📈 economics

Investigating continuance intention to use m-banking apps among older adults using partial least squares modeling with income group differences

This study investigates the factors influencing older adults' continued use of mobile banking apps in Bangladesh using PLS-SEM, revealing that system quality, trust, and perceived risk are critical drivers while highlighting significant disparities in digital financial inclusion experiences between lower- and higher-income groups.

Original authors: Hamida Akhter, Abu Naser Mohammad Saif, Nusrat Jafrin, Rasheda Akter Rupa, Francesca Dal Mas, Maurizio Massaro

Published 2026-07-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Hamida Akhter, Abu Naser Mohammad Saif, Nusrat Jafrin, Rasheda Akter Rupa, Francesca Dal Mas, Maurizio Massaro

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a world where your bank is always in your pocket, accessible with just a tap on a screen. This is the promise of m-banking apps. While younger generations have embraced this technology like fish in water, older adults (people aged 60 and above) often find themselves standing on the shore, hesitant to dive in.

This research paper is like a detective story trying to figure out why some older people in Bangladesh keep using these banking apps, while others stop after the first try. The researchers wanted to know: What makes an older person say, "I'll keep using this," versus "This is too much trouble"?

Here is the story of their investigation, broken down into simple concepts.

The Cast of Characters (The Theories)

To solve the mystery, the researchers didn't just guess; they used three well-known "rulebooks" from the world of technology and psychology, mixing them together like a special recipe:

The IS Success Model: This checks if the "machine" works well (Is the app fast? Is the info clear?).
The Expectation Confirmation Model (ECM): This asks, "Did the app do what I hoped it would?" If you expect a car to drive smoothly and it does, you are happy. If it stalls, you are not.
The UTAUT Model: This looks at human factors like "Is it easy to use?" and "Did my friends tell me to use it?"

They also added two special ingredients crucial for older adults: Trust (Do I believe this app won't steal my money?) and Perceived Risk (Am I scared I'll get hacked?).

The Investigation (The Method)

The researchers went into the bustling city of Dhaka, Bangladesh, and interviewed 160 older adults (60+ years old) who already knew how to use these apps. They asked them to rate their experiences on a scale, similar to how you might rate a restaurant after a meal.

They also split the group into two teams to see if money mattered:

Team Lower-Income: Those earning less than 50,000 Bangladeshi Taka a month.
Team Higher-Income: Those earning more.

The Big Discoveries (The Results)

1. The "Reliability" Rule (System Quality)

Think of the app as a bus. If the bus is late, breaks down, or gets stuck in traffic, you stop taking it.

Finding: The most important thing for older adults was that the app worked reliably. If the app was fast, didn't crash, and was easy to navigate, they felt their expectations were met.
Surprise: Surprisingly, how "pretty" the information looked or how fast customer service answered didn't matter as much as the app simply working. Older adults just wanted the "bus" to run on time.

2. The "Safety" Factor (Trust & Risk)

Imagine walking into a dark alley. If you trust the neighborhood, you walk in. If you think there's a mugger, you run away.

Finding: Trust was a huge booster. If an older adult trusted the bank, they were more likely to keep using the app.
The Fear: Perceived Risk was the biggest brake. If they felt even a little bit of danger (like losing their savings to a scam), they immediately stopped believing the app was a good idea. For older adults, the fear of losing hard-earned money is a massive barrier.

3. The "Friends" Myth (Social Influence)

You might think, "If my grandkids tell me to use it, I will."

Finding: The researchers found that friends and family didn't really matter once the person started using the app. Older adults relied on their own experience, not what others said. If the app worked for them, they stayed; if not, they left, regardless of what their neighbors thought.

4. The "Easy to Use" Paradox

Usually, we think "easy to use" is the most important thing.

Finding: For this specific group, "ease of use" wasn't the main driver for staying with the app. Why? Because if the app was reliable and safe, they felt it was "good enough." They weren't looking for a Ferrari; they just wanted a bicycle that didn't fall apart.

The Money Twist (Income Differences)

This is where the story gets interesting. The researchers asked: "Does having more money change how people feel about the app?"

The Result: Yes, but only in specific ways.
- Higher-Income Seniors: They were like critics. They cared deeply about the quality of the information (is the data detailed?), the system (is it fast?), and the service (is support good?). Because they used the apps for more complex things, they noticed every little flaw.
- Lower-Income Seniors: They were like survivors. They cared mostly about the basics. As long as the app didn't crash and they could send money, they were happy. They didn't demand high-end features.

The Bottom Line

This paper tells us that to keep older adults using m-banking apps, banks shouldn't just try to make the app "fancy" or rely on friends to convince them. Instead, they need to focus on three things:

Make it work perfectly: No crashes, no delays.
Make it feel safe: Prove that money won't disappear.
Reduce the fear: Show them that the risk of scams is low.

If banks can build a "reliable and safe bus" for older adults, those passengers will keep riding, helping them stay connected to the modern economy.

Technical Summary: Continuance Intention to Use m-Banking Apps Among Older Adults in Bangladesh

Problem Statement
While mobile banking (m-banking) applications significantly contribute to digital financial inclusion and the United Nations Sustainable Development Goals (SDGs 1, 8, 9, and 10), sustained usage among vulnerable populations, specifically older adults, remains limited. In Bangladesh, despite rapid growth in smartphone penetration and m-banking adoption, there is a lack of empirical research regarding the factors influencing the continuance intention of older adults (aged 60+). Existing literature predominantly focuses on initial adoption or general user bases, often overlooking the specific post-adoption dynamics, risk perceptions, and socioeconomic heterogeneity of the elderly demographic in developing economies. This study addresses the gap by investigating the antecedents of continuance intention among older adults in the Dhaka metropolitan area, with a specific focus on how income levels moderate these relationships.

Methodology
The research employs a quantitative approach utilizing Partial Least Squares Structural Equation Modeling (PLS-SEM).

Theoretical Framework: The study integrates three established models: the Information Systems (IS) Success Model (DeLone & McLean, 2003), the Expectation Confirmation Model (ECM) (Bhattacherjee, 2001), and the Unified Theory of Acceptance and Use of Technology (UTAUT) (Venkatesh et al., 2003). Additionally, Trust and Perceived Risk are incorporated as contextual variables critical for the elderly demographic.
Data Collection: A purposive non-probability sampling technique was used to recruit 160 valid respondents aged 60 and above with prior m-banking experience in Dhaka, Bangladesh. Data was collected via face-to-face surveys and online forms over a four-month period (June–September 2024).
Analytical Approach:
- Measurement Model: Validated using Composite Reliability (CR), Average Variance Extracted (AVE), Fornell-Larcker criterion, Heterotrait-Monotrait (HTMT) ratio, and cross-loadings to ensure reliability and discriminant validity.
- Structural Model: Path coefficients were estimated using bootstrapping (5,000 subsamples) to test hypotheses.
- Multigroup Analysis (MGA): To address income-based heterogeneity, respondents were divided into low-income (monthly income < BDT 50,000) and high-income (monthly income ≥ BDT 50,000) groups. Measurement Invariance of Composite Models (MICOM) was conducted to ensure comparability before performing parametric and Welch-Satterthwaite tests for path differences.

Key Results
The empirical analysis yielded the following findings regarding the drivers of confirmation, satisfaction, and continuance intention:

Significant Drivers of Confirmation:
- System Quality (SYQ): Positively and significantly impacts confirmation ( $\beta$ = 0.128, $p$ < 0.05).
- Trust (TR): Positively and significantly impacts confirmation ( $\beta$ = 0.206, $p$ < 0.01).
- Perceived Risk (PR): Negatively and significantly impacts confirmation ( $\beta$ = -0.260, $p$ < 0.001).
- Non-Significant Factors: Information Quality, Service Quality, Effort Expectancy, Performance Expectancy, and Social Influence were found to have no statistically significant impact on confirmation in the overall model.
The Continuance Chain:
- Confirmation significantly and positively influences Satisfaction ( $\beta$ = 0.594, $p$ < 0.001).
- Satisfaction significantly and positively influences Continuance Usage Intention ( $\beta$ = 0.568, $p$ < 0.001).
Multigroup Analysis (Income Differences):
- The study identified that income acts as a selective moderator. While the core causal chain (Confirmation $\to$ Satisfaction $\to$ Intention) remains consistent across groups, the influence of IS quality dimensions on confirmation differs significantly between income groups.
- Information Quality $\to$ Confirmation: Significant difference between groups ( $p$ = 0.005/0.006).
- System Quality $\to$ Confirmation: Significant difference between groups ( $p$ = 0.001/0.002).
- Service Quality $\to$ Confirmation: Significant difference between groups ( $p$ = 0.022/0.023).
- The remaining paths (including Trust, Perceived Risk, and UTAUT constructs) did not show significant differences between low and high-income groups.

Key Contributions

Theoretical Integration: This study is among the first to unify the IS Success Model, ECM, and UTAUT within a single framework specifically tailored to older adults in a developing country context. It theoretically positions UTAUT constructs as antecedents of confirmation rather than direct drivers of continuance, aligning with post-adoption logic.
Contextual Nuance: By focusing on the elderly in Bangladesh, the research highlights that for this demographic, confirmation is driven primarily by functional reliability (System Quality) and safety (Trust/Risk), rather than by the perceived ease of use or social influence often cited in general adoption studies.
Equity and Digital Divide: The study provides empirical evidence of socioeconomic heterogeneity in digital financial inclusion. It demonstrates that while the psychological mechanism of satisfaction leading to continuance is universal, the antecedents related to information and system quality are income-dependent, suggesting that lower-income elderly users may have different thresholds for what constitutes "quality" or "confirmation."

Significance and Claims
The authors position this study as a contribution to the sustainable development agenda by identifying the specific behavioral conditions required to achieve long-term digital financial inclusion for older adults. The paper claims that:

Improving system quality, fostering trust, and mitigating perceived risk are critical levers for enhancing the continuance intention of older adults.
Financial institutions and policymakers can use these findings to design inclusive digital services that support equitable economic participation.
The findings offer actionable insights for tailoring m-banking services to different income segments, noting that higher-income elderly users place distinct value on information and service quality dimensions compared to their lower-income counterparts.

The study explicitly acknowledges limitations, including its cross-sectional design, reliance on self-reported data, geographic restriction to Dhaka, and the use of purposive sampling. Consequently, the authors frame their conclusions as indicative of the specific context studied, calling for future longitudinal and multi-regional research to validate these dynamics further.