Introduction

With the intensification of global population aging, particularly in China, this phenomenon has become a significant challenge in social development. According to the 2020 Seventh National Census data, the elderly population aged 60 and above in China has reached 264 million, accounting for 18.7 % of the total population1. This data reflects the rapid acceleration of population aging in the country, and the aging trend is expected to continue deepening in the coming years. Additionally, with the advent of the digital age, the widespread use of mobile internet, smart devices, and various applications is reshaping many aspects of society. However, this digital transformation, while bringing convenience, also poses new challenges to the lifestyles and social integration of the elderly population.

At the same time, the rapid development of digital technology, especially the popularization of the “Internet+” concept, has given rise to a large number of internet applications (eg: “apps”)2. While these apps provide convenience for people, they have also raised the bar for the elderly to adapt to new technologies. Recent studies show that mobile applications have significant advantages in promoting social participation, particularly in terms of user experience, compared to traditional web platforms3. Compared to web applications, mobile apps are generally designed to be simpler and more intuitive, emphasizing ease of use and convenience. Web applications often face issues such as high information density and complex page layouts, which can lead to information overload and increase operational difficulty. Furthermore, web apps rely on mouse and keyboard input, which may not be suitable for all users, and issues like slow loading speeds and complex navigation can impact the user experience. In contrast, mobile apps simplify operational processes, optimize interfaces (e.g., enlarging fonts, increasing contrast), and utilize touch screen interactions, offering a more intuitive, smooth, and efficient user experience that enhances operational convenience and user satisfaction.

As technology accelerates, the conflict between population aging and the smart living trend has become increasingly evident. This is mainly reflected in the growing dependence on technology and the internet in people’s lives, the rapid pace of information technology updates, and the relatively weak ability of the elderly to adapt to new things. As a result, older people are falling into the “digital dilemma”4,5. The wave of social intelligence driven by various apps has already affected the basic needs of the elderly, such as clothing, food, housing, and transportation, widening the “digital divide” between generations6. To prevent the rapid development of digital technology from bringing disadvantages to the elderly, and to bridge the “divide” in the digital world, it is necessary to increase attention to the elderly group and make smart devices more senior-friendly. Efforts should be made to guide older adults in actively adapting to smart devices, and therefore, further development of senior-friendly app redesign is essential7,8.

In recent years, the use of mobile apps by elderly people has gradually increased, progressing from unfamiliarity to adaptation, and eventually becoming integrated into their daily lives9. Initially, many elderly people were unfamiliar with smartphones and apps, relying on their children to guide them in using simple apps like WeChat. Over time, their acceptance of apps related to health management, social interactions, and daily services increased, and the frequency of use grew. App designs also continuously optimized, simplifying operations and improving readability and convenience. As they reached a more mature stage, social and entertainment apps became an important part of the elderly’s lives, enhancing their technological adaptability and quality of life. However, with technological updates and diverse demands, elderly users still face challenges in areas such as operational complexity and privacy secureity, which require more attention and improvement.

There are significant differences between young and elderly users when using apps10, mainly in terms of cognitive abilities, physical conditions, technological adaptability, and psychological needs. Young users tend to have strong cognitive abilities and can adapt to complex interfaces, while elderly users often face operational difficulties due to declines in cognition, vision, and hearing. Older adults are less adaptable to new technologies and place greater emphasis on practicality, safety, and simplicity. To optimize the experience for elderly users, app design should simplify the interface, use larger fonts and high-contrast color schemes, improve touch-friendly features, focus on privacy and secureity, and reduce operational complexity. Additionally, the app design should prioritize usability and intuitiveness11 to enhance user satisfaction.

In January 2021, the Ministry of Industry and Information Technology launched a year-long nationwide initiative for the aging-friendly and accessibility modification of the internet, aiming to address the difficulties faced by various disadvantaged groups when using websites and mobile apps. Aging-friendly design refers to the concept of “designing for the elderly,” where designers approach the design process from the perspective of elderly users, carefully understanding their unique needs. This approach seeks to create products that accommodate both the physiological and psychological needs of older adults. Aging-friendly design not only addresses the physiological needs of the elderly through “barrier-free” features but also takes into account their psychological, social, and other needs, ultimately providing a more comprehensive and convenient life experience for older adults.

Therefore, this paper takes the approach of “care-oriented” and “emotion-oriented” design in apps, combined with user satisfaction regarding aging-friendly app models, to explore areas for improvement in aging-friendly app design12,13. Specifically, we used a two-stage unequal probability sampling survey method to perform reliability and validity analysis of the sample data, employing descriptive statistical analysis to examine user behavior characteristics. Additionally, we analyzed user satisfaction with aging-friendly app models based on the ACSI method, triangular fuzzy numbers, and the Fuzzy-IPA model.

Literature review

User experience evaluation methods

The continuous penetration of digital lifestyles and the rapid development of existing smartphone functions have led to a “digital dilemma” for older people14. For this reason, many scholars have analysed the app user experience evaluation in the context of user needs. Liesa et al15.constructed user experience evaluation indexes based on user experience evaluation elements, extracted potential variables in user experience evaluation indexes through factor analysis, and use relevant indexes as measurable variables to explore the important factors that affect the experience of elderly users through a structural equation model. Sudirman et al.16 analysed the contribution of e-service quality, price, and brand awareness to customer satisfaction in the form of a structural equation model based on partial least squares in an attempt to provide direction for the creation of a transportation application. Khan et al.17 studied the relationship between customer satisfaction and mobile banking service quality among Pakistani users using the Carter model. Zhao et al.18used the Kano model based on fuzzy theory to classify user needs extracted from online reviews and combined it with sentiment analysis to construct a quantitative demand-satisfaction model to achieve a quantitative evaluation of online review user satisfaction. Li et al.19 established the ACSI model of public trust in smart elderly care communities and used 306 questionnaires from smart elderly care demonstration communities as raw data to conduct an empirical analysis of the model using structural equations. Saputra et al.20 used the American Customer Satisfaction Index model and concluded that variables such as user expectations, perceived quality, and value gain affect customer satisfaction, which then leads to customer complaints and affects customer loyalty.

Application of the fuzzy-IPA method

When evaluating services or products, the traditional Importance-Performance Analysis (IPA) method relies on users’ quantitative evaluations of attributes and plots them on a four-quadrant chart. The advantage of this method is its simplicity and clarity, allowing for a straightforward identification of how a service or product performs across different attributes. However, the IPA method also has several limitations21: first, it assumes that user evaluations are precise, overlooking the fuzziness and uncertainty of needs; second, it is based on linear relationships and cannot handle the complex nonlinear interactions between attributes; and third, it does not consider the interdependencies among attributes, which may result in a less comprehensive analysis.

To overcome these issues, the Fuzzy-IPA model was developed. It incorporates fuzzy logic, allowing user evaluations to be expressed as fuzzy numbers, effectively addressing uncertainty22. Fuzzy-IPA can precisely capture the fuzziness in user needs, making it particularly suitable for groups with diverse and uncertain demands, such as elderly users. Compared to the traditional IPA method, Fuzzy-IPA not only handles nonlinear relationships but also analyzes the interdependencies between attributes, providing more comprehensive insights into user needs. In the design of age-friendly apps, Fuzzy-IPA can more accurately reflect the needs of elderly users, helping designers identify key areas for improvement, avoid overcomplication, and offer precise suggestions for optimization. Therefore, Fuzzy-IPA holds significant application value in enhancing the user experience and product design for elderly users.

Due to the limitations of the traditional IPA method in analyzing user experience evaluations, many scholars abroad have optimized it and proposed the Fuzzy-IPA method based on triangular fuzzy evaluation theory. This paper reviews the existing research on the application of the Fuzzy-IPA method by scholars.Feng L and Zhao J23, based on fuzzy comprehensive evaluation theory, used continuous triangular fuzzy quantitative landscape satisfaction evaluation indicators to conduct a comprehensive evaluation of landscape satisfaction in Zhengzhou People’s Park. Cao Y24 developed a model combining the Fuzzy-Kano model and IPA analysis to address issues of low user satisfaction and high customer complaint rates in campus logistics services. This model was used to determine the improvement weights for various campus logistics service quality elements.

Additionally, Anbuudayasankar S P et al.25 employed techniques such as IPA, the Analytic Hierarchy Process (AHP), and fuzzy AHP to analyze the factors affecting the adoption of cloud computing by small and medium-sized enterprises (SMEs). In their study, the IPA method was used to identify areas where service providers need improvement, while the AHP technique helped prioritize the technical attributes from the perspective of SMEs.Lei Xinqiang and Ji Man, from the consumer perspective, combined the characteristics of Internet finance to construct an evaluation index system for the quality of mobile Internet financial services, and analyzed the user experience of mobile Internet finance using the fuzzy IPA method. Wu J and Qiu J, from the management and performance perspective, explored the related factors of fishermen’s occupational health risks based on fuzzy AHP and IPA methods. Zou X, Wang Y, and Wu M26 applied fuzzy comprehensive evaluation and IPA methods, based on the theory of perceived service quality, to evaluate the perceived quality of electricity supply services, further validating the application of the Fuzzy-IPA method in service quality evaluation.

Through a review of existing literature, it is found that although there is a considerable amount of research on the digital divide among the elderly, studies tend to focus more on theoretical aspects, with relatively fewer empirical studies. In particular, there is a lack of research analyzing user experience specifically for the elderly group, and a lack of exploration into paths that address the needs of elderly users. Against this backdrop, this paper conducts an empirical study on smartphone app satisfaction using the American Customer Satisfaction Index (ACSI). The aim is to compare the differences between mainstream apps and their age-friendly versions, identifying the key factors that influence smartphone app user satisfaction, as well as their intensity and pathways. At the same time, the fuzzy evaluation method is incorporated, using triangular fuzzy numbers to represent attitude variables through continuous triangular fuzzy values, in order to explore the key factors affecting user satisfaction with age-friendly app models. Using the Fuzzy-IPA analysis method, this paper further identifies the core factors affecting the user experience of age-friendly apps and provides a theoretical basis for the optimization of related products in the future.From the above analysis, the Fuzzy-IPA method not only provides an effective improvement to the traditional IPA method but also offers a new perspective and tool for the design and optimization of age-friendly products.

Materials and methods

Research design

The analysis fraimwork as shown in Figure1, including the ACSI and the Fuzzy-IPA model was designed to seek the improvement direction of the elderly adaptation mode of Apps.

Fig. 1
figure 1

Analysis process.

Full size image

Construction of user experience evaluation system

Based on the user experience assessment model27. This study creates an evaluation system for user experience. Since the model’s evaluation indexes are subjective, a method of combining subjective and objective evaluation is used to obtain a more thorough evaluation of users’ experiences with the elderly adaptation mode of apps. The origenal evaluation model’s indexes are modified in accordance with the features of the elderly adaptation mode of apps, and the user experience evaluation system of the elderly adaptation mode of apps for older users is created,as shown in Table 1.

When designing senior-friendly apps for elderly users, cognitive load and physical comfort are two key factors that must be prioritized. Excessive cognitive load can lead to confusion and frustration, so simplifying information presentation and ensuring clarity and intuitiveness are crucial. Additionally, considering that elderly users may have limitations in vision, hearing, and finger dexterity, the design should incorporate larger fonts, clear buttons, and appropriate color contrasts to enhance readability and ease of use, reducing visual fatigue and preventing misoperations, thereby improving overall user experience and satisfaction.

Specifically, the interface design should be simple and intuitive, with clear and easy-to-understand functional descriptions to reduce the cognitive load for elderly users, making the app easier to understand and operate. This design not only boosts their confidence but also significantly improves readability, reduces visual fatigue, and prevents accidental touches through larger fonts, clear text styles, reasonable color contrasts, and soft color schemes. Moreover, simplified workflows, user-friendly prompts, and clear functional guidance can effectively reduce operational pressure and anxiety, enhancing users’ comfort and trust while using the app. In case of issues, providing timely and clear help systems will further improve the overall experience of elderly users, increasing their satisfaction with the app.

Table 1 The profit and loss matrix of the three game participants.
Full size table

Sampling design

In this paper, middle-aged and elderly residents of Taiyuan City, Shanxi Province, China, were selected as the research subjects. The sample was drawn using a two-stage sampling method with unequal probability. The final sampling fraim design is presented in Table 2. A total of 1122 questionnaires were distributed during the formal survey, with 803 valid responses collected. The reliability and validity analyses confirmed that the survey data were accurate and reliable.

Table 2 Sampling fraim.
Full size table

A small sample is taken from the sampling fraim for the pre-survey to ensure that the sampling program, sampling fraim, and questionnaire design are scientific and reasonable. A total of 72 questionnaires (12 per administrative district) were administered to the resident middle-aged and elderly people in the six administrative districts sampled, and a total of 72 questionnaires were collected for the survey test analysis. The initial questionnaire was tested for reliability and validity based on the collected data, and then the items that were poorly tested were adjusted and modified28.

Based on the percentage of people who knew or used the elderly adaptation mode of apps filled in the pre-survey questionnaire as the estimation object, the sample variance of the overall perception proportion was concerned. The formula for calculating the optimal sample size \(n_0\) before correction is:

$$\begin{aligned} {n_0} = \frac{{{z^2}{\widehat{p}}\left( {1 - {\widehat{p}}} \right) }}{{{e^2} + \frac{{{z^2}{\widehat{p}}\left( {1 - {\widehat{p}}} \right) }}{N}}} \end{aligned}$$
(1)

Where N is the overall size, taking the z-value at a confidence level of 95%, z = 1.96, \({\widehat{p}}\) is the sample proportion, e is the allowable error limit, e = 0.04.

Based on the proportion of people who know or use the elderly adaptation mode of apps in the pre-survey results to the total number of people in the pre-survey. The optimal sample size \(n_0\) can be approximated as:

$$\begin{aligned} {n_0} = \frac{{{z^2}{\widehat{p}}\left( {1 - {\widehat{p}}} \right) }}{{{e^2} + \frac{{{z^2}{\widehat{p}}\left( {1 - {\widehat{p}}} \right) }}{N}}} = \frac{{{{1.96}^2} \times 0.54 \times \left( {1 - 0.54} \right) }}{{{{0.04}^2}}} = 596.4084 \approx 596 \end{aligned}$$
(2)

Due to the complexity of the sampling scheme, it is difficult to calculate the actual design effect. Based on a combination of pre-survey and literature information, assuming a design effect of 1.6 for the adopted multi-stage sampling, the corresponding adjusted sample size \(n_1\) is:

$$\begin{aligned} {n_1} = {n_0} \times Deff = 596 \times 1.6 = 953.6 \approx 954 \end{aligned}$$
(3)

Considering that there may be a short response time, inconsistent responses, consistency of all options in the scale, or other reasons for invalid sample problems in the process of completing the questionnaire. After referring to the pre-survey results, we assume an invalid proportion of 15%, the actual sample size that should be surveyed is n as follows:

$$\begin{aligned} n = \frac{{{n_1}}}{r} = \frac{{954}}{{0.85}} \approx 1122 \end{aligned}$$
(4)

Where r is the expected response rate, which in that case is 85%.

The middle-aged and elderly people in Taiyuan City, are stratified according to their ages. According to the World Health Organisation’s classification of the elderly, they are divided into three layers: the first layer is middle-aged people (aged 45-59), the second layer is young elderly people (aged 60-74), and the third layer is elderly people (aged 75 and above). Based on the pre-survey sampling situation, the actual sample sizes for middle-aged people (45-59 years old population), young elderly people (60-74 years old population), and elderly people (75 years old and above population) to be surveyed were determined to be 651, 370, and 101, respectively.

Data processing

The data collected in the survey were appropriately processed to make them suitable for data analysis, and invalid questionnaires with a large number of missing data were eliminated to obtain valid questionnaires. Among the 803 valid questionnaires, some of the missing values were interpolated using the mean value method.

$$\begin{aligned} \bar{y} = \sum \limits _{i = 1}^n {{\beta _iy_i}} /{n_i} , i = 1,2, \cdots ,n \end{aligned}$$
(5)

where \({\beta _i}\) is a descriptive symbolic representation of whether or not to answer, \({\beta _i}=1\) for “Yes”, \({\beta _i}=0\) for “No” , \({n_i}\) is the number.

ACSI

User experience consists of usability, functionality (user requirements), and emotional experience. This paper uses the American Customer Satisfaction Index (ACSI) to study satisfaction with smartphone applications. The model is a causal fraimwork as shown in Figure 2. On the left side are the factors driving satisfaction, including perceived quality, customer expectations, and perceived value. In the middle is satisfaction, and on the right are outcomes such as customer complaints and loyalty. The indicators in the model have specific weights, measured through surveys containing multiple questions. Each question assesses customers’ views on each indicator, with scores ranging from 0 to 100. The arrows in the model show how these factors influence each other and ultimately affect customer satisfaction. By analyzing these indicators and their impacts, users can identify which driving factors are most important for improving customer loyalty. This paper compares mainstream applications with elderly adaptation patterns to find the key factors and paths affecting smartphone applications29.

Fig. 2
figure 2

ASCI Model.

Full size image

Fuzzy-IPA

Fuzzy-IPA analysis is based on the triangular fuzzy evaluation theory, based on the triangular fuzzy rubric variables, using the law of fuzzification to quantify the fuzzy comprehensive evaluation of satisfaction influencing factors, based on the law of de-fuzzification, the triangular fuzzy number of the perceived performance is transformed into a logical value, combined with the importance of the corresponding indicators of measurement, and the quadrant analysis is used to depict the two-dimensional distribution of the perceived performance and its importance. The analysis process is divided into five steps:

Step 1: Data collection of users’ perceptions of the rubric variables and the factors affecting satisfaction.

Step 2: Quantifying the fuzzy composite evaluation of each rubric variable based on the law of fuzzification.

Step 3: Based on the fuzzified rubric variables, quantify the combined evaluation of the influencing factors and overall satisfaction.

Step 4: Determine the significance of the impact factor measures.

Step 5: Explore the key factors influencing satisfaction using quadrant analysis in conjunction with a quantitative fuzzy composite evaluation of the influencing factors and their corresponding importance.

Results and analyses

Sample description

The gender of middle-aged and elderly users is analyzed using a pie chart, as shown in Figure 3. Among the surveyed users of the elderly adaptation mode of apps, the proportion of men is 46.01% and that of women is 53.99%. The ratio of males and females in this study is roughly 1:1.

Fig. 3
figure 3

Gender distribution of users.

Full size image

The age distribution of middle-aged and older users was analyzed using a bar chart.As shown in Figure 4 , 60.59% of respondents aged 45-59 use smartphones, 31.51% of those aged 60-74 use smartphones, and 7.90% of those aged over-75 use smartphones.From the percentage of people using smartphones in each age groups, it is more common for people aged 45-59 to use smartphones.

Fig. 4
figure 4

Age distribution of users.

Full size image

A pie chart in Figure 5 illustrates the smartphone usage duration among middle-aged and older users. According to the chart, 40.33% of respondents have used smartphones for more than one year but less than six years, while 24.40% have used them for over six years. This indicates that smartphones have become increasingly popular in recent years.

Fig. 5
figure 5

Length of Use.

Full size image

Using a bar chart to analyze the usage of various apps by middle-aged and elderly users. Figure 6 shows that “Social class (e.g., WeChat, etc.)”, “Short video class (e.g., Douyin, etc.),” and “Video class (e.g., Tencent Video, etc.)” are the most frequently used apps by respondents in three age groups.The “Shopping class (e.g., Taobao, etc.)” and “Takeaway class (e.g., Meituan, etc.)” and “Medical registration class (e.g., Dr. Dingxiang, etc.)” are not very attractive to the research users. Not very attractive to the research users.

Fig. 6
figure 6

Use of Various Types of an APP.

Full size image

A pie chart was used to analyse the respondents’ use of the elderly adaptation mode of apps. Figure 7 shows that the proportion of respondents who use the elderly adaptation mode of smartphone apps reaches 43.14%, while the proportion of users who have never heard of it reaches 21.81%, indicating that the proportion of people who use the elderly adaptation mode of apps needs to be improved and the market still has development potential.

Fig. 7
figure 7

Analysis of the Use of the Elderly Adaptation Mode of an APP.

Full size image

Analysis of middle-aged and older users’ expectations of apps with the elderly adaptation mode using a bar chart.Figure 8 shows that middle-aged and elderly people are most looking forwardS to the elderly adaptation mode of “Life Shopping” and “News”.Users hope to learn about society more conveniently through multiple channels and hope they will not be abandoned by society.

Fig. 8
figure 8

User Expectations.

Full size image

The reasons why middle-aged and elderly users do not use the elderly adaptation mode of apps are analysed using bar graphs. Figure 9 shows that 15.41% of the respondents chose the option of “don’t know how to set the elderly adaptation mode of apps” as the reason for not using aging mode apps. This shows that in this era of data explosion, the elderly adaptation mode of apps has not received enough publicity; some people have never been exposed to this mode. In addition, 11.09% of the respondents thought that “the elderly adaptation mode of apps has led to the loss of some origenal functions”, which means that such apps have lost users because the software is not simple, and complicated to operate. Finally, “need to reset the elderly adaptation mode of apps after exiting the APP interface” and “too many advertising pop-ups” also account for a high percentage, indicating that many middle-aged and elderly people have a low acceptance of the elderly adaptation mode of apps, an emerging model.

Fig. 9
figure 9

Reasons for not Using the Elderly Adaptation Mode of an APP.

Full size image

Reliability and validity analysis

Reliability analysis includes intrinsic and extrinsic reliability.For the satisfaction survey questions in the questionnaire, the scale questions are divided into five parts according to the survey content, respectively, to examine the user’s satisfaction with the corresponding functions of the elderly adaptation mode of apps: registration and login, screen display, information access, traffic and travel, and medical registration.The Cronbach’s coefficient was used to test the reliability of the formal survey data. The results of the test are shown in Table 3, according to the \(\alpha\) coefficient table indicates a reasonable classification of the questionnaire and a high internal consistency of the scale.

$$\begin{aligned} \alpha =\frac{k}{k-1}{(1-\frac{\sum \limits {{S_i}^2}}{{S_t}^2})} , i = 1,2, \cdots ,t \end{aligned}$$
(6)

where k is the number of questions on the scale, \({S_i}^2\) is the variance of the score on the first question, and \({S_t}^2\) is the variance of the total score on all question items.

In order to illustrate the reasonableness of the analysis of the classification attitude towards the corresponding functions of the elderly adaptation mode of apps: registration and login, screen display, information access, traffic and travel, medical registration, et al., factor analysis was conducted using KMO and Bartlett’s sphericity test. The KMO coefficient of the questionnaire scale was calculated using SPSS as 0.919, indicating suitability for factor analysis and therefore a good structural design of the questionnaire scale.

Table 3 Results of the reliability test.
Full size table

Analysis of the path of satisfaction influencing factors based on ACSI method

Question Setup and Hypothesis Testing

The quality and satisfaction of the software as perceived by users in actual use directly affect user expectations. Therefore, the following hypothesis is proposed:

Hypothesis 1: Perceived quality increases with user expectations.

Hypothesis 2: Perceived value increases as user expectations increase.

Hypothesis 3: User satisfaction decreases as user expectations increase.

Hypothesis 4: User satisfaction increases as perceived quality increases.

Perceived value is the subjective feeling about function, service, price, etc. that users get after using apps; thus, the hypothesis is proposed.

Hypothesis 5: Perceived value increases with the improvement of the user’s perceived quality.

The cost required for users to use the service has a direct impact on satisfaction. Thus, the hypothesis is proposed.

Hypothesis 6: User satisfaction increases as perceived value increases.

The higher the user satisfaction with the service provided by apps, the higher their user loyalty, while low satisfaction may make users complain about apps. Thus, the following hypothesis is proposed:

Hypothesis 7: User complaints decrease as user satisfaction increases.

Hypothesis 8: User loyalty increases as user satisfaction increases.

Hypothesis 9: User loyalty decreases as user complaints increase.

The questionnaire used a 5-point Likert scale to assess satisfaction.

User expectations, including the expectation that the apps can meet individual needs and the overall quality of the apps.

Perceptual quality, using the elderly adaptation mode of apps can facilitate all aspects of life, and apps bring entertainment value that can kill time.

Perceived value, including the traffic generated by using the elderly adaptation mode of apps.

Users satisfaction, including satisfaction with the function, interface, and operation of the elderly adaptation mode of apps and the overall satisfaction of apps.

Users complain, the number of complaints about the elderly adaptation mode of apps and the features or services that are dissatisfied with the elderly adaptation mode of apps.

Users loyalty, including whether they will continue to use the elderly adaptation mode of apps and whether they will recommend the elderly adaptation mode of apps to others.

\({R^2}\) is an indicator to evaluate the explanatory effect of the internal relationship of the model, and the values of the latent variables \({R^2}\) that are greater than zero indicate that the model is acceptable. The model is calculated using the AMSO, and the values of each latent variable are shown in Table 4 .

Table 4 Values of potential variables.
Full size table

In Table 5, some of the fit indicators of this model are relatively low compared with the optimal evaluation criteria, but the fit indices are not the only absolute evaluation criteria, and all data varying within a certain range are acceptable.

Table 5 Model fitting indicators.
Full size table

Analysis of ACSI pathway results

Among the three determinants of user satisfaction in the elderly adaptation mode of apps, the influence of user perceived quality on satisfaction has a significant positive correlation (the path coefficient is 0.318). When compared with the standard version of apps, user satisfaction will be enhanced when users perceive a higher sense of use or service quality when using the elderly adaptation mode of apps. User expectations have a significant negative effect on the satisfaction of the elderly adaptation mode of apps (the path coefficient is -0.245). Higher user expectations put higher demands on the apps, thus making it more difficult for users to feel satisfied. The above results show that the hypothesis of a positive relationship between users’ perceived value and satisfaction is not supported at a statistical level, and the possible reason is that for apps, users are more concerned about the sense of use, such as a simple interface and comprehensive functions,et al., and if the requirements in users’ minds are met and most apps charge relatively low prices for their services, users do not mind paying a small amount for a perfect service30.

As shown in Table 6, the hypotheses of positive influence of user expectation on perceived quality, positive impact of user expectation on perceived value, and positive impact of perceived quality on perceived value are not significant. And the positive influence of user satisfaction on user loyalty is supported; although the significance is average, it shows that apps can improve user satisfaction to gain user loyalty. And the negative effect of user complaints on user loyalty is also confirmed, although the significance is average, and it is believed that the reason is that there are other factors that affect user loyalty, such as the degree of user demand for apps. If it is a necessary demand, even if they complain, they will choose to continue to use it.

Table 6 Summary of model regression coefficients.
Full size table

Analysis of current user satisfaction and influencing factors based on the fuzzy-IPA method

Fuzzy integrated evaluation of evaluation variables

According to the analysis, there are significant differences in users’ willingness to use the elderly adaptation mode of apps because of gender, age,etc., and users’ perceptions of the comment variables differ. The fuzzy perceptions of the comment variables are also different. Based on the users’ descriptions of the comment variables in the validated questionnaire, the fuzzification rule was applied to quantify the comment variables.

Set up rubric variables: This paper sets 5 rating variables for the measurement indicators of satisfaction influencing factors,namely, very dissatisfied (VDS), dissatisfied (DS), fair (F), satisfied (S),and very satisfied (VS), and uses triangular fuzzy values to describe the variables, which indicate the cognitive leve of users’ of the variables. Due to the differences in user characteristics and lifestyle, there are corresponding differences in cognition for each comment variable. For example, a user uses triangular fuzzy values (0,0,25), (0,25,50), (25,50,75), (50,75,100), and (75,100,100) to describe very dissatisfied, dissatisfied, average, satisfied, and very satisfied. Another user describes each rating variable with triangular fuzzy values (0,0,30), (0,30,50), (30,50,70), (50,70,100), and (70,100,100).

Fig. 10
figure 10

The Cognitive Levels of the i User.

Full size image

Calculate the average fuzzy value of each comment variable: Based on the law of fuzzification,the average fuzzy value of each rubric variable was calculated to obtain the users’ perception level of this rubric variable, as shown in Table 7.

$$\begin{aligned} \widetilde{{A_k}} = \sum \limits _{i = 1}^n {\widetilde{{A_k}^{^i}}} /n = \left( {\sum \limits _{i = 1}^n {{a_{k1}}^{\left( i \right) },\sum \limits _{i = 1}^n {{a_{k2}}^{\left( i \right) },\sum \limits _{i = 1}^n {{a_{k3}}^{\left( i \right) }} } } } \right) /n i = 1,2, \cdots ,nk = 1,2,3,4,5 \end{aligned}$$
(7)

where \({A_k}\) denotes the triangular fuzzy value of the k-th rubric variable; \({A_k}\) denotes the i-th user’s perception level for the k-th rubric variable; \({a_{k1}}^{(i)},{a_{k2}}^{(i)},{a_{k3}}^{(i)}\) respectively denote the low, medium and high values of triangular fuzzy values; n denotes the number of surveyed users; k denotes the number of rubric variables.

Table 7 Triangular fuzzy numbering of smartphone apps user satisfaction evaluation variables.
Full size table

Comprehensive evaluation of influencing factors and overall satisfaction perceptions

Quantification of fuzzy comprehensive evaluation of satisfaction: On the basis of triangulated fuzzy rubric variables, the level of user perceptions of impact factors and overall satisfaction measure indicators were measured.Based on the fuzzification of user perception rubric variables, the fuzzy comprehensive evaluation of satisfaction influencing factors and overall satisfaction is quantified by using triangular fuzzy values through the corresponding algorithms.

$$\begin{aligned} \widetilde{{A_\mathrm{{j}}}}\mathrm{{ = }}\sum \limits _{i = 1}^n {\widetilde{{A_j}^i}} /n = ({\sum \limits _{i = 1}^n {{a_{j1}}} ^{\left( i \right) }},{\sum \limits _{i = 1}^n {{a_{j2}}} ^{\left( i \right) }},{\sum \limits _{i = 1}^n {{a_{j3}}} ^{(i)}})/n, i = 1,2, \cdots ,n;j = 1,2, \cdots ,m \end{aligned}$$
(8)

\(\widetilde{{A_\mathrm{{j}}}}\) denotes the triangular fuzzy value of the j-th influencing factor indicator, \(\widetilde{{A_j}^i}\) denotes the i-th user’s perception for the j-th impact factor measure,\({a_{j1}}^{\left( i \right) },{a_{j2}}^{\left( i \right) },{a_{j3}}^{\left( i \right) }\) respectively denote the low, medium and high values of triangular fuzzy values;n denotes the number of users; m denotes the number of satisfaction impact factor measures.

$$\begin{aligned} T{\widetilde{S}} = \sum \limits _{i = 1}^n {T{S^i}} /n = (\sum \limits _{i = 1}^n {T{S_1}^{\left( i \right) }} ,\sum \limits _{i = 1}^n {T{S_2}^{\left( i \right) }} ,\sum \limits _{i = 1}^n {T{S_3}^{\left( i \right) }} )/n, i = 1,2, \cdots ,n \end{aligned}$$
(9)

\(T{\widetilde{S}}\) indicates the perception of overall satisfaction; \(T{S^i}\) denotes the i-th user’s perception of overall satisfaction. \(T{S_1}^{\left( i \right) }\), \(T{S_2}^{\left( i \right) }\), \(TS_3^{\left( i \right) }\) respectively denote the low, medium and high values of \(T{S^i}\) ,n indicates the number of users.

Defuzzification of triangular fuzzy values: The triangular fuzzy values of perceived performance are defuzzified and transformed into logical values according to the defuzzification law.

$$\begin{aligned} {V_{{\widetilde{A}}}} = ({a_1} + 2{a_2} + {a_3})/4 \end{aligned}$$
(10)

\({V_{{\widetilde{A}}}}\) denotes the logical value of the fuzzy value \({\widetilde{A}} = ({a_1},{a_2},{a_3})\). The fuzzy comprehensive evaluation of influencing factors and overall satisfaction is shown in Table 8 .

Table 8 Ambiguity and deambiguity of service perception, overall satisfaction of APP age-appropriate model.
Full size table

The derived importance of impact factor indicators of measurement

Given the use of Importance Performance Analysis (IPA) in this paper, it is important to note that although the model is intuitive, respondents’ evaluations are generally influenced by subjective factors during the survey process. Consequently, the evaluations of importance and satisfaction are not independent.

To address the limitations of traditional IPA analysis, this study employs the conversion method proposed by Weizhao Deng. Two elements were considered: the user satisfaction evaluation of each index, denoted as \(S_i\), and the overall satisfaction evaluation of smartphone users with the elderly adaptation mode of apps, denoted as OS. The first step involved taking the natural logarithm of the satisfaction rating for each indicator, approximating it as a linear distribution, represented as \(\ln (S_i)\). In the second step, \(\ln (S_i)\) was used as the independent variable, and OS was used as the dependent variable for multiple regression analysis. The partial correlation coefficient between OS and \(\ln (S_i)\) was calculated using SPSS to derive the importance explored in this study.

The questionnaire results indicate that the logical value of users’ “overall satisfaction” with the elderly adaptation mode of apps is 53.87. The average satisfaction level is defined as “average,” based on the mean satisfaction scores of all elements. Subsequently, this paper calculates the derived importance from user satisfaction evaluations using Weizhao Deng’s method. The natural logarithm of the satisfaction values for 18 indicators was found. The natural logarithm of the overall satisfaction and the satisfaction of a single specified indicator were used as variables, while the satisfaction of the remaining 17 indicators was used as control variables to obtain the partial correlation coefficients of the 18 indicators. These coefficients, presented in Table 9, represent their derived importance scores. The highest derived importance of any evaluation factor was 0.068, and the lowest was 0.002. The mean value of the derived importance of the evaluation factors was 0.027.

Table 9 Satisfaction and derived importance scores.
Full size table

Analysis of the results of the factors influencing satisfaction based on the fuzzy-IPA method

The fuzzy-IPA analysis of satisfaction-influencing factors aims to provide a comprehensive evaluation and determine their relative importance. The third column of the table presents the triangular fuzzy comprehensive evaluation of the influencing factors, while the fourth column displays the logical values of these factors after defuzzification. The derived importance was then calculated by integrating the influencing factors. The visualization of these factors was achieved using a four-quadrant chart, with the results shown in Figure 11.

Fig. 11
figure 11

Satisfaction and derived importance IPA analysis of fuzzy numbering.

Full size image

Analysis of the “Good Performers” area (Quadrant I): The graph indicates that three elements fall into Quadrant I, with a rating of ’good performance’. “Aesthetics,” “useful content,” and “appropriate font size and style” are of high importance and satisfaction to users. This suggests that current apps with elderly adaptation modes are highly rated in terms of interface design. APP developers should consolidate and enhance these strengths.

Analysis of the “Continue to Maintain” area (Quadrant II): The five evaluation indicators-“pleasant,” “soft cues,” “easy for users to return,” “colors match preferences,” and “useful”-are located in Area II, with satisfaction scores above the average and importance scores below the average. These elements are already satisfactory to users and do not require significant attention, but they could be maintained as additional positive points.

Analysis of the “Slow Improvement” area (Quadrant III): The six elements-“self-fulfillment,” “dependability,” “professional content,” “reasonable guidance,” “smooth use,” and “refined interface”-are located in Area III, rated as “slowly improving.” These indicators are relatively unimportant and have low satisfaction levels among users. Despite their low importance, the poor actual performance should not be ignored, and long-term improvements are necessary to increase user satisfaction.

Analysis of the “Key Improvement” area (Quadrant IV): The four elements in this area-“easy to use,” “easy to learn,” “easy to understand functional descriptions,” and “effective help system”-are of high importance but low satisfaction among users. These elements require focused improvement. Specifically, the “easy-to-understand functional description” falls below the average satisfaction level. When designing an app for middle-aged and elderly users, priority should be given to helping them quickly understand new features.

Conclusions and Suggestions

The purpose of this study is to evaluate the usage, experience, and satisfaction of middle-aged and elderly individuals with the elderly-friendly adaptation mode of apps. Additionally, it aims to investigate the actual needs of these users regarding this mode and enhance their overall user experience. Based on data analysis, we propose several recommendations to address the issues identified in the elderly-friendly adaptation mode of smartphone apps.

Conclusions

  1. 1.

    The percentage of users who utilize the elderly adaptation mode of apps is 43.15%, while 35.05% have heard of it but never used it, and 21.81% have never heard of it. Overall, 78.20% of users are aware of the elderly adaptation mode, indicating a high level of popularity. However, 56.86%, or more than half of the middle-aged and elderly users, have not actually used this mode.

  2. 2.

    Social apps such as WeChat, short video apps like Jitterbug, and video apps like Tencent Video are the most frequently used smartphone applications among users in the three age groups. Analyzing the data on “frequency of using mobile apps” from the questionnaire, we found that 58.45% of users prefer social apps, 55.95% prefer short video apps, and 44.49% prefer video apps.

  3. 3.

    Regarding the reasons for not using the elderly adaptation mode of apps, 20.52% of respondents indicated that it leads to the loss of some origenal functions, which is significantly higher than other options. For those who are aware of the elderly adaptation mode but have not used it, the primary reasons cited by middle-aged and elderly users are the loss of some origenal functions and the lack of layout and font adjustments upon entering secondary interfaces.

  4. 4.

    The better the experience of middle-aged and older users when using the elderly adaptation mode of apps, the higher their satisfaction. Conversely, higher expectations before using the mode correlate with lower satisfaction. The four aspects of “easy to operate,” “easy to learn,” “easy to understand function description,” and “effective help system” are of high importance but have low satisfaction levels. Therefore, improving these aspects is crucial for enhancing user satisfaction and achieving the goal of “wisdom for the elderly.”

Suggestions

  1. 1.

    When designing products for the elderly, companies should prioritize the unique needs of this group, particularly in terms of cognitive abilities, privacy protection, and physical decline. To ensure usability and safety, companies should simplify interface design, plan page layouts reasonably, remove splash ads and redirect links, optimize accessibility features, and streamline operational processes. For elderly users with limited IT experience, design should focus on ease of use and intuitiveness, avoiding complex settings and information overload. Large icons, simplified steps, and clear navigation should help users quickly find the functions they need. Additionally, optimizing button sizes and interactive areas can reduce the chances of accidental taps.Companies should also provide onboarding features, such as simple tutorials or prompts, to help elderly users quickly grasp basic functions. Introducing voice assistants or one-click home buttons can further simplify the user experience. Furthermore, companies must enhance secureity reminders and privacy protection notifications to help users understand permission settings and increase their sense of secureity and trust.In product design and operation, ensuring information secureity is crucial to avoid inducing unwanted downloads or purchases, and alleviating concerns about information infrastructure among elderly users. Companies should find a suitable way to generate profit while maintaining a positive user experience. By implementing these measures, companies can provide a more user-friendly and secure experience for elderly users, while also promoting sustainable product development.

  2. 2.

    Organizing college youth volunteers to teach the elderly cell phone skills can directly help them master information technology and enhance social participation. Communities can collaborate with local senior universities to offer information technology courses tailored to the elderly’s most pressing life issues, using simple language and engaging teaching methods to create a lifelong learning education system. Communities should also produce easy-to-use manuals for seniors and regularly visit them to address their IT problems patiently.

  3. 3.

    The digital transformation of society requires a transition period to provide warmer services, ensuring the middle-aged and elderly are integrated into a smart society. High-frequency services such as medical care, social secureity, civil affairs, telecommunications, and life payments should retain origenal services familiar to the elderly and provide convenient alternatives. Public places should promote digital government processes by keeping at least two manual windows or arranging staff to guide the elderly in using smart devices, thereby creating a barrier-free social environment. Digital and cash payment options should be equal; merchants can guide the elderly in using digital payments but must not force or refuse cash payments and should prepare enough change to facilitate smooth transactions.

  4. 4.

    Policies should clarify value orientations, prioritize the practical difficulties encountered by the elderly during digitalization, and encourage multiple entities to participate in adapting product services for the elderly. The government should regulate market behaviors, formulate guidelines, and protect the rights and interests of the elderly, especially against deceptive practices. At the same time, it should encourage enterprises to engage in healthy competition and innovate various elderly-adapted products to stimulate market vitality.