Investigating the Influencing Factors of the Purchase Intention of the Continuing Care Retirement Community: A Case Study of Shenzhen

Abstract

China officially entered the aging society in 2000, witnessing a rapid surge in demand for senior care services. In response, the real estate industry introduced the concept of Continuing Care Retirement Communities (CCRCs). The relationship between this model and complex risks in urban security resilience encompasses various factors, including filial care, demand for senior care, and urban sustainable development. The strategic layout and planning of CCRC enhance the security resilience of urban operational systems in the face of intricate senior care risks. However, the development and operation of CCRCs have encountered the challenge of sluggish project progress, primarily due to a lack of robust purchase intention. This study investigates the factors influencing the purchase intention of CCRCs in mainland China, using the Theory of Planned Behavior (TPB) as the foundational theoretical model. Additionally, three contextual constructs (economic cost, product performance, and external stimuli) were introduced to form the initial model. Based on the initial model, six factors were identified and nine hypotheses were proposed. A questionnaire survey was conducted to collect data, and Structural Equation Modeling (SEM) analysis was employed to test the proposed hypotheses. The results indicate that consumers’ purchase intention of CCRCs is primarily influenced by product performance and subjective norms, followed by economic cost. At the same time, external stimuli have a significant indirect effect on it.

1. Introduction

Based on the 2020 data from the seventh national population census, China’s population aged 60 and above reached 264 million, comprising 18.7% of the total population, marking a 3-percentage-point increase from 2000. The population over 65 years old amounted to 190 million, representing 13.5% of the total population, with an increase of 5.44 percentage points from 2010. Projections indicate that by the mid-21st century, China’s elderly population will escalate to 400 million, constituting approximately 26.53% of the total population. The substantial and rapid growth of China’s elderly population underscores the pressing need to address aging-related challenges [1].

The expansion of Continuing Care Retirement Communities (CCRCs) entails multifaceted risks and considerations for urban safety resilience. Firstly, China’s aging population, particularly those aged 60 and above, is on a steep upward trajectory, highlighting the critical demand for suitable senior care solutions. Secondly, as urbanization progresses, an increasing number of seniors are opting to reside in urban areas, challenging traditional familial aging arrangements. Moreover, shifts in China’s family dynamics, such as smaller household sizes and a rise in dual-income households, pose challenges for adult children in providing full care for their parents, intensifying the necessity for comprehensive care services. Consequently, the diversification of elderly living needs has spurred a growing demand for holistic care encompassing medical services, social engagement, and daily assistance. Urban areas can address the rising demand for elderly care by establishing CCRCs, thereby elevating the overall quality of urban elderly care services [2].

Furthermore, the urban welfare system is intricately linked to urban safety resilience, with its effectiveness directly influencing the quality of life and social security of the elderly. As a pivotal component of the urban welfare system, CCRCs offer assisted living and medical services for the elderly, effectively alleviating the burden on elderly individuals and their families. This enhances the affordability of CCRC services, aligning with the principles of our welfare state-type social security characterized by universal coverage, comprehensive protection, bolstered personal security, and enhanced quality of life. Simultaneously, the concept of sustainable development encompasses various aspects, including meeting present and future generations’ needs, enhancing overall quality of life, and promoting equality. These principles and objectives form the cornerstone of sustainable development. Urban welfare system development can enhance urban residents’ quality of life while ensuring prudent resource utilization and environmental preservation [3]. Accordingly, the planning and management of CCRCs must adhere to sustainable development principles, safeguarding natural resources, improving quality of life, and minimizing adverse environmental impacts. Therefore, a linear correlation exists among these elements. Integrating the urban welfare system and CCRC development with the principles of sustainable development is crucial for achieving balanced economic, social, and environmental advancement, fostering a more sustainable living environment for present and future generations [4].

However, the development of CCRCs in China is still in its nascent stage, and both residents and the construction of CCRCs need to adapt and evolve with time [5]. A crucial aspect in promoting CCRC development is to ensure that projects align with the actual needs of seniors, thereby effectively balancing supply and demand. While existing research has delved into various aspects of CCRCs, there is a notable absence of studies exploring the factors affecting the purchase intention of CCRCs [6]. This study aims to fill this gap by investigating the factors influencing the purchase intention of CCRCs in Shenzhen [7,8]. Firstly, it systematically reviews and analyzes the existing literature on CCRCs, identifying potential influencing factors and classifying them based on the Theory of Planned Behavior (TPB). Subsequently, a questionnaire survey is employed to investigate the demand for CCRCs in Shenzhen, followed by an analysis of the key factors influencing the purchase intention using Structural Equation Modeling (SEM) [9]. This study seeks to provide relevant solutions to enhance public understanding of CCRCs, fostering a conducive environment for industry development and a harmonious social atmosphere, thus promoting the healthy growth of CCRCs to facilitate positive aging [10].

The study innovations of this paper are twofold. Firstly, by conducting the study before Shenzhen officially entering the aging stage, it provides insights into the demand for CCRCs and the purchase intention of consumers before the city reaches this critical phase. Secondly, this study focuses on Shenzhen, a rapidly aging immigrant city. By exploring the perspective of “population batch aging” in this developed urban setting, the research offers a more comprehensive and accurate estimation of market demand for CCRCs. The primary focus of this paper is the potential elderly clientele of Shenzhen’s CCRCs. The survey and analysis process serves a dual purpose: firstly, it provides an avenue for disseminating and comprehensively educating residents on concepts related to CCRCs, enhancing public understanding of the subject; secondly, it facilitates the exchange of perspectives and recommendations between elderly individuals and residents, thereby enriching the discourse surrounding CCRCs. This not only broadens the understanding of CCRCs but also holds theoretical significance for future research endeavors in this domain. This paper will proceed as follows: Section 2 will summarize the current state of research on CCRCs from the international literature. Section 3 will introduce the study’s model variables, data sources, and research methodologies. Section 4 will analyze and evaluate the questionnaire data and detail the construction and analysis of the Structural Equation Modeling. Finally, Section 5 will conclude the research.

2. Literature Review

2.1. Research on CCRCs

The global development of CCRCs has reached a high level of maturity, with a comprehensive system in place. As researchers worldwide delve deeper into this field, they are increasingly focusing on the spiritual and cultural requirements of seniors and their demand for CCRCs. Consequently, much of the scholarship is centered on operational management, long-term care, and demand for such properties.

Operational management assumes a crucial role, emphasizing services tailored to seniors in the realm of senior care. Providing high-quality senior care services is central to the effective operational management of CCRCs. This theme has been the subject of extensive global research. McHugh et al. [11] distilled feedback from seniors residing in Sun City concerning the operational management of a CCRC. They based their recommendations on the site’s environment, residents’ habits, and social and economic impacts, ultimately concluding that the success of the CCRC depends on its integration with local traditional culture. Chiu et al. [12] employed a population projection model to forecast future demand for senior housing, suggesting that senior housing should be operated and managed through leasing arrangements. Byun [13] underscored the significance of safety in the planning of senior housing in South Korea, emphasizing the imperative to augment safety considerations in the planning of residential areas catering to seniors. Yang and Sui [14] examined how residential environments impact the emotional well-being of seniors, conducting an analysis that encompassed both built and social environments. Through a combination of self-assessed health, living conditions, and mobility, they explored the correlation between residential environment factors and emotional health, revealing significant differences in the impact of built and social environmental factors on emotional well-being.

The next topic of consideration is long-term care for seniors. Scholars in various countries have been utilizing the concept of “long-term care” in their study of senior care models. Long-term care encompasses a series of nursing services provided by healthcare professionals to ensure that semi-disabled, disabled, and dementia-afflicted senior individuals can lead fulfilling lives and achieve physical and psychological well-being while keeping their values. Portrait et al. [15] conducted a study on long-term care through the creation of mathematical models and refined the concept of a “multidimensional health status” for assessing the long-term care approach. This model can be applied to analyze the factors influencing senior individuals’ choices of long-term care as well as to assess the scope and targets of informal care, family care, and institutional care, in addition to accurately predicting the healthcare requirements of seniors. SunWoo [16] examined the governmental perspective on addressing the challenges in developing the long-term care model and asserted that national policy adjustments in support facilities, institutionalizations, and care are crucial for senior individuals’ values. Furthermore, it is imperative to prioritize the enhancement of cultural, recreational, and social activities for seniors. Simultaneously, offering medical and nursing care support to address both physiological and psychological needs is crucial for ensuring a secure and tranquil old age. Oesterle [17] conducted a research project on the current status of long-term care development in seven countries in Central and Southeastern Europe and found that family care remains the mainstay of long-term care development. Williams et al. [18] conducted an empirical study on the healthcare integration model using data from the Social Survey Cycle Report released by Statistics Canada. Their study classified the long-term care model into home care, institutional care, and hospice care (end-of-life care), providing a strategic framework for the future development of long-term care. Kim and Lee [19] surveyed the current status of nursing homes in South Korea implementing a long-term care insurance policy. The study revealed the effective management of human resources and resident care in these nursing homes and identified deficiencies in community resource utilization and financial and environmental facilities management. Addressing these deficiencies is crucial for continuous improvement in nursing homes within CCRCs to enhance the quality of care services, optimize manpower allocation, secure funding, and improve support management systems.

The third aspect pertains to demand intention. Cohen [20] argued that the diverse physiological and psychological conditions of seniors impact their considerations related to lifestyle, health status, and the selection of CCRC locations. These scholarly findings have implications for the design of long-term care services for seniors. Westerholm [21] conducted a study on the demand for CCRCs among seniors with dementia or multiple diseases, noting that such seniors preferred permanent residence options and sought round-the-clock medical care services. Japan, recognized as the most rapidly aging nation, underwent a transition from an aging to an aged society over a span of 24 years. The dire situation of the aging society compelled Japanese authorities to increase the construction of the senior service system. With the enhancement of the service system, there has been an increasing focus on the demand for senior-friendly environments. Trahutami [22] focused on examining how the environment influences the extended life expectancy of seniors. Through a comprehensive analysis of data related to senior care and health, it was concluded that natural factors such as water quality, air quality, natural food, proximity to nature, and a clean and comfortable environment influence the high life expectancy of senior individuals in Japan. Aung et al. [23] evaluated the contribution of 20 factors to the dependent variable in the theoretical framework of the environmentally friendly environment recommended by the WHO. Through a study of 243 senior people in Japan, two factors were found to be statistically significant in relation to positive aging: (1) participation in social and cultural activities and (2) participation in group sports activities during leisure time. Additionally, quality of life was strongly associated with (1) locally available information about health problems and service needs, (2) personal care or assistance needs that could be met in a home setting with the use of formal services, and (3) income sufficient to meet basic needs in the past 12 months without public or private assistance.

2.2. Theory of Planned Behavior

To address the shortcomings of rational behavior theory in elucidating individual behavior, Ajzen proposed the Theory of Planned Behavior (TPB). When external factors impinge upon human behavior, or when there exists a willingness to act beyond an individual’s control or ability to execute fully according to their intentions, considerations beyond individual volition become necessary. TPB posits that apart from attitudes and subjective norms, the inclination to act is also shaped by perceived behavioral control. Human behavioral processes are driven by personal intentions while being regulated by perceived control over behaviors. Moreover, subjective norms constrain behavioral control, and these norms are influenced by personal attitudes, implying that positive attitudes are imperative for fostering positive behavioral processes. Numerous scholars have conducted extensive research on behavioral inclination, spanning various domains such as the online purchase of fresh agricultural products, life insurance purchase, engagement in public affairs, elderly medical care involvement, rural entrepreneurship, public housing withdrawal, clean energy consumption, low-carbon travel, and green food purchasing. Furthermore, the willingness to renovate old neighborhoods underscores the adaptability and explanatory potency of TPB, highlighting its applicability in elucidating and predicting consumer behavior. In this study, we employ TPB to construct a model elucidating the relationship between the purchase intention of CCRCs and its influencing factors, thus exploring the dynamics between these factors and the inclination to invest in CCRCs.

2.3. Research Gap

In existing studies, it has been observed that the development process of CCRCs is primarily influenced by three main factors: government policies, business enterprises, and consumers’ preferences. Firstly, the issue of aging needs to be guided and driven at the government level. A robust social security system can offer policy support to CCRC enterprises, mobilize the enthusiasm of relevant enterprises, and create a conducive business environment for the development of CCRCs. Subsequently, building upon favorable policies, enterprises need to actively engage social groups in the development of CCRCs, focusing on transformation and innovation in financing models, management modes, and other aspects. Lastly, consumers’ purchase intention of CCRCs is crucial. As a relatively new product, CCRCs have not yet gained widespread popularity, and consumers’ recognition and acceptance need to be improved. Furthermore, the diverse services offered by CCRCs lead to a higher price than ordinary housing, with the range and quality of services becoming important considerations for consumers. Hence, factors such as price and service ultimately influence consumers’ purchase intention of CCRCs, serving as a key control point for the sustainable development of the entire CCRC industry. Based on this, this paper selects Shenzhen city as the research site to analyze the factors affecting the purchase intention of CCRCs. The objective is to address important issues in the development of CCRCs, effectively promote their healthy development, and truly realize the strategic goal of active aging.

3. Research Methodology

3.1. Research Methods

The research methods utilized in this study encompass the following:

(1)

Literature Review Method: This method plays a pivotal role in identifying research gaps, defining objectives, and proposing methodologies. Throughout the drafting and writing of the research proposal, the relevant literature from both domestic and international databases underwent a thorough review. Through summarization, categorization, and analysis, this study delineated its research direction and approach.

(2)

Questionnaire Survey Method: This approach involved the creation of online or offline questionnaires to explore the factors influencing the inclination to invest in CCRCs. The reliability and validity of the questionnaire data were assessed using SPSS 26.0 software, ensuring the authenticity and effectiveness of the data and materials essential for the paper.

(3)

Field Interview Method: Field interviews were conducted with pertinent personnel from elderly associations across various districts in Shenzhen. This methodology aimed to gain firsthand insights into the challenges encountered during the aging process in Shenzhen, thereby acquiring the most pertinent data for forecasting the demand for CCRCs and devising policies and measures for their sustainable development.

(4)

SEM Analysis Method: SEM is adept at analyzing causal relationships among multiple influencing factors and outcomes, especially in scenarios with numerous influencing factors and diverse causal relationships. Given the intricate nature of the factors influencing CCRCs and the array of outcomes they yield, SEM was employed in this study to conduct statistical analyses of multiple factors and investigate the primary determinants of the purchase intention of CCRCs. The specific execution process of SEM is illustrated in Figure 1.

3.2. Influencing Factors

This study adopts the TPB to identify behavioral attitudes, subjective norms, and perceived behavioral control as crucial factors of the purchase intention of CCRCs. Building on this framework, the theoretical model incorporated three additional variables—economic cost, product performance, and external stimuli—to comprehensively analyze the relationship between purchase intention and its influencing factors [24].

In this paper, a theoretical model of TPB is proposed, highlighting three newly introduced variables that are influential in the development of CCRCs. Firstly, price directly affects consumers’ purchase intention, making it crucial to incorporate it into the CCRC purchase model. Furthermore, as a nascent and policy-oriented industry in China, CCRCs are influenced by societal perceptions, which impact consumers’ purchase intention. This aligns with the notion that external stimuli shape consumers’ purchasing behavior [25]. Moreover, CCRCs comprise housing and services specifically designed for seniors, requiring consumers to gain a comprehensive understanding before making a purchase, thereby enhancing their acceptance of the facilities. Therefore, the performance of CCRC products serves as a solid foundation for industry growth and acts as an appealing factor for consumers.

Through a systematic exploration, 30 observational variables related to the purchase intention of CCRCs were scientifically categorized and organized. The specific details and codes for these variables are presented in

Table 1. Categorization of measured and latent variables.

Latent Variables	Observed Variables	Coding
Behavioral attitudes	Investing in CCRCs is the right choice.	ATB1
	Investing in CCRCs is a healthy choice.	ATB2
	Investing in CCRCs is a sensible choice.	ATB3
	Investing in a retirement property is a valuable choice.	ATB4
Subjective norms	Families support investment in CCRCs.	SN1
	Friends support investment in CCRC.	SN2
	The government supports investment in CCRCs.	SN3
Perceived behavioral control	I understand the characteristics and information of CCRCs.	PBC1
	I would be more inclined to purchase a retirement property if there were comprehensive information and evaluation criteria available for such properties.	PBC2
	I prefer to invest in CCRCs if they’re available for my needs.	PBC3
	I can afford to buy a retirement property now.	PBC4
Economic cost	I prefer to buy a retirement property if the price is right.	EC1
	I am more willing to buy if my family has good financial strength.	EC2
	I am more willing to buy if the average house price is suitable.	EC3
	I am more willing to buy if the price is lower than the average price of goods.	EC4
External stimuli	I am more willing to buy if the government gives financial subsidies.	EST1
	I am more willing to buy if the market popularity is high.	EST2
	Promotional campaigns by enterprises will boost my willingness to buy.	EST3
	The high visibility of a company will promote my willingness to buy.	EST4
	The high-quality programs of a company will promote my willingness to buy.	EST5
	The reputation of a company will promote my willingness to buy.	EST6
Product performance	CCRCs can improve health.	PTP1
	CCRCs can improve convenience.	PTP2
	CCRCs can reduce the pressure on society to provide for seniors.	PTP3
	CCRCs can meet medical care needs.	PTP4
	CCRCs can meet spiritual and cultural needs.	PTP5
Purchase intention	I will take the initiative to pay attention to information related to CCRCs.	ESI1
	I would prioritize senior housing projects.	ESI2
	I would like to live in a retirement property.	ESI3
	I would recommend CCRCs to my friends and family.	ESI4

.

(1)

Behavioral attitudes

Behavioral attitudes refer to consumers’ favorable or unfavorable disposition towards purchasing CCRCs, reflecting their subjective stance on the matter. CCRCs offer both residential and pension services tailored to the elderly, promising a wholesome and enjoyable experience. Awareness of the benefits associated with senior housing tends to elevate consumers’ purchase intention. Accordingly, it was hypothesized that positive behavioral attitudes significantly enhance the intention to purchase CCRCs (H1).

(2)

Subjective norms

Subjective norms encompass the pressure and influence emanating from the attitudes and behaviors of family members, friends, influential individuals, and society at large when consumers contemplate the acquisition of senior housing. The precedent of real estate purchases by family, friends, or significant others heightens the pressure and influence on consumers, thereby augmenting their purchase intention [26]. Furthermore, subjective norms directly affect both behavioral attitudes and purchase intention [27]. Hence, it was hypothesized that positive subjective norms significantly correlate with favorable behavioral attitudes towards CCRCs (H2) and increase the purchase intention of CCRCs (H3).

(3)

Perceived behavioral control

Perceived behavioral control denotes consumers’ individual cognitive assessment when making purchase decisions, encompassing the perceived ease or difficulty of the purchasing process. Generally, if consumers feel sufficiently informed about senior housing options, find suitable projects in the market, and possess the means to make a purchase, their purchase intention strengthens. Consequently, it was hypothesized that robust perceived behavioral control among consumers significantly enhances their purchase intention of CCRCs (H4) [28,29].

(4)

Economic cost

Economic cost encompasses not only the price of the product itself but also the financial burden borne by consumers to initiate consumption behavior, directly influencing their purchase intention. If the CCRC price aligns with a family’s purchasing power and the subsequent cost of living falls within an acceptable range, consumers’ purchase intention escalates. Consequently, it was hypothesized that substantial economic costs exert a significant positive impact on consumers’ purchase intention of CCRCs (H5).

(5)

External stimuli

Drawing on insights from the four theories of purchasing behavior, this paper incorporated external stimuli as a variable in the model examining the purchase intention of CCRCs. Additionally, consumers’ purchase intentions are influenced by external factors such as prevailing policies and the operational strategies of CCRC ventures. Specifically, heightened purchase intentions are observed when governmental policies offer robust protection, companies enjoy a favorable reputation, and market demand is substantial [30]. Moreover, research indicates that external stimuli can directly shape consumers’ behavioral attitudes [25]. Consequently, it was hypothesized that external stimuli exert a significant positive impact on both consumers’ behavioral attitudes towards CCRCs (H6) and their purchase intention in this domain (H7).

(6)

Product performance

CCRCs remain in their nascent stages, and their product performance serves as a pivotal tool to penetrate the market, establish foundations, and foster long-term development. Furthermore, they serve as a sustained driver to enhance consumers’ potential purchase intentions. Product performance significantly influences consumers during the information processing phase, thereby bolstering their intentions to make purchases. Additionally, product performance positively impacts behavioral attitudes and purchase intention [31]. Hence, the following hypotheses were posited: High product performance significantly enhances consumers’ behavioral attitudes towards CCRCs (H8) and high product performance significantly boosts consumers’ purchase intention of CCRCs (H9).

Based on the initial theoretical model, nine hypotheses were proposed, as shown in

Table 2. Model assumptions for factors influencing demand for CCRCs.

Hypothesis	Hypothetical Content
H1	Positive behavioral attitudes significantly enhance the intention to purchase CCRCs.
H2	Positive subjective norms significantly correlate with favorable behavioral attitudes towards CCRCs.
H3	Positive subjective norms significantly increase the purchase intention of CCRCs.
H4	Robust perceived behavioral control among consumers significantly enhances their purchase intention of CCRCs.
H5	Substantial economic costs exert a significant positive impact on consumers’ purchase intention of CCRCs.
H6	External stimuli exert a significant positive impact on consumers’ behavioral attitudes towards CCRCs.
H7	External stimuli exert a significant positive impact on consumers’ purchase intention.
H8	High product performance significantly enhances consumers’ behavioral attitudes towards CCRCs.
H9	High product performance significantly boosts consumers’ purchase intention of CCRCs.

. The relationship between the hypotheses and the factors is shown in Figure 2.

3.3. Data Collection

Based on the conceptual model of the TPB, this paper developed a questionnaire to assess the factors influencing the inclination to purchase CCRCs. Three variables—economic cost, product performance, and external stimuli—were introduced, drawing on both the established scale design method of TPB and insights from the related literature. The questionnaire comprised three sections: basic information, current senior living situation, and factors influencing purchase intention. Questions covered respondents’ demographic characteristics and variables observed by the SEM. The questionnaire included dimensions such as behavioral attitude (4 items), subjective norms (3 items), perceived behavioral control (4 items), economic cost (4 items), external stimulus (6 items), product performance (5 items), and willingness to buy (4 items). Items were scored on a 5-point Likert scale, with 1 denoting “strongly disagree” and 5 denoting “strongly agree”. Following the determination of the required sample size based on previous research experience and the current study’s context, a combination of field and online surveys was conducted in Shenzhen. The target respondents were individuals aged 50 and above, encompassing both current purchasers and potential consumers of CCRCs, to comprehensively assess the willingness to purchase such properties of the elderly population of Shenzhen. A total of 110 questionnaires were collected, of which 104 were deemed valid after screening and excluding invalid responses, yielding a questionnaire validity rate of 94.5%, meeting the sample capacity requirements.

3.4. Data Analysis Process

To investigate the factors influencing the purchase intention of CCRCs, both online and offline questionnaires were utilized. SPSS 26.0 software was employed to assess the normality, reliability, and validity of the questionnaire data, ensuring the accuracy and effectiveness of the information used in this study.

For data analysis, SEM was conducted using SPSS 26.0 and AMOS 24.0 software. The SEM process consisted of two main components: measurement modeling and structural modeling. In the measurement modeling stage, latent variables were measured using observed variables, while the relationships between latent variables were examined in the structural modeling phase.

Before conducting the structural model testing, Confirmatory Factor Analysis (CFA) was necessary to verify the validity of the measurement model. This process can involve the removal of certain observed variables. Following CFA, the goodness of fit of the structural model was assessed and adjusted accordingly. Once an optimized model was obtained, the results were analyzed and interpreted to identify significant influences and regression weights.

Furthermore, field interviews were carried out with pertinent personnel from the Shenzhen Senior Citizens Association in each city district. The purpose of these interviews was to gain insights into the challenges faced by Shenzhen in addressing the aging process. This valuable information contributed to predicting the demand for CCRCs and proposing policies and measures for their healthy development.

4. Results and Discussion

4.1. Questionnaire Data Test

4.1.1. Descriptive Statistical Analysis

SPSS 26.0 software was employed to analyze the frequency of variables in the questionnaire data, and

Table 3. Frequency analysis of demographic variables.

Question	Item Option	Frequency	Percentage	Mean	Standard Deviation
Genders	Male	33	31.70	1.68	0.47
	Female	71	68.30
Age	50–60 years old	14	13.50	2.09	0.59
	61–70 years old	67	64.4
	71–80 years old	23	22.10
Duration of residence	1 year and below	1	1.00	3.36	1.17
	2–3 years	30	28.80
	4–5 years	29	27.90
	6–10 years	19	18.30
	11 years and above	25	24.00
Household registration	Shenzhen households	70	67.30	1.33	0.47
	Non-Shenzhen households	34	32.70
Health status	Very good	17	16.30	2.41	0.87
	Good	36	34.60
	Average	42	40.40
	Poor	9	8.70
Number of children	None	3	2.90	2.77	0.70
	1	31	29.80
	2	57	54.80
	3 and above	13	12.50
Whether living with children	No children	10	9.60	2.46	0.76
	Living in the same city	43	41.30
	Not living in the same city	44	42.30
	Different cities, different residence	7	6.70
Occupation before retirement	State-owned	7	6.70	5.01	3.12
	Private enterprise	14	13.50
	Civil servants	23	22.10
	Institution	15	14.40
	Education	15	14.40
	Medical	4	3.80
	Self-employed	1	1.00
	Farming	8	7.70
	Freelance	11	10.60
	Unemployed	4	3.80
	Others	2	1.90

presents some of the extracted analysis outcomes. The results of the frequency analysis for each variable demonstrate a substantial adherence to sample survey requirements. Table 3 illustrates a prevailing prevalence of individuals in senior demographics, specifically those aged between 61 and 70 years, manifesting a pronounced inclination towards senior housing arrangements and exhibiting a discernible level of financial reserves. Notably, these respondents exhibited a robust intention to purchase CCRCs. Furthermore, it is evident from the data that most senior individuals enjoy good health. Accordingly, drawing upon Maslow’s Hierarchy of Needs, one may infer a relatively heightened pursuit of spiritual and cultural aspirations within this demographic. Consequently, this analysis of respondent characteristics underscores a vast potential market for CCRCs.

4.1.2. Normality Test

The assumption of normal distribution underpins most statistical methods. In this study, descriptive statistics, such as mean, standard deviation, skewness, and kurtosis, were computed for each dimension using the scale as the measurement tool. An absolute skewness value exceeding 3 and an absolute kurtosis value surpassing 0 for observed sample variables could indicate a departure from normal distribution.

Employing SPSS 26.0 software, the normality of the survey data was assessed, with the results presented in

Table 4. Normality test results.

Latent Variables	Coding	Minimum	Maximum	Mean	Standard Deviation	Skewness	Kurtosis
Subjective norms	SN1	1	5	2.919	0.921	−0.022	−0.428
	SN2	1	5	3.152	0.868	0.045	−0.318
	SN3	1	5	3.203	0.851	−0.036	−0.199
Perceived behavioral control	PBC1	1	5	3.034	0.924	0.089	−0.371
	PBC2	1	5	3.294	0.938	0.003	−0.492
	PBC3	1	5	3.081	0.921	−0.057	−0.227
	PBC4	1	5	3.264	0.994	−0.069	−0.457
Economic cost	EC1	1	5	3.118	0.993	−0.030	−0.431
	EC2	1	5	3.091	0.925	0.024	−0.424
	EC3	1	5	2.889	0.938	0.100	−0.334
	EC4	1	5	3.037	0.899	0.067	−0.329
External stimuli	EST1	1	5	3.405	0.948	−0.122	−0.680
	EST2	1	5	3.030	0.884	−0.089	−0.023
	EST3	1	5	3.280	0.802	0.045	0.030
	EST4	1	5	3.260	1.000	−0.008	−0.607
	EST5	1	5	3.368	0.865	−0.062	−0.308
	EST6	1	5	3.051	0.9710	−0.124	−0.583
Product performance	PTP1	1	5	2.983	0.937	0.133	−0.351
	PTP2	1	5	2.821	0.949	0.077	−0.479
	PTP3	1	5	3.142	0.823	0.098	0.074
	PTP4	1	5	3.196	0.877	0.094	−0.201
	PTP5	1	5	3.503	0.932	−0.023	−0.635
Behavioral attitudes	ATB1	1	5	3.152	0.946	−0.187	−0.293
	ATB2	1	5	2.919	0.924	0.240	−0.254
	ATB3	1	5	3.095	0.846	0.055	−0.315
	ATB4	1	5	3.574	0.782	−0.077	−0.164
Purchase intention	ESI1	1	5	3.182	0.856	0.032	−0.336
	ESI2	1	5	2.895	0.938	0.136	−0.641
	ESI3	1	5	3.179	0.923	0.079	−0.540
	ESI4	1	5	2.953	0.977	−0.037	−0.428

. The findings reveal that the absolute skewness values for all question items were less than 3 and the absolute kurtosis values were less than 10. These results indicate general adherence to the normal distribution, establishing suitability for subsequent analyses. Simultaneously, the standard deviation for each question item approximated or equaled 1, signifying considerable variability among the seven potential variables.

4.1.3. Reliability Test

The scale-type measurement tool underwent an analysis of internal consistency reliability, with Cronbach α serving as the reliability index. A measurement dimension achieving an α value exceeding 0.7 signified satisfactory internal consistency [32].

To evaluate reliability, we employed SPSS 26.0 software, and the outcomes of the reliability analysis for each dimension are presented in

Table 5. Internal consistency Cronbach α values.

Question Item	Scaled Mean After Deletion of Items	Scaled Variance After Deletion of Items	Corrected Item–Total Correlation	Squared Multiple Correlation	Cronbach α After Deletion of Items	Cronbach α
SN1	6.071	2.425	0.690	0.477	0.680	0.799
SN2	6.355	2.325	0.644	0.429	0.728
SN3	6.122	2.555	0.602	0.369	0.769
PBC1	9.591	5.917	0.658	0.443	0.810	0.844
PBC2	9.409	5.409	0.717	0.515	0.784
PBC3	9.639	5.764	0.698	0.487	0.793
PBC4	9.378	5.904	0.643	0.420	0.817
EC1	9.098	5.851	0.668	0.469	0.802	0.841
EC2	9.044	5.568	0.720	0.528	0.780
EC3	9.017	5.474	0.669	0.448	0.803
EC4	9.247	5.773	0.647	0.430	0.811
EST1	16.135	13.087	0.773	0.639	0.866	0.894
EST2	16.345	13.210	0.783	0.692	0.865
EST3	16.365	14.531	0.647	0.439	0.886
EST4	16.027	14.528	0.667	0.469	0.883
EST5	15.990	13.535	0.751	0.586	0.870
EST6	16.115	14.834	0.680	0.491	0.881
PTP1	12.662	8.326	0.730	0.539	0.819	0.861
PTP2	12.449	8.743	0.702	0.504	0.826
PTP3	12.503	9.315	0.628	0.402	0.845
PTP4	12.142	8.563	0.683	0.478	0.831
PTP5	12.824	8.600	0.656	0.435	0.838
ATB1	9.257	5.086	0.692	0.481	0.773	0.832
ATB2	9.030	5.284	0.695	0.484	0.771
ATB3	9.027	5.748	0.635	0.416	0.799
ATB4	9.314	5.477	0.622	0.395	0.805
ESI1	9.821	4.758	0.662	0.447	0.799	0.838
ESI2	9.645	4.853	0.727	0.542	0.770
ESI3	9.588	4.623	0.679	0.477	0.792
ESI4	9.166	5.406	0.622	0.427	0.816

. The Cronbach α coefficients displayed in the table range from 0.799 to 0.90, indicating a high level of internal consistency. Notably, upon removal of any question item, the α value did not witness a significant increase compared to the original α value. This signified that the retention of all question items was advised, without the need for deletion.

4.1.4. Factor Analysis and Convergent Validity

To assess the adequacy of questionnaire items, a factor analysis was performed on the dimensional items. This analysis involved two stages: exploratory factor analysis and validation factor analysis. In the present subsection, exploratory factor analysis was conducted initially for each dimension. This required meeting specific criteria, including a KMO value greater than 0.7, a significant Bartlett’s test of sphericity, and cumulative variance explained exceeding 40% [33]. Subsequently, validated factor analysis was employed to test the significance of factor loadings. The Combined Reliability (CR) and Average Variance Extracted (AVE) were computed for each dimension using the standardized factor loadings. Fornell [34] considered the measurement dimension to have good convergent validity when CR > 0.7 and AVE > 0.5.

The analysis was conducted using SPSS 26.0 software, and the results are presented as follows:

(1)

Exploratory factor analysis

Validity test results obtained through exploratory factor analysis are shown in

Table 6. KMO and Bartlett sphericity test results.

Dimension	KMO	χ2	Bartlett Sphericity Test df	p	Percentage of Variance Explained (%)
Subjective norms	0.700	281.061	3	0.000	71.496
Perceived behavioral control	0.815	464.568	6	0.000	68.089
Economic cost	0.804	464.439	6	0.000	67.891
External stimulation	0.871	976.925	15	0.000	65.477
Product performance	0.866	621.148	10	0.000	64.396
Behavioral attitudes	0.808	430.886	6	0.000	66.541
Purchase intention	0.792	466.985	6	0.000	67.562

. The findings reveal that the KMO value for each dimension exceeded 0.7, Bartlett’s test of sphericity reached a significant level (p < 0.001), and the percentage of explained variance exceeded 60%. Hence, the preliminary validity test suggested that each dimension demonstrated good validity.

(2)

Validation factor analysis

Subsequently, the convergent validity test was conducted through validated factor analysis, and the findings are documented in

Table 7. Gathering validity test results.

Dimension	Item	Standardized Factor Loadings	Z-Value	p-Value	CR	AVE
Subjective norms	SN1	0.748	13.348	***	0.803	0.577
	SN2	0.694	12.559	***
	SN3	0.831	15.024	***
Perceived behavioral control	PBC1	0.778	14.754	***	0.844	0.576
	PBC2	0.713	13.132	***
	PBC3	0.734	13.641	***
	PBC4	0.807	15.480	***
Economic cost	EC1	0.734	13.682	***	0.843	0.573
	EC2	0.821	15.908	***
	EC3	0.718	13.271	***
	EC4	0.751	14.094	***
External stimuli	EST1	0.806	16.245	***	0.894	0.587
	EST2	0.667	12.487	***
	EST3	0.720	13.817	***
	EST4	0.836	17.155	***
	EST5	0.687	12.981	***
	EST6	0.859	17.916	***
Product performance	PTP1	0.804	15.810	***	0.862	0.557
	PTP2	0.772	14.909	***
	PTP3	0.715	13.404	***
	PTP4	0.682	12.584	***
	PTP5	0.751	14.354	***
Behavioral attitudes	ATB1	0.726	13.366	***	0.833	0.556
	ATB2	0.697	12.668	***
	ATB3	0.787	14.892	***
	ATB4	0.769	14.436	***
Purchase intention	ESI1	0.763	14.576	***	0.841	0.570
	ESI2	0.733	13.798	***
	ESI3	0.811	15.899	***
	ESI4	0.710	13.229	***

Note: *** p < 0.001, two-tailed.

. The analysis revealed that the question items within each dimension exhibited factor loadings ranging from 0.667 to 0.859, all statistically significant at the p < 0.001 level. Furthermore, the dimensions demonstrated satisfactory conformity with the prescribed criteria for CR and AVE. These results affirmed the favorable convergent validity of the measurement dimensions.

4.1.5. Correlation Analysis and Discriminant Validity

The structural validity of the data encompassed both convergent validity and discriminant validity. To assess discriminant validity, the square root of the AVE for each dimension was compared with the correlation coefficients between that dimension and other dimensions. The scores of each dimension question item were averaged, and a correlation analysis was conducted to examine the relationship between the dimensions. The AVE results were presented along the diagonal, and the obtained findings are presented in

Table 8. Correlation analysis and distinguishing validity results.

Dimension	M	SD	Correlation
			Subjective Norms	Perceived Behavioral Control	Economic Cost	External Stimuli	Product Performance	Behavioral Attitudes	Purchase Intention
Subjective norms	3.09	0.74	0.76
Perceived behavioral control	3.17	0.78	0.10	0.76
Economic cost	3.03	0.77	0.05	0.04	0.76
External stimuli	3.23	0.74	0.04	0.07	0.03	0.77
Product performance	3.13	0.73	0.04	0.05	0.08	0.01	0.75
Behavioral attitudes	3.05	0.75	0.17 **	0.05	0.05	0.18 **	0.24 ***	0.75
Purchase intention	3.18	0.72	0.31 ***	0.20 ***	0.35 ***	0.09	0.43 ***	0.34 ***	0.76

Note: ** p < 0.01, *** p <0.001, two-tailed. Bolded numbers on the diagonal of the table are the square root of the mean variance extracted for the corresponding dimension (√AVE); off-diagonal numbers are inter-dimensional correlation coefficients.

.

The results in Table 8 demonstrate a significant positive correlation between subjective norms, external stimuli, product performance, and behavioral attitudes (p < 0.05) as well as between subjective norms, perceived behavioral control, economic cost, product performance, and purchase intention (p < 0.05). This provides a preliminary validation of the hypotheses presented in this study. Furthermore, the correlation coefficients between any two dimensions were smaller than the square root of the average variance extracted (diagonal bolded numbers), indicating adequate discriminant validity between the dimensions and effective distinctions of the concepts.

4.2. Structural Equation Modeling

4.2.1. Initial Structural Equation Model Construction

This study employed AMOS 24.0 software to conduct SEM of the factors influencing the purchase intention of CCRCs [35]. The model included five independent variables: subjective norms, perceived behavioral control, economic cost, external stimuli, and product performance, as well as two dependent variables: behavioral attitudes and purchase intention. The structural equation model constructed is presented in Figure 3.

4.2.2. Initial Tests and Modifications to the Model

• Basic fitness test

Table 9. Basic fit test.

Observed Variable Errors	Estimate	S.E.	C.R.	p
e1	0.223	0.039	5.667	***
e2	0.372	0.045	8.237	***
e3	0.389	0.042	9.353	***
e4	0.390	0.041	9.514	***
e5	0.344	0.044	7.859	***
e6	0.335	0.039	8.566	***
e7	0.431	0.044	9.828	***
e8	0.351	0.038	9.278	***
e9	0.278	0.037	7.505	***
e10	0.453	0.048	9.400	***
e11	0.425	0.044	9.763	***
e12	0.301	0.033	9.128	***
e13	0.246	0.029	8.394	***
e14	0.432	0.039	11.027	***
e15	0.394	0.036	10.875	***
e16	0.309	0.029	10.733	***
e17	0.314	0.032	9.746	***
e18	0.309	0.035	8.798	***
e19	0.309	0.033	9.446	***
e20	0.362	0.034	10.536	***
e21	0.377	0.038	9.782	***
e22	0.439	0.043	10.201	***
e23	0.388	0.045	8.665	***
e24	0.323	0.039	8.218	***
e25	0.346	0.037	9.462	***
e26	0.451	0.046	9.917	***
e27	0.394	0.04	9.758	***
e28	0.244	0.03	8.220	***
e29	0.373	0.04	9.301	***
e30	0.302	0.03	10.057	***

Note: *** indicates a level of significance p < 0.001.

reveals that all error variances (e1 to e29) of the observed variables were positive and ranged between 5.667 and 11.027. Additionally, all error variances demonstrated statistical significance at 0.001 or higher.

The Critical Ratio (C.R.) value represents the critical ratio of the test, while p indicates the significance coefficient. A CR value exceeding 1.96 with a p value of less than 0.05 indicates significance at the 95% confidence level, and a CR value surpassing 2.58 with a p value lower than 0.01 signifies significance at the 99% confidence level.

Table 10. Parameter estimation results of the initial model.

Pathway			Unstandardized Regression Coefficient	Standardized Regression Coefficient	S.E.	C.R.	p
Behavioral attitudes	←	Subjective norms	0.137	0.183	0.051	2.702	0.007
Behavioral attitudes	←	External stimuli	0.136	0.181	0.048	2.807	0.005
Behavioral attitudes	←	Product performance	0.207	0.276	0.050	4.163	***
Purchase intention	←	Behavioral attitudes	0.182	0.202	0.056	3.262	0.001
Purchase intention	←	Product performance	0.267	0.395	0.041	6.524	***
Purchase intention	←	External stimuli	0.026	0.039	0.036	0.728	0.466
Purchase intention	←	Economic cost	0.231	0.342	0.039	5.878	***
Purchase intention	←	Perceived behavioral control	0.100	0.147	0.037	2.678	0.007
Purchase intention	←	Subjective norms	0.180	0.267	0.040	4.516	***
SN1	←	Subjective norms	0.706	0.831	0.047	15.021	***
SN2			0.688	0.748	0.051	13.374	***
SN3			0.601	0.694	0.049	12.224	***
PBC1	←	Perceived behavioral control	0.675	0.734	0.049	13.634	***
PBC2			0.800	0.807	0.052	15.445	***
PBC3			0.718	0.778	0.049	14.709	***
PBC4			0.668	0.713	0.051	13.129	***
EC1	←	Economic cost	0.675	0.751	0.048	14.102	***
EC2			0.758	0.821	0.048	15.863	***
EC3			0.728	0.734	0.054	13.585	***
EC4			0.671	0.717	0.051	13.222	***
EST1	←	External stimuli	0.834	0.836	0.049	17.141	***
EST2			0.833	0.859	0.047	17.872	***
EST3			0.589	0.667	0.047	12.427	***
EST4			0.594	0.687	0.046	12.907	***
EST5			0.576	0.720	0.042	13.791	***
EST6			0.763	0.806	0.047	16.232	***
PTP1	←	Product performance	0.752	0.804	0.048	15.810	***
PTP2			0.675	0.772	0.045	14.903	***
PTP3			0.560	0.682	0.045	12.560	***
PTP4			0.699	0.751	0.049	14.353	***
PTP5			0.677	0.715	0.051	13.380	***
ATB1	←	Behavioral attitudes	1.000	0.769			***
ATB2			0.968	0.787	0.078	12.426	***
ATB3			0.827	0.726	0.070	11.822	***
ATB4			0.871	0.697	0.077	11.275	***
ESI1	←	Purchase intention	1.000	0.732			***
ESI2			1.014	0.811	0.081	12.547	***
ESI3			1.065	0.762	0.086	12.416	***
ESI4			0.820	0.710	0.074	11.126	***

Note: *** indicates a level of significance p < 0.001.

illustrates that, apart from the path coefficient between “external stimuli” and “purchase intention”, which was insignificant, all other paths exhibited significance.

2.

Overall fit test

When conducting an overall analysis of the structural equation model, the initial consideration should prioritize the assessment of the fit index. The chosen fit index should include measures for absolute fit evaluation, incremental fit evaluation, and parsimonious fit evaluation.

Table 11. Initial model fitting results.

Category	Indicator	Judgment Criteria	Modified Model	Fitness Judgment
Absolute fit evaluation	χ2	The smaller, the better	440.736	Passed
	p	p > 0.05	0.028	Failed
	χ2/df	<3	1.142	Passed
	GFI	>0.9	0.914	Passed
	AGFI	>0.9	0.896	Failed
	RMSEA	<0.08	0.022	Passed
Incremental fit evaluation	NFI	>0.9	0.897	Failed
	IFI	>0.9	0.986	Passed
	CFI	>0.9	0.986	Passed
	RFI	>0.9	0.884	Failed
Simple fit evaluation	PGFI	>0.5	0.759	Passed
	PNFI	>0.5	0.796	Passed

presents the specific index used and the fitting results of the initial model derived from applying AMOS 24.0 software for calculation.

3.

Model correction and re-testing

Table 11 reveals a poor initial model fit, prompting the need for corrections. As per the Modification Indices (MIs) suggestions offered by AMOS, the initial model was rectified by introducing sequential adjustments to the residual correlations between EST2 and EST4, EST2 and EST3, EST2 and PTP2, and ESI2 and ESI4.

In the first correction, it became evident that the MI value between e13 and e15 (15.932) surpassed all other observation errors. Consequently, a path was added between e13 and e15. However, despite this modification, the χ² value only diminished to 417.560, signifying that the model fit remained suboptimal.

Following a similar approach, subsequent corrections inevitably led to an improved model fit. Ultimately, after the fourth correction, the χ² value decreased to 376.911, meeting the criteria for an ideal model fit.

Table 12. Results of changes in model fit indicators during the correction process.

Fitting Metrics	Absolute Fit Evaluation						Incremental Fit Evaluation				Simple Fit Evaluation
	χ2	p	χ2/df	GFI	AGFI	RMSEA	NFI	IFI	CFI	RFI	PGFI	PNFI
Judgment criteria	The smaller, the better	>0.05	<3	>0.9	>0.9	<0.08	>0.9	>0.9	>0.9	>0.9	>0.5	>0.5
Initial model	440.736	0.028	1.142	0.914	0.896	0.022	0.897	0.986	0.986	0.884	0.759	0.796
Judgment result	Passed	Failed	Passed	Passed	Failed	Passed	Failed	Passed	Passed	Failed	Passed	Passed
First revision	417.560	0.122	1.085	0.918	0.901	0.017	0.903	0.992	0.992	0.890	0.760	0.799
Judgment result	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Failed	Passed	Passed
Second revision	395.397	0.333	1.030	0.922	0.906	0.010	0.908	0.997	0.997	0.896	0.762	0.801
Judgment result	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Failed	Passed	Passed
Third revision	387.823	0.422	1.013	0.924	0.908	0.007	0.910	0.999	0.999	0.897	0.761	0.801
Judgment result	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Failed	Passed	Passed
Fourth revision	376.911	0.564	0.987	0.926	0.910	0.000	0.912	1.000	1.000	0.900	0.761	0.801
Judgment result	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed	Passed

depicts the changes in model fitting indicators during the correction process and highlights the corresponding results of the adjustments.

The fit of the refined model was enhanced, with each indicator meeting its designated criteria. Notably, the CFI and RFI indicators were considered as 1 since their actual values exceeded 1. The corrected structural model is displayed in Figure 4.

4.2.3. Parameter Estimation and Hypothesis Testing of the Model

Further examination of the individual path coefficients, path coefficient significance, and standardized path coefficient results is presented in

Table 13. Parameter estimation results.

Pathway			Unstandardized Regression Coefficient	Standardized Regression Coefficient	S.E.	C.R.	p
SN1	←	Subjective norms	0.707	0.831	0.047	15.039	***
SN2			0.687	0.747	0.051	13.368	***
SN3			0.602	0.695	0.049	12.232	***
PBC1	←	Perceived behavioral control	0.674	0.733	0.049	13.622	***
PBC2			0.801	0.807	0.052	15.449	***
PBC3			0.718	0.779	0.049	14.711	***
PBC4			0.668	0.713	0.051	13.133	***
EC1	←	Economic cost	0.675	0.752	0.048	14.115	***
EC2			0.759	0.821	0.048	15.878	***
EC3			0.727	0.733	0.054	13.566	***
EC4			0.671	0.717	0.051	13.224	***
EST1	←	External stimuli	0.806	0.807	0.049	16.524	***
EST2			0.888	0.917	0.045	19.724	***
EST3			0.631	0.715	0.047	13.370	***
EST4			0.640	0.741	0.046	14.050	***
EST5			0.569	0.711	0.041	13.903	***
EST6			0.737	0.778	0.047	15.611	***
PTP1	←	Product performance	0.751	0.803	0.048	15.777	***
PTP2			0.687	0.779	0.045	15.178	***
PTP3			0.562	0.684	0.045	12.626	***
PTP4			0.695	0.747	0.049	14.257	***
PTP5			0.678	0.716	0.051	13.410	***
dATB3	←	Behavioral attitudes	0.826	0.725	0.070	11.823	***
ATB2			0.968	0.788	0.078	12.427	***
ATB1			1.000	0.769			***
ATB4			0.869	0.696	0.077	11.268	***
ESI3	←	Purchase intention	1.071	0.787	0.085	12.644	***
ESI2			0.939	0.770	0.078	12.061	***
ESI1			1.000	0.752			***
ESI4			0.739	0.656	0.072	10.198	***

Note: *** p < 0.001, two-tailed.

. The path coefficients for each observed variable achieved significance at the 0.001 level, with overall values ranging from 0.656 to 0.917. Additionally, the t-values of the C.R. test all exceeded 1.96 and exhibited significance at the 0.001 level, signifying that the parameter estimates successfully passed the significance test. This indicated that the latent variables of consumers’ behavioral attitudes, subjective norms, perceived behavioral control, economic cost, external stimuli, product performance, and purchase intention were significantly influenced by the observed variables.

The above structural equation model was evaluated and tested using AMOS 24.0 software, as shown in

Table 14. Path coefficient estimates and hypothesis testing results.

Pathway			Unstandardized Regression Coefficient	Standardized Regression Coefficient	S.E.	C.R.	p	Hypothesis Testing
Behavioral attitudes	←	Product performance	0.213	0.284	0.050	4.265	***	Passed
Behavioral attitudes	←	External stimuli	0.117	0.155	0.047	2.479	0.013	Passed
Behavioral attitudes	←	Subjective norms	0.137	0.183	0.051	2.689	0.007	Passed
Purchase intention	←	Behavioral attitudes	0.194	0.210	0.058	3.358	***	Passed
Purchase intention	←	External stimuli	0.024	0.035	0.036	0.677	0.498	Failed
Purchase intention	←	Subjective norms	0.190	0.274	0.041	4.591	***	Passed
Purchase intention	←	Perceived behavioral control	0.104	0.150	0.039	2.686	0.007	Passed
Purchase intention	←	Product performance	0.268	0.386	0.042	6.362	***	Passed
Purchase intention	←	Economic cost	0.242	0.350	0.041	5.968	***	Passed

Note: *** p < 0.001, two-tailed.

.

The results from Table 14 and Figure 4 indicate that out of the nine research hypotheses in this study, only one hypothesis (H7) was rejected. The path standardized coefficients for the remaining eight hypotheses (H1, H2, H3, H4, H5, H6, H8, and H9) ranged from 0.150 to 0.386, with a significance level of p < 0.01, demonstrating a significant direct positive effect on purchase intention for behavioral attitudes, subjective norms, perceived behavioral control, economic cost, external stimuli, and product performance.

Based on the empirical analysis, it was observed that among the direct factors influencing “behavioral attitudes”, the rank order of influence for each reason variable was as follows: product performance > subjective norms > external stimuli. This indicated that “product performance” was the most influential factor on “behavioral attitudes”, while both subjective norms and external stimuli also contributed to promoting a positive attitude. Furthermore, regarding the direct factors influencing “purchase intention”, the order of influence for each reason variable was as follows: product performance > economic cost > subjective norms > behavioral attitudes > perceived behavioral control > external stimuli. This underscores the significance of “product performance” in driving changes in “behavioral attitudes” and its central role in the ultimate “purchase intention”, warranting additional attention.

4.2.4. Analysis of Model Effects

The SEM in this study’s path diagram only captured the direct influence relationship between variables and lacked the intuitive reflection of indirect effects. Thus, the indirect and total effects between variables were calculated based on the direct effects to facilitate further analysis and explanation. To calculate the indirect and total effects of “product performance” on “purchase intention”, the following formula was utilized:

Indirect effect = direct effect 1 × direct effect 2

Total effect = direct effect 3 + indirect effect

With reference to Figure 4 and Table 13, the indirect effect was calculated as follows:

Indirect effect = 0.284 × 0.210 = 0.446

and the total effect was calculated as follows:

Total effect = 0.386 + 0.446 = 0.832

In this context, “direct effect 1” corresponds to the direct effect of “product performance” on “behavioral attitudes”, “direct effect 2” represents the direct effect of “behavioral attitudes” on “purchase intention”, and “direct effect 3” denotes the direct effect of “product performance” on “purchase intention”.

With reference to

Table 15. Values of each effect between latent variables.

Pathway			Direct Effect	Indirect Effect	Total Effect
Behavioral attitudes	←	Subjective norms	0.183		0.183
Behavioral attitudes	←	External stimuli	0.155		0.155
Behavioral attitudes	←	Product performance	0.284		0.284
Purchase intention	←	Behavioral attitudes	0.210		0.210
Purchase intention	←	Subjective norms	0.274	0.312	0.586
Purchase intention	←	External stimuli	0.035	0.033	0.068
Purchase intention	←	Product performance	0.386	0.446	0.832
Purchase intention	←	Perceived behavioral control	0.150		0.150
Purchase intention	←	Economic cost	0.350		0.350

, the direct effects of potential variables were outlined. Notably, product performance exhibited the most significant impact on purchase intention. Hence, it is crucial for CCRCs to prioritize quality assurance during the development process to fundamentally enhance consumers’ purchase intention.

When analyzing Figure 4 and Table 15, the standardized path coefficient was categorized into three levels. If the coefficient exceeded 0.3, this indicated a highly significant influence; if it fell between 0.2 and 0.3, this suggested a significant influence; and if it was below 0.2, this implied a weak influence. Applying this criterion, this study calculated the total effect of each observed variable on purchase intention and conducted sorting and classification. The obtained results are presented in

Table 16. Classification of effects of observed indicators on the willingness to purchase CCRCs.

Degree of Impact	Influencing Factors	Total Effect	Sort
First class	PTP1	0.68	1
	PTP2	0.648	2
	PTP4	0.622	3
	PTP5	0.596	4
	PTP3	0.569	5
	SN1	0.487	6
	PBC3	0.456	7
	SN2	0.438	8
	SN3	0.407	9
Second class	EC2	0.287	10
	EC1	0.263	11
	EC3	0.257	12
	EC4	0.251	13
Third class	ATB2	0.165	14
	ATB1	0.161	15
	ATB3	0.152	16
	ATB4	0.146	17
	PBC2	0.121	18
	PBC1	0.110	19
	PBC4	0.107	20
	EST2	0.062	21
	EST1	0.055	22
	EST6	0.053	23
	EST4	0.050	24
	EST3	0.049	25
	EST5	0.048	26

.

4.3. Discussion and Recommendations

4.3.1. Discussion

Examining Table 16 reveals that consumers primarily prioritize five factors within the “product performance” dimension (PTP1, PTP2, PTP4, PTP5, PTP3), which include the healthiness of CCRCs, convenience, medical and nursing care, cultural and recreational services, and alleviation of elderly pressure.

Moreover, the subjective norms (SN1, SN2, SN3) exhibit a notable influence on the inclination to purchase CCRCs. Particularly, the endorsement of real estate acquisition by family members (SN1) significantly impacts purchase intention. In China, the prevailing notion of familial elderly care and the weight of familial consensus in major decisions, such as real estate acquisition, underscore the importance of familial support. Thus, enhancing societal acceptance and acknowledgment of CCRCs is essential for fostering consumer confidence and purchase intention. Furthermore, in the era of information technology, favorable public opinion can steer societal perceptions positively. Consequently, in the realm of CCRCs, the opinions of acquaintances (SN2) and governmental stances (SN3) wield significant influence throughout the industry lifecycle.

The substantial influence of subjective norms on the CCRC market also highlights an important trend, namely, the pursuit of homophily [36]. In the context of CCRCs, homophily refers to a community of residents with similar needs, backgrounds, or characteristics. This homophily fosters communal interactions and resource sharing, nurturing a tighter-knit community and sense of belonging. As consumers seek to mitigate costs and optimize social health performance while yearning for a sense of camaraderie and quality living environments, the promotion of diverse socialization pathways becomes paramount. Homophily also guides the planning and design of CCRCs; by tailoring amenities and services to specific homogeneous groups, developers can better meet resident needs, thereby enhancing the living experience and quality of life. Moreover, homophily cultivates a distinct community culture within CCRCs, promoting collaboration and collective advancement.

Furthermore, amidst the pursuit of homophily, CCRCs are transitioning towards socialization. Cousin et al. [37] emphasized the critical role of socialization in community formation and cohesion. In mixed social health environments like senior living, where individuals from diverse backgrounds converge, fostering socialization networks among residents fosters collaboration and communion within the community. Such insights provide valuable guidance for promoting interaction and cooperation within mixed-society health environments like CCRCs, promoting sustainable community development and a thriving environment.

Consumer evaluations of whether CCRCs meet their needs (PBC3), a key aspect of rational consumption decision-making, significantly influence purchase intentions. Hence, the construction of CCRCs should prioritize age-friendly design and diversification to cater to varying senior citizen needs.

In summary, CCRCs should initially establish a reputable brand through exemplary “product performance” to assuage consumer concerns. Strengthening public awareness and understanding of CCRCs can enhance societal recognition, thereby fostering positive “subjective norms”. Simultaneously, tailored marketing strategies for CCRCs should be devised. Additionally, diversified CCRC projects should be developed to accommodate diverse elderly needs, thus fueling sustainable development.

4.3.2. Recommendations

(1)

Foster a conducive market environment for CCRCs.

Traditional social perceptions of aging and consumption habits in China somewhat constrain the uptake of CCRCs, hindering their development. Hence, the government should take the lead in fostering a supportive atmosphere for social awareness and CCRC enterprise development. Through publicizing knowledge related to CCRCs, the government can enhance public understanding, foster societal acceptance, and gradually shift aging paradigms.

(2)

Enhance policy incentives for CCRCs.

As a nascent and government-centric industry, initial CCRC development necessitates leveraging governmental support. Consumers prioritize product performance and economic costs when purchasing CCRCs, highlighting the importance of quality assurance and economic regulation. Therefore, the government should utilize policies to incentivize real estate enterprises, bolstering their commitment to public benefit. Additionally, preferential policies such as loans, subsidies, and tax breaks can reduce construction and consumer spending costs, bolstering CCRC attractiveness and demand.

(3)

Refine the evaluation index system for CCRCs.

Given the burgeoning potential of CCRCs, a robust evaluation index system is imperative for their controlled and optimized development. Similar evaluation systems exist in CCRC fields, offering objective, fair, and comprehensive project evaluations to aid consumer decision-making. Consequently, as the golden era of CCRCs approaches, the government should collaborate with enterprises to establish an evaluation index system, fostering higher-quality projects and enhanced consumer experiences.

(4)

Transition occupancy models of CCRCs.

As a migrant city, Shenzhen grapples with a significant net population inflow and limited land resources, posing challenges to its housing reform system. Consumer concerns about economic costs underscore the need to explore occupancy modes in the CCRC sector. Consequently, examining occupancy modes can alleviate enterprise costs, reduce consumer expenditure pressures, and align with the industry goals of public benefit.

(5)

Adapt existing properties for the aging community.

Given the aging nature of CCRCs, repurposing existing housing with age-adaptive renovations offers a viable solution. Retrofitting existing structures with barrier-free facilities can expedite construction timelines, lower costs for enterprises and consumers, and facilitate aging in place, thereby addressing the scarcity of land resources in Shenzhen.

5. Conclusions and Prospects

5.1. Conclusions

This study employed a literature analysis to determine that, with the phenomenon of increased aging in China, CCRCs have emerged as a crucial focus for the transformation and upgrading of the real estate industry, representing a new avenue for pursuing profits. However, the current consumer behavior in the CCRC market presents a less optimistic outlook. Consequently, there is an urgent need to investigate the factors influencing the purchase intention of CCRCs to devise strategies for the development of the commercial CCRC market and advancing positive aging initiatives. As a result of this investigation, the following conclusions have been made.

(1)

This paper introduced three variables—“economic cost”, “external stimuli”, and “product performance”—within the framework of the TPB, establishing an extended model to analyze the factors influencing the purchase intention of CCRCs. Empirical analysis revealed that “behavioral attitudes”, “subjective norms”, “perceived behavioral control”, “economic costs”, and “product performance” significantly impact purchase intention. Specifically, heightened positive behavioral attitudes, greater pressure from subjective norms, and perceived ease of purchase all contribute to a stronger purchase intention. The order of influence for each variable was as follows: product performance > economic cost > subjective norms > behavioral attitudes > perceived behavioral control > external stimuli, signifying those consumers place the greatest emphasis on the product performance of CCRCs, followed by price.

(2)

The path coefficient between “external stimuli” and “purchase intention” was found to be insignificant, suggesting a lack of a direct influence between these factors. However, the path coefficients between “external stimuli” and “behavioral attitudes” as well as between “behavioral attitudes” and “purchase intention” were both significant. Consequently, the indirect effect of “external stimuli” on “purchase intention” emerged as significant. This finding indicates that while external stimuli do not directly impact consumers’ purchase intention, they do exert influence through a process of information transmission and transformation that ultimately shapes their purchase intention.

(3)

The three variables introduced in this paper significantly impacted the respective outcome variables. Among them, the three causal variables of “behavioral attitudes” had a significant effect on them. At the same time, “behavioral attitudes” had a significant impact on the final “purchase intention”, which connected the indirect influence path of these three causal variables on “purchase intention”. This underscores the influence of “behavioral attitudes”, shaped by multiple factors.

(4)

The direct influences of “subjective norms” and “product performance” on “purchase intention” were 0.274 and 0.386, respectively, surpassing the impact of “behavioral attitudes” on “purchase intention”, which stood at 0.210. This indicates that consumers’ purchase intention of CCRCs is more profoundly influenced by “product performance” and “subjective norms”.

5.2. Future Prospects

CCRCs in China are still in their infancy, lacking complete theoretical research and practical experience. While this paper has provided a detailed discussion and deep analysis of the influence of seven potential variables on the purchase intention of CCRCs, utilizing SEM for testing and analysis, it is constrained by the complexity of the research subject and the limitations of other factors. The following issues or deficiencies remain for further improvement:

(1)

The questionnaire data collected in this study were limited due to the high population mobility in Shenzhen and dynamic changes in demand. Future research should aim to enhance the generalizability of the findings by enlarging the sample size.

(2)

This paper primarily focused on the cognitive factors influencing the formation of purchase intention before actual consumption behavior of CCRCs occurs. However, this study did not track actual consumption behavior over time or evaluate post-consumption satisfaction. Future research should delve into these aspects.

(3)

This study primarily examined the influence of six potential variables on purchase intention, without analyzing demographic variables such as age and health status. We hope to address this gap in subsequent research by incorporating and discussing demographic variables for a more comprehensive understanding.