This section describes empirical case selection, data collection, and the analytical procedure in detail. Moreover, based on the INRM and the performance evaluation result, the priority of improvement and comprehensive analysis are primarily discussed.
3.1. Empirical Case Description
The case selected for this study is an open space in the main campus of the Harbin Institute of Technology (HIT) (
Figure 3). The campus is located in the city center with a total area of approximately 66.8 ha, a building area of 1,270,000 square meters, and a floor area ratio of nearly 1.3. According to historical records, several functional areas, such as student and teacher living areas, teaching areas, sports zones, and logistics areas, have been gradually built in the main campus of HIT for different purposes and ages, without following the holistic principles of campus planning. Moreover, no clear boundary restricts access to the main campus, and an urban arterial road has even been constructed across the POS (public open space). This POS serves an educational function, is part of the urban POS system, and is open for public usage because of the POS insufficiency in Harbin city center. The reasons for selecting the open space in the main campus of HIT as a case study are as follows:
First, Harbin city is facing a serious trend of population aging. It is predicted that the population aged over 60 years will increase from approximately 1,840,000 (19.2% of the total population) to approximately 2,150,000 (approximately 22% of the total population) from 2015 to 2020. The aging society, the “empty nest” phenomenon, shrinking family sizes, and increasingly elderly households have become more prominent; moreover, the total elderly population has a certain proportion with low income levels. Thus, developing a sustainable society for the elderly population is necessary.
Second, the study site is located in the inner city, which has a high proportion of elderly people as residents, as was confirmed by the data collected from several street organizations located close to the POS (collection date, 29 August 2016). Some communities have a high proportion of elderly residents. For example, elderly people numbering 1743 comprise approximately 24.5% of the Fuhua community, 1483 of the HIT community (30.9%), 1091 of the Guixin community (14.2%), 1397 of the Chunming community (18.7%), 1488 of the Nonghang community (23.6%), and 1009 of the Tiewa community (13.3%). According to the statistical information from the questionnaire, elderly people who participate in outdoor activities in the POS at HIT are mainly from these nearby communities.
Third, the government and HIT implemented a joint program to gradually improve the environment (i.e., POS) for elderly people. With the development of the concept of urban sustainability concept and because of new environmental demands resulting from the changing population structure—such as the built urban environment being unable to meet the comprehensive needs of the elderly population—the local government has implemented an urban redevelopment process to improve the physical environment for the elderly population. The interim assessment report of the 12th Five-Year Plan for the elderly population revealed that Nangang district lacks open spaces and infrastructure facilities compared with other districts in Harbin city [
68]. This POS insufficiency is attributed to the high building density and the lack of storage land. Therefore, the 13th Five-Year Plan [
69] addresses the issues of the elderly population through an urban development strategy and accelerates the development of a liveable environment system and activity places for the elderly. Thus, to improve the POS in Harbin city center, the local government assisted HIT in redeveloping the campus and open space for the elderly population. Sharing POS with the campus is an effective approach to enhancing utilization efficiency and maximizing joint use opportunities of compatible facilities and space. These redevelopment projects in HIT and the environmental reconstruction projects for an aging society promoted by local government in the city center have been implemented near simultaneously. After obtaining funding, the building schedule and other key projects have been established gradually for the redevelopment of the POS.
The current status of the redevelopment project in the campus is as follows. For the regional boundary, the construction of the front gate has been completed at the entrance to monitor external vehicles and individuals. No large area lawn exists due to the dual effects of restrictive land use and climate. Thus, for the purposed of greening, the larger green area includes three nearly 1-ha green spaces (Vineyard, Alumni Woods, and the backyard of the Hang Tian building;
Figure 4b,c,j, respectively), and the remaining small green area consists of trees beside the street and flowerbeds in front of each college (
Figure 4h,q, respectively). For the waterscape, the POS initially included an outdoor swimming pool and a fountain, but these features were abandoned because of the difficulty of restoration and maintenance (
Figure 4a,b, respectively). The public facilities of HIT are a large supermarket, three small kiosks (
Figure 4k), and toilets on the ground floor of various departments including the teaching building and the canteen. The POS has two historic buildings and two monuments or sculptures of cultural significance, including the “Star of HIT” (depicting the minor planet named after HIT;
Figure 4o) and a globose sculpture relying on the technology of Seamless Welder (
Figure 4m). Other types of construction, such as ramps, sidewalks, and building entrances, have currently achieved the minimum standards of the Codes of Accessibility Design (GB50763-2012;
Figure 4g,h,k,l). In addition, the university and activity centers for elderly people have been designated for open use.
In conclusion, based on the rapid growth of the elderly population in Nangang district and Harbin city and the progressive urban redevelopment measures of the local government, the POS in HIT was selected as a highly appropriate empirical case to research continuous improvement strategies for a healthy aging society.
3.2. Data Collection
This section explains the operation of data collection and the choice of interviewees, including the expert and user groups. A DEMATEL questionnaire was employed to explore the relationships between the 16 criteria, meaning that experts were suitable interviewees to answer the questions. A VIKOR questionnaire was used to collect data on the degree of satisfaction with each criterion, so this questionnaire was completed by the elderly participants. Decisions being made by experts on behalf of users was formerly a conventional means of decision-making, but is now often considered controversial. So this study interviewed two groups (experts and users) for the satisfaction section and compared their results.
The expert questionnaires were separately administered in the first and third phases (
Figure 2). In the first phase, 13 DEMATEL questionnaires were returned, comprising four invalid questionnaires and nine valid questionnaires. The expert interviewees were selected as follows. First, interviewees came from a wide variety of industry, official, and university circles; second, the interviewees were required to have a deep understanding of the topic at hand, whether theoretically or practically. In a detailed description, eight experts engaged in urban planning or landscape planning and had landscape design experience related to the pension industry, and the remaining experts were from the Department of Health Industry Management and had considerable research experience related to the health of the elderly population. Five experts were professional scholars, of whom two were professors, and the others worked in design institutes and had the highest qualification. Third, the chosen interviewees were required to understand the purpose and operating process of the DEMATEL questionnaire. The aforementioned four questionnaires were judged invalid largely for this reason. To improve validity, questionnaires were administered in the form of structured interviews, and any confusion about the questions was cleared by the authors. The average time for questionnaire completion was approximately 3.5 h.
In the third phase (
Figure 2), 40 performance (modified VIKOR) questionnaires were administered over the Internet to another expert group in person-to-person form; therefore, all 40 questionnaires were valid. All of the experts graduated from the architecture school of HIT, implying that they had sufficient knowledge of the study site, as well as being landscape planners or urban designers with professional understanding of environmental gerontology. Hence, these experts could answer the questionnaires, which contained 16 evaluation criteria of POS and the satisfaction degree of each criterion. Moreover, the responses about the each criterion of the study site were rated on a scale from 1 to 10, with 1 representing extremely dissatisfied and 10 extremely satisfied.
The interviewees who were administered the modified VIKOR questionnaire in the third phase were people aged over 65 years (defined as elderly according to the WHO) and were active on the HIT campus from 5 July to 5 October 2016. According to their own experience of the HIT campus, these interviewees provided their satisfaction degree for the 16 evaluation criteria of POS in the paper-and-pencil questionnaire. The author and four trained graduate students randomly distributed the questionnaire on campus at five different periods (05:00–07:00, 07:00–11:00, 11:00–17:00, 17:00–21:00, and 21:00–24:00). When recruiting elderly people, the researchers confirmed whether they met the study requirements and obtained their consent for the interviews. In the survey, the preferred participants were elderly people who were resting, reading books, or exercising (stretching or walking) rather than those performing collective activities and looking after children. The researchers explained that the survey data will remain confidential, participation is voluntary, and that the elderly people could abstain from participating in this study. The research questions about social statistical information, detailed address, and other confidential information had to be as clear as possible to avoid any misunderstanding from the elderly participants. Two questions (“Whether Living Alone” and “Monthly income”) were deleted from the questionnaires, because a large number of participants did not respond to these questions. Some participants expressed distrust for the research generally, had concerns about privacy, and believed that the information might be used to make unwanted changes in their neighborhood. In addition, for respondents with visual impairment, researchers read the questionnaire and helped them to complete the questionnaire. A total of 205 questionnaires were administered and collected, including 59 invalid questionnaires. A questionnaire was considered invalid when a substantial number of questions in the survey were left unanswered. Among the target population (65 years or older), 79 (54.1%) were women. The majority (91.8%) of participants who completed the questionnaire were healthy and active, and 8.2% reported some mobility limitation. A total of 100 (68.5%) persons visited the POS daily.
3.3. Result and Discussion
The INRM in
Figure 5 illustrates the structure of interactions in the improvement model. Social interaction (
D3) was influenced by mobility (
D1) and nature enjoyment (
D2), whereas mobility (
D1) was influenced by nature enjoyment (
D2). Therefore, to improve the social interaction (
D3) dimension of POS, the improvement of criteria in the mobility (
D1) and nature enjoyment (
D2) dimensions should be emphasized. Moreover, to improve the mobility (
D1) dimension of POS, the improvement of criteria in the nature enjoyment (
D2) dimension should be emphasized. Thus, the priority of influence of the three dimensions could be ordered as nature enjoyment (
D2), mobility (
D1), and social interaction (
D1). Similarly, from the perspective of INRM, the route pattern (
C14) was the most influential criterion in mobility (
D1), and the remaining criteria could be ordered as
C13,
C15,
C12, and
C11. High-quality vegetation (
C22) was the most influential criterion in nature enjoyment (
D2), and the remaining criteria could be ordered as
C23,
C24,
C21, and
C25. Co-maintenance spaces (
C33) were the most influential criterion in social interaction (
D1); the remaining criteria could be ordered as
C36,
C35,
C32,
C31, and
C34.
The empirical case for using POS to achieve a healthy aging society in Harbin can be established through performance evaluations and improvements that incorporate VIKOR, based on the IWs found through DANP (
Table 2). The total gap value is a sum of all criterion gap values, which can be obtained from the product of each local weight and its gap. A gap represents the distance between a criterion’s performance and the aspiration level. In this study, two gap values were derived from the expert and elderly groups. Furthermore, these gaps were improved to enhance each dimension/criterion’s performance.
The integration values for performance and gaps are presented in
Table 2. The nature enjoyment (
D2) dimension exhibited the largest gap value (0.473), and among its criteria, water features (
C23) exhibited the largest gap value (0.674). Therefore, the POS can proceed with priority improvement by using the criteria with the largest gap value. The following criteria exhibited large gap values: waste management (
C24) (0.612), Co-maintenance spaces (
C33) (0.606), natural art design (
C25) (0.405), and high-quality vegetation (
C22) (0.403). The following criteria exhibited relatively small gap values: route patterns (
C14) (0.303), seats on pathways (
C31) (0.320), and high-density vegetation (
C21) (0.322); however, their absolute values were still large (to 0). Thus, this POS has serious problems, and several improvements need to be implemented to achieve a POS that realizes healthy aging goals.
For the continuous improvement strategy for the POS in this empirical case, the influence relationships and the influential degree of the dimensions and criteria are discussed. Based on the INRM (
Figure 5) and performance evaluation results (
Table 2), two primary discussions are elaborated as follows: priority of improvement and comprehensive analysis according to the evaluation results of the expert and elderly groups.
To determine improvement priorities, gap values should be considered first. The absolute integration gap values of all three dimensions in
Table 2 were significantly high and ranked as follows: nature enjoyment (
D2), 0.473 (to 0); social interaction (
D3), 0.387; and mobility (
D1), 0.360. This result indicates that the development of POS in HIT for healthy aging is still in the initial stage, and the performance level is relatively low. Mobility (
D1) exhibited a relatively small gap value, indicating that this dimension was regarded as more essential and received relatively more support from decision makers. The last stage of the improvement project focuses on enhancing mobility and accessibility. Moreover, it has achieved the minimum standards of Codes of Accessibility Design (GB50763-2012). Despite these measures, the needs of the elderly population have not been fully met. Yung et al. [
8] argued that POS planned and designed simply based on universal accessibility standards cannot completely fulfill the needs of elderly people. They also suggested the effectiveness and the inclusiveness of the evaluation standards of POS should be regularly reviewed and adjusted. Some criteria exhibited extremely large gap values, such as water features (
C23) (0.674), waste management (
C24) (0.612), co-maintenance spaces (
C33) (0.606), natural art design (
C25) (0.405), and high-quality vegetation (
C22) (0.403), all of which belong to the dimensions of nature enjoyment (
D2) and social interaction (
D3). This finding implies that these aspects of the POS in HIT should be improved first. POS improvement was previously implemented based on the priority order from the criterion with the largest gap value to the criterion with the smallest value. By contrast, the DANP-V model uses a systematic approach to address the problems from the root. For example, in this case, water features (
C23), which exhibited the largest gap value among the 16 criteria, was given the first priority for improvement following the conventional approach. In fact, the POS of HIT previously had two water features (a swimming pool and a fountain), which the elderly interviewees reminisced about. These areas had been transformed into small squares and a parking area (
Figure 4a), because of land use limitations and the difficulty of maintaining them in a cold climate. Moreover, it would be difficult to build a new water feature. However, the DANP-V model could identify the criteria affecting the satisfaction degree of water features.
Figure 5 shows that the satisfaction degree of high-quality vegetation (
C22) can affect the satisfaction degree of water features. This finding suggests that for this case, decision makers should allocate more resources to the root of the problem. Moreover, waste management (
C24) exhibited the second largest gap value. Some residents believed that more disposal facilities should be provided for waste from pets, although the existing facilities met the current standards (1/100–200 m) in the Code of Urban Environmental Sanitation Facilities Planning (GB 50337-2003) (
Figure 4f,l,n). However, others argued that even with sufficient sanitation facilities, it is difficult to guarantee a clean and tidy environment (
Figure 4e); they considered unsanitary behavior to be a problem of upbringing. By contrast, this result revealed a new perspective: that the satisfaction degree of waste arrangement (
C24) was influenced by water features (
C23) and high-quality vegetation (
C22), as shown in
Figure 5. As mentioned, the satisfactoriness of water features (
C23) can be improved by providing high-quality vegetation (
C22). According to the performance and gap values of high-quality vegetation (
C22) (ranked fifth) in
Table 2, the satisfaction degree of this criterion was also very low. In recent years, the greening projects of the POS in this case mainly concentrated on planting more trees. The satisfaction degree of high-density vegetation (
C21) was higher than that of high-quality vegetation (
C22), indicating that the elderly respondents gained more travel motivation from the quality rather than quantity of natural elements (
Figure 4c,i). To date, many problems have been detected with respect to vegetation arrangement in this POS, such as the plant combination pattern, coordination with existing landscape elements, and landscape conservation and management (
Figure 4d,l,p,r). From the results, the low quality of the natural landscape is likely to result in uncivilized behavior, which may be regarded as a sign that elderly people are not safe and unwelcome. As indicated by the broken windows theory, a low quality environment can powerfully affect and induce the formation of negative attitudes that exhibit no awareness of environmental protection and are inclined to immoral behavior [
70]. Similarly, Gibson and Werner [
71] used the empirical method to explore the relationship between smokers’ behavior and the environment, and they discovered that uncivilized behavior is not necessarily related to personality problems and is more likely to be influenced by environmental designs and their implications. Therefore, attention should be paid to improving the quality of vegetation design. As demonstrated by Finlay et al. [
42], environmental interventions that positively influence elderly adults’ quality of life need not even be on a largescale. Smaller features, such as a small and exquisite green plot or a courtyard fountain, can have a significant effect.
In this study, performance and gap values were synthetically obtained from the survey data of both experts and elderly participants. First, the gap values from the designer questionnaires were generally larger than that from the elderly questionnaires. This difference can be attributed to the following reasons: (1) the standard of professional assessment may be stricter than that of the populace and (2) difference in living quality expected by different age groups. Kane [
72] found that the expectations of quality of life among seniors decline with age. Regarding subjective impression of well-being, elderly participants seemed more optimistic and therefore often scored higher than younger adults. (3) Compared with designers, the elderly people interviewed in this study have typically lived around HIT for a long time and experienced a long self-adaptive process. According to the person–environment fit theory proposed by Lawton [
16], the self-adaptive ability of the elderly population enables their adjustment to external environmental stimuli when exposed to those stimuli for a long time. This concept can explain why elderly people express comparatively more satisfaction, even when facing inclement environments.
Second, these findings, derived from the comparison of the results of two groups, indicate that experts and elderly people provide generally consistent arrangement patterns for the criteria, particularly the first three criteria. However, we also need to pay attention to the significant difference in two criteria orders. For sufficient lighting (C15) (nine different orders), it is more likely that designers believe that elderly people prefer exercising in the daytime, and the existing lighting facilities can meet their demands. However, 54.1% of elderly respondents were active in the evening (approximately 17:00–21:00), and some believed that the brightness of the lighting facilities was insufficient. For intimate seating C34 (six different orders), the privacy degree required by the elderly in the case is relatively low. Nonetheless, designers use numerous evaluation standards, which are mostly obtained from their own experience of nursing projects for the elderly population, causing the larger gap values. Finally, considering the special needs of elderly people, the findings of this study indicate that public participation should be emphasized. Elderly people should be encouraged to communicate their preferences for the improvement process of POS for healthy aging. Combining the evaluation results from experts and users, decision makers can more favorably identify the requirements and priority orders from user groups.