolski et al. 2023).

From a government perspective, it can be very useful for evaluating national financial literacy policies. Evidence indicates that financial education from multiple sources has greater potential impacts on consumer financial capability (Xiao and O’Neill 2016). Additionally, international organizations (Organization for Economic Co-Operation and Development 2018) have set goals for governments, including developing programs to promote digital financial education and creating special programs for vulnerable groups, along with tools to assess outcomes. Thus, the DFCS emerges as an alternative for evaluating the implementation of these policies.

The financial system can use the DFCS to assess current and potential clients. The results of extensive investments in service digitalization depend directly on users’ digital financial capability. If users are unable to use technology effectively, the anticipated results may not be achieved, or the system’s risks may be increased. Understanding clients’ DFCS can also assist in developing marketing and relationship strategies, including creating nudges, courses, and promotional materials to enhance clients’ ability to use the systems and their confidence.

Our scale considers digital financial capability from the individual’s perception, thus not accounting for external factors. However, given evidence that financial capability may also have a dimension of financial opportunities available in the environment (Xiao and Bialowolski 2023), future research may seek to incorporate such aspects into the scale. Other factors may be incorporated into the scale. For example, product suitability, product quality, gender social norms, and financial experience. The scale also still requires cross-cultural validation and comparative testing with other scales.