Applying multi-criteria decision analysis to prioritise age-friendly criteria for policy implications

ABSTRACT

Significant increase in the proportion of elderly and its expected further increase in the days to come draw attention of policymakers, practitioners, and researchers to enhance a community’s age-friendliness. Enhancement of a community’s age-friendliness requires appropriate policy implications recommended by various stakeholders based on Multi-Criteria Decision Analysis (MCDA) followed by Friedman’s rank test. Results showed that ‘safety and security’ has the highest priority amongst all the criteria considered for the assessment of age-friendliness. Based on Friedman’s rank test, community boundary with security check posts has the highest priority to increase safety and security of the community. A combination of policy measures having higher priority such as ‘police patrolling’, ‘mixed land-use urban planning’, and ‘accessibility to services’ could be the effective solution to increase safety and security of the community environment and foster independent living ability of its residents which in turn play important role in increasing age-friendliness of the community.

KEYWORDS:

• Policy
• enhancement
• safety
• assessment
• age-friendliness

1. Introduction

Elderly and the disabled are facing increasing challenges because of their reduced capacity to perform normal activities of daily living. With the advancement of time, they feel increasingly dependent upon their younger generations for their subsistence. This situation is a matter of concern now and in days to come as their number and proportion in the society go on increasing due to better healthcare facilities resulting in both fertility and mortality rate decline and increase in dependency ratio (Office of the registrar general & census commissioner India 2006). The traditional Indian joint family system, mostly the outcome of India’s agrarian economy, also tends to break and results in the emergence of the nuclear family. As a result of this, care for the elderly and the disabled in the family cannot be properly taken as was witnessed earlier. The elderly feel socially isolated and the elderly abuse increases. They feel themselves worthless and a burden to their family and society. In this context, their community environment plays significant role in safeguarding their interest if it is made age-friendly. Such age-friendly environment adapts its services and structures and fosters independent living capability of all age groups especially the elderly and the disabled. People stay healthy, feel socially connected, and actively participate in making decisions that make them feel their self-worth (World Health Organization 2007; The State of Queensland 2017). The quality of life of people is enhanced as it maximises health, participation, and security (World Health Organization 2007). The concept of age-friendliness is to enable ‘active ageing’ and ‘independent living’ among people in the community that are vital to keep them healthy, wealthy, and prosperous (World Health Organization 2002; Kalache and Gatti 2003). This concept of age-friendliness was developed by WHO during its policy initiatives in the 1990s and early 2000s (Buffel, Phillipson et al., Buffel et al. 2012, 2012). In 2002, its ‘Medical model’ concept was changed to ‘Social Model’ (Ostroff 2010). While ‘Medical model’ concept was based on the assumption that disability is the feature of an individual, ‘Social Model’ concept assumes that the disability is the outcome of the interaction between a person and the environment (Ostroff 2010). So far, many developed countries have taken steps to make their cities age-friendly. World Health Organization also in 2006 had taken a significant step to improve the age-friendliness of 33 cities in 22 countries. The major task of this step was to enlist the presence or absence of age-friendly features with a mission of providing the missing ones. However, this idea of providing age-friendly features at one go does not seem apparent in developing countries like India where per capita income is low and resources are inadequate (Pérez-Cuevas et al. 2015). Also, it is not so necessary to provide less important features at the very first step if the resource is the main constraint since these features do not contribute much to increasing age-friendliness of the community. Therefore, to tackle such a scenario, it is necessary to prioritise the age-friendly criteria based on their weights. Analytical Hierarchy Process (AHP) was found to be the most suitable for determining weights and to prioritise the criteria. This method, alternatively known as Multi-Criteria Decision Analysis (MCDA), determines weights of criteria based on pair-wise comparison of criteria or sub-criteria. Although this method involves various steps aimed to score and categorise the alternatives, i.e. communities, it is used here especially to assign weights and prioritise these age-friendly criteria. However, the scoring and categorisation of communities is beyond the scope of this paper. The prioritisation of criteria and sub-criteria would help academicians, policymakers, practitioners, and researchers to set priorities of action for the implication of policies in governmental and private interventions to improve the age-friendliness of the community. Various policies recommended by the experts during the survey were assigned ranks using Friedman’s rank test. This would help policymakers, and practitioners in deciding the importance of a particular policy for its implementation. Economists from Mumbai University define that consumers spending US $2 to $10 per capita per day belong to middle class (Roy 2018). According to an estimate made by Roy, the population of this middle-income group of India would surpass China, the United States, and Europe (Roy 2018) by the year 2027. It was also estimated that the percentage of middle-income group would increase from 5 to 10% in 2005 to 90% in 2039 (Roy 2018). Keeping in view of such dramatic growth, respondents were chosen from families of middle-income groups belonging to selected MIG communities.

2. Community profile of India

As discussed earlier in the introduction section, in developing countries like India it is impossible to implement all missing age-friendly features at one go because of low per capita income and limited resources. In this context, it is important to prioritise the criteria, and various policy recommendations, which would help policymakers, researchers and practitioners in setting the priority of action in governmental and private interventions. Therefore, the study on the community profile of developing countries is important to know the socio-economic status of the target group i.e. the elderly, and the target family i.e. middle-income group family. In this study, community profile of India is illustrated to acknowledge policymakers, practitioners, and researchers on the applicability of the methods in other developing countries.

In India, the trend of both urbanisation and rapid rate of growth of the elderly has been noticed from the recent past. If India is to follow this trend, it will have 19% of its total population in the elderly category by the year 2050 (Agarwal et al. 2016a). World Health Organization (WHO) conducted a survey in 33 cities of 22 countries in 2006 to find the issue of cities not being age-friendly. This age-friendly initiative was implemented in three states of India, i.e. Delhi, Udaipur, and Kolkata. In Delhi, mobility and accessibility were found to be the major issues (Singh 2016). The application process to avail of concessions in travel was complex and cumbersome. The elderly found difficulty in embarking and disembarking when bus drivers did not stop close to the curb. The overcharging by auto-rickshaws, the ill-designed or inadequate signage, illegal incursion by shops or the hawkers, and inadequate or absence of lighting were some of the major issues. 55% of the total road length was only available as footpaths and 10–15% of the footpath was constructed according to the building by laws. The common areas of buildings were not maintained properly. Staircases were not given importance for which these remained dirty and dark. Public toilets were insufficient in number and did not have arrangements for the disabled. In the survey conducted for prioritising age-friendly criteria in the study area, i.e. Bhubaneswar Municipal Corporation through expert’s opinion, safety, security, affordability, and accessibility were found as the major issues. Experts were chosen from various fields within study area having in-depth knowledge of age-friendliness for pair-wise comparison among age-friendly criteria considered for the study. Before the expert’s interview, both the pilot study and focus group discussion were also conducted for the identification of age-friendly criteria. 133 communities consisting of middle-income families were identified for the age-friendliness assessment. 13 communities were selected out of these 133 communities by simple random sampling method to carry out the assessment survey. The communities selected for the study were Arya village, Bomikhal, Chandrasekharpur, Khandagiri Bari, Khandagiri Vihar, Kharvel Nagar, Nakhara, Niladri Vihar, Palasuni, Soubhagya Nagar, Sri Vihar, Utkal Nagar, and V.S.S. Nagar. These communities are representatives of the urban region of India as the selection was based on the criteria set for the middle-income group by the Ministry of Housing and Urban Affairs, Government of India (Ministry of Housing and Urban Affairs 2021).

3. Age-friendly criteria and sub-criteria

The literature review gives a comprehensive summary of previous research on steps to improve the age-friendliness of a city or community. It is concluded that the age-friendliness covers a wide range of aspects. According to WHO, it could be assessed by four types of indicators such as ‘input’, ‘output’, ‘outcome’, and ‘impact’. Available resources and structures considered as input indicators would enable interventions in the form of programmes, services, and policies for which these interventions could be considered as output indicators. Output indicators help improve the age-friendliness of both social and physical environment and therefore age-friendliness of a community could be considered as outcome indicators. These outcome indicators in turn would help improve health and well-being of the elderly or the population of the community as a whole for which health and well-being could be regarded as impact indicators (World Health Organization 2015), World Health Organization in its own guide has demonstrated that a significant step to improve age-friendliness is to prepare a checklist of age-friendly features on eight domains classified by it which would help identify and facilitate the missing ones (World Health Organization 2007). Other notable age-friendly initiatives include ‘Visiting Nurse Society of New York’s Advantage’ initiative, U.S. Environmental Protection Agency’s ‘Building Healthy Community for Active Ageing’ initiatives and AARP’s ‘Livable Communities’ initiatives (Scharlach et al. 2014). Consistent with these initiatives countries across the globe have adopted various mythologies to enhance the age-friendliness of their cities. These age-friendly practices would help build the abilities of the elderly to fulfill their basic needs, make decisions, be mobile, and contribute and maintain social relationships (Marston and Samuels 2019). Online learning alternatively known as web-based learning, virtual learning, e-learning, or distance learning has become a popular modality enabling participants to access information and interact with others to enhance their learning abilities (Lehning et al. 2009). It reflects the process of sharing knowledge not only by experts but also by participants which could alternatively be named as ‘collective intelligence’ (Boulos and Wheelert 2007). The goal of ‘creating Age-friendly communities online conference’ developed by policymakers and researchers was to disseminate emerging knowledge on age-friendliness and to make age-friendly communities available on the easily accessible websites (Lehning et al. 2009). Access to services such as physical and social infrastructure makes the elderly remain active and engaged. In a study of comparative analysis of areas of Victoria and Austria, the results showed that regional municipalities have fewer services to access than their metropolitan counterparts (Lowen et al. 2015). Colangeli in his PhD thesis demonstrated different barriers to making cities and communities age-friendly. He argued that professional planners are not ready to solve the problems arising out of demographic change because of the lack of resources, lack of credibility, and perception of the elderly being the neglected section of society. These constraints force them to implement short-term and immediate strategies. In this context, he stated that to make cities age-friendly professional planners need to focus on long-term agenda, land-use planning, improving visionary skills, creating the ability to act as facilitators or educators, maintaining relationship with politicians, and engaging community residents and stakeholders (Colangeli 2010). In addition to the lack of resources, lack of strong leadership and direction was identified by participants as barriers to becoming age-friendly in Manitoba Province of Canada. Therefore, a strong leadership at all levels of government is essential to support communities in making them age-friendly (Menec et al. 2014). Age-friendly policies and programmes adopted by the local Government of San Francisco Bay Area are incentives for creating mixed land use neighbourhoods, improving walkability by the change in services and infrastructure, improving accessibility of public transportation facility, and allowing discount on public transportation fare (Lehning 2010). In Canada, Sao Paulo, South Australia, and Andalusia policy interventions focus mainly on the joint effort of state and municipal Governments, enabling initiatives attractive and visible, collaborative approach through partnership, and ensuring consistency in implementation to make their cities more age-friendly (Plouffe and Kalache 2011a, 2011b). In a study conducted in Quebec city of the Canadian province, the findings showed that collaborative partnership among stakeholders of municipal apparatus, political representation, and community organisations was more effective in improving the age-friendliness of cities than solely involving the elderly in the implementation process (Garon et al. 2014). Europe has implemented the ‘healthy ageing approach’ adopted by World Health Organization on Active Ageing. This approach includes raising awareness among the elderly, empowering both person or community as a whole, increasing accessibility to services, and creating a supportive physical and social environment. This, in turn, would enhance the independent living ability of the elderly to contribute to the social and economic life of a city (Green 2012) (Buffel, Phillipson et al., Buffel et al. 2012)., In a study of the potential role of the village model, research findings suggest that villages have tremendous potential to enhance its resident’s ability to age in place consistent with the goals of age-friendly initiatives of WHO (Scharlach et al. 2014).

Various age-friendly practices implemented worldwide as discussed above give a brief idea of the age-friendly features of a city or community. In addition, a pilot study and a focus group discussion were conducted in the study area along with case studies of old-age homes in Odisha state of India to gather more age-friendly features. The pilot study was conducted not only to identify age-friendly features but also to evaluate the feasibility, resources available, time, and cost for conducting the study. According to Browne, the sample size for a pilot study should be 30 or greater (Browne 1995) to estimate a parameter, whereas Kieser and Wassmer estimated a sample size between 20 and 40 for a main study sample size of 80–250 using 80% upper confidence limit (Kieser and Wassmer 1996). Julious suggested a minimum sample size of 12 subjects per group as a thumb rule (Julious 2005). By considering these rules, 30 respondents were selected for the pilot study. During the pilot study, respondents were asked about the problems they faced or would anticipate while staying in their own home or community, and suggestions for improvement. In addition to this, they were asked about their social, economic, and environmental activities, safety, security, and emergency response experiences, and various ways of communicating and disseminating information while living in the community. The pilot study was followed by a ‘focus group discussion’ involving experts and stakeholders to assimilate knowledge about features of age-friendliness.

All these identified features documented by various means were taken into consideration for classifying them into different heads known as ‘criteria’ and sub-heads known as ‘sub-criteria’. The interpretation of all the criteria and sub-criteria are described below in detail.

3.1 Independence level

A person lives in the society with assistance of his family members or any one of the neighbourhood where he or she lives in. However, our prime concern is to measure the capability of the person to act independently without the help of others for performing activities of daily living. This capability is termed as ‘independence level’, which merely depends upon the availability of services, infrastructure, and equipment that a person requires while performing activities of daily living. In a study conducted by 14 focus groups of residents of age 35 or above, this was identified as one of the four factors responsible for measuring age-friendliness. The following sub-criteria foster the independent living capability of a person.

1. Housing options: People need different types of housing units depending upon their functional ability and age. People, especially the elderly and the disabled could act independently if the available houses are physically designed to meet their specific needs. According to the international review of housing options and concepts, these housing options should be sufficient to facilitate their choice and social contact while maintaining independence (Davey et al. 2004).

2. Transportation options: People require different types of the transportation system to move independently within the community. Novak stated that available transportation options in a community should be adequate to maintain independence and self-sufficiency (Novak 2015). The various types of transportation systems available in a community are public transportation, specialised transportation, assisted personal and voluntary transport services.

3. Home adaptability: A person’s own home should have the option of easily modifying it so that his or her everyday activities can be easily managed. Common types of home modification include widening doors, installing ramps, hand-held showers, and grab bars. Home modification facilitates independence by providing safety, security, dignity, and accessibility and saving the cost (The National Resource Center on Supportive Housing and Home Modification 2003).

4. Equipment usability: The products that are used by the people in the community to perform daily activities are important so far as their mobility is concerned (Cook and Polgar 2014). Some products are specially made to be usable by the people particularly the elderly and the disabled. Products can also be modified to make it usable for people of all age groups depending upon their needs.

5. Accessibility: Accessibility is the feature of products, services, and facilities by which people with a variety of disabilities can use them independently. A person’s own home or the public places could be built with ramps in addition to stairs or step-climbers so that people with disabilities or the elderly find it easy to access. Similarly, public transportation system if made accessible increases independence and freedom of choice of people (Crewe and Zola 2001). In addition to public transportation, access to housing, public places, and information are also equally important.

6. Affordability: A person’s income determines his or her purchasing ability to use services, products, or housing in activities of daily living to become independent. Older people having low income are more likely to have inadequate or poorly maintained housing which causes physical and psychological distresses (Spillman et al. 2012). Therefore, the control over the cost of these services is a prime concern for bringing them within one’s financial ability.

3.2 Social inclusion

Social inclusion is defined as the inclusion of a person in community life for making intra-community relationships. This depends upon the ability to participate in social activities. Examples of different types of social activities include interacting with friends, families, and community members. Social inclusion is promoted if the communities are made age-friendly (Saloojee 2003; Johner 2011; Scharlach and Lehning 2013). The following sub-criteria facilitate the social inclusion of people in the community.

1. Proximity: The community environment can be designed to reduce both physical and social distances among people. This, in turn, reduces social isolation, marginalisation, and depression. Common spaces that facilitate social inclusion are parks, libraries, mixed-income housing units, and integrated classrooms (Mitchell and Shillington 2002).

2. Material well-being: A person’s material resources and safe housing allow him or her to participate fully in social activity that promotes social inclusion (Huxley 2015). Material well-being is assessed by a person’s income, consumption patterns, and the assets that the person possesses.

3. Volunteering: It is the quality of a person by which he or she helps others in terms of doing some activities willingly and without being paid. The voluntary work, if properly organised, would indicate strong economic development (Wu 2011). It is considered a tool for promoting social inclusion among the elderly according to European Union policies (Naegele et al. 2010). The level of volunteering work varies in countries of the European Union. Some countries like Netherlands and UK give importance to voluntary work but there are other countries that give less importance to it (Naegele et al. 2010).

4. Valued recognition: The phenomenon of a person by virtue of which he or she feels encouraged when people in the community give respect to his or her contributions to society is called valued recognition. The person may help the society in terms of doing some activities or providing services. The valued recognition promotes social inclusion (Donnelly and Coakley 2002).

5. Involvement and engagement: If a person is involved in different decision-making activities such as making decisions of his or her own health, housing, and well-being, he or she feels like a part of the community and socially included. Social inclusion has a greater impact on those living in poverty and the marginalised in the way that allow them to participate in decision-making activities. This increases their self-worth (Fitzduff 2007).

6. Family support: In a family, members support each other in performing everyday activities. Usually, the head of the family represents the family before the society or community by participating in decision-related matters. The adolescent with disabilities seeks identity and relationships with the outside members of the community. These people rely heavily on their families (Maxey and Beckert 2017). Therefore, efficient planning and a deterministic approach by the family help their teenager or young adults with disability to maintain social relationships with others in the society (Abery 2006). Family support was identified as an essential component of social inclusion of the elderly according to the expert’s opinion.

3.3 Well-being

It is the state in which a person feels happy, healthy, and comfortable. This is an important feature of age-friendliness (Steels 2015, 2015). The well-being criterion has again been categorised into following sub-criteria.

1. Physical well-being: It is the state of being active physically so that the person is able to perform everyday activities independently. It is obtained by doing regular exercise, good nutrition, and maintaining healthy lifestyle choices. The safe and secure environment of the age-friendly community affects the physical health and well-being of its members (World Health Organization 2007).

2. Social well-being: It is the degree of measure of the social relationship of a person with others in the community by virtue of interaction, communication, and socialisation. The age-friendly environment creates a social environment that supports well-being of elders by enabling them to stay with others (Rootman et al. 2012).

3. Emotional Wellness: It is the state in which a person feels a strong sense of self-worth and self-esteem and develops the ability to share his or her feelings with others in a positive way.

4. Environmental well-being: The pleasant and clean environment has an impact on the health and well-being on individuals. A person having environmental well-being would be able to understand the impact of nature on human beings and take necessary steps to protect the surrounding environment (Smith 2009).

5. Spiritual well-being: When a person is involved in spiritual practices and develops spiritual belief, he or she feels a sense of peace and contentment. This spiritual well-being is experienced more in an age-friendly community (Plouffe et al. 2013).

3.4 Safety and security

An age-friendly environment is safe and secure as it reduces the risk of falls among the elderly and prevents vulnerable elderly from being abused and neglected (WHO, World Health Organization 2007). City’s physical environment has strong influence on mobility, safety, security, health, and social participation. Therefore, the design, construction, and maintenance of outdoor spaces, buildings, and transportation systems of the city are important in making it age-friendly (WHO 2007; World Health Organization 2007).

1. Land use safety: The type of land use has a direct impact on the safety of the community environment. For example, mixed land use type planning protects the community from theft and crime because of all-time human activity. However in modern city planning practice importance is given more to segregation of land uses that results in an unsafe environment and urban sprawl (WHO 2007; World Health Organization 2007).

2. Transportation safety: The life and property in the community can be protected if its transportation system is properly managed in terms of proper zoning, regulation, and development of innovative technology. Available alternative transportation options in the community are particularly helpful to the elderly and the disabled in providing safety to them (Dickerson et al. 2007).

3. Indoor safety: The indoor environment should be safe particularly for the elderly, as they have reduced vision and mobility impairment. Some examples of safety measures inside a home are non-slip floors, grab bars, handrails, floors producing no glare, adequate lighting, and avoidance of levels (Gawron and Rojek-Adamek 2015).

4. Outdoor safety: Safety of the outdoor is also important for the elderly and the disabled. Some examples of safety measures of the outdoor environment are adequate street lighting, police patrolling, aid posts, blue light poles, GPS tracking, and the creation of awareness through community education (Gawron and Rojek-Adamek 2015).

3.5 Information and communication provision

Provision of information at the time of need to the people of the community is important as it helps manage their life and stay up to date through appropriate action. An age-friendly community provides information to its people at right time enabling them to participate in the community and to stay healthy, active, and secure (Everingham et al. 2009).

1. Information ground: Information ground is the medium in which information is shared through social interaction and word-of-mouth. Some examples of information ground environments are newspapers, local media, public posters, and notices through which information is disseminated directly to the community residents (Everingham et al. 2009).

2. Mobile networking: Mobile networking is another important medium for disseminating information with the help of mobile technology. Elderly and the disabled can use the mobile device by acquired knowledge to get access to vital services at any point of time for their security and autonomy (Abascal and Civit 2000).

3. Internet services: Internet service is also a medium of disseminating information. However, with the advancement of age, people find it difficult to use (Everingham et al. 2009). Therefore, due care should be given to the internet to make it accessible and user-friendly.

4. Materials and methods

The criteria and sub-criteria identified through literature, case study, pilot study, and focus group discussion were selected by ‘Modified Fuzzy Delphi Method MFDM’ (Raut and George 2018). These selected criteria and sub-criteria were then assigned weights by the Analytical Hierarchy Process (AHP) or Multi-Criteria Decision Analysis (MCDA). AHP, introduced by Thomas Saaty in the year 1980 was found suitable for this research as it involves decision through pair-wise comparison of both tangible and intangible parameters (criteria or sub-criteria). The error committed by the experts is minimum as the method carries is a test to check inconsistency in the judgement. The hierarchical structure of AHP method is represented graphically in Figure 1. In this hierarchy, the goal or objective is at the top, criteria are at level 1, sub-criteria are at level-2, alternatives are at the level 3 (lowest level).

Figure 1. Hierarchical structure of AHP, Source (Raut and George 2020).

The whole AHP process involves the following steps:

1. Establishing the hierarchy of criteria and sub-criteria with respect to the goal or objective.

2. Making pair-wise comparison matrices for each criterion or sub-criterion over the others.

3. Assigning the weights to criteria and sub-criteria.

4. Checking the consistency of judgements.

5. Calculating the scores of available alternatives.

6. Ranking and categorising the alternatives.

The first four steps were used here for prioritising criteria based on their obtained weights. However, the scoring, ranking, and classification of alternatives, i.e. communities were performed by separate methods which are beyond the preview of this paper.

If nxn matrix A represents pair-wise comparison of n evaluation criteria, its element a_ij represents the importance of i^th criterion over j^th criterion. The significance of a_ij is illustrated below in the Table 1

Table 1. Representation of the importance of one criterion over the other.

Sl. No.	Value	Significance
1	aij < 1	i^th criterion is less important than the j^th criterion
2	aij > 1	i^th criterion is more important than the j^th criterion
3	aij = 1	i^th criterion is equal to the j^th criterion

The entries a_ij and a_ji satisfy the condition a_ij.a_ji = 1 and for i=j, a_ij = 1. The relative importance in a scale ranging from 1 to 9 with interpretation is shown in Table 2.

Table 2. Interpretation of relative importance one criterion over the other.

Value of aij	Reciprocal value	Definition	Interpretation
1	1	Equal Importance	Two activities contribute equally to the objective
2	1/2	Weak or slightly more importance
3	1/3	Moderately importance	Experience and judgement slightly favour one activity over another
4	1/4	Moderately strong importance
5	1/5	Strong importance	Experience and judgement strongly favour one activity over another
6	1/6	Moderately very strong importance
7	1/7	Very strong importance	An activity is favoured very strongly over another; its dominance demonstrated in practice
8	1/8	Very, very strong importance
9	1/9	Extreme importance	The evidence favouring one activity over another is of the highest possible order of affirmation

After establishment of the matrix A, a normalised pair wise comparison matrix A _norm is derived by normalising each element ${\bar{a}}_{ij}$ by the following computation.

${\bar{a}}_{ij}=\frac{a_{ij}}{{\sum }_{l=1}^n{a}_{ij}}$

The weight vector w is built by averaging all the elements in each row of A _norm as given by following equation.

$w_j=\frac{{\sum }_{l=1}^n{\bar{a}}_{jl}}{n}$

During pair-wise comparison, inconsistency may arise sometimes in decision-maker’s judgement. Therefore to check the consistency of the judgement, consistency ratio (CR) should be less than 0.1 as suggested by Saaty.

$\frac{\mathrm{C}\mathrm{I}}{\mathrm{R}\mathrm{I}}=\mathrm{C}\mathrm{R}<0.1$

RI is the consistency index when the elements of matrix A are taken randomly. The values of RI for n ≤ 10 are shown below in Table 3.

Table 3. Interpretation of relative importance of one criterion over the other.

n	2	3	4	5	6	7	8	9	10
RI	0	.58	.9	1.12	1.24	1.32	1.41	1.45	1.51

Data were collected during the research work conducted at Indian Institute of Technology, Kharagpur in the year 2017. There is no exact number of required samples to carry out AHP survey. For convenience, 37 set of questionnaires consisting of pair-wise comparison questions were prepared and given to experts and users having in-depth knowledge of Age-friendliness. However, 31 responses were received showing a response rate of 83.7%. Data obtained were analysed by AHP software available online in excel format (American Society for Quality 2008). The respective consistency index was then checked. It was found that in 26 cases, the results obtained were consistent as consistency ratios were less than 0.1. A pair-wise comparison matrix A enlisting scores of all possibles comparisons among the criteria as obtained from the responses of the experts is shown below in Table 4. There are 26 such comparison matrices but all can not be shown in this paper because of page constraints.

Table 4. Pair-wise comparison matrix of the main criteria with respect to the Goal.

	Independent Living	Social Inclusion	Well-being	Information and communication Provision	Safety and security	Priority
Independent Living	1	2	5	1	2	.294
Social Inclusion	1/2	1	5	1/3	1/2	.129
Well-being	1/5	1/5	1	1/6	1/7	.038
Information and communication provision	1	3	6	1	1/3	.230
Safety and security	1/2	2	7	3	1	.309

Similarly, the pair-wise comparison matrix for sub-criteria with respect to Independent living criterion is shown in Table 5 below.

Table 5. Pair-wise comparison matrix of sub-criteria with respect to Independent living.

	Housing options	Transportation options	Home modifications	Equipment use	Accessibility	Affordability	Priority
Housing options	1	1/4	1/5	1/2	3	1/3	.076
Transportation options	4	1	1	3	5	1	.262
Home modifications	5	1	1	3	5	1	.274
Equipment use	2	1/3	1/3	1	3	1/3	.106
Accessibility	1/3	1/5	1/5	1/3	1	1/3	.047
Affordability	3	1	1	3	3	1	.235

The weights obtained here are the local weights. Therefore, the global weight of each sub-criterion is obtained by multiplying its local weight with the local weight of its respective criterion. The global weights of all criteria and sub-criteria obtained from the responses of all 26 experts are given in Table 6.

Table 6. Global weights of all criteria and sub-criteria.

Criteria and Sub-criteria	Expert 1	Expert 2	Expert 3…	… Expert 26	Global weights
Independence	.294	.442	.301	.098	0.234
Housing options	.022	.040	.066	.007	0.024
Transportation options	.077	.118	.023	.010	0.026
Home modifications	.081	.177	.044	.008	0.022
Equipment use	.031	.065	.019	.019	0.027
Accessibility	.014	.015	.032	.024	0.037
Affordability	.069	.028	.118	.031	0.042
Social inclusion	.129	.078	.060	.456	0.093
Proximity	.003	.006	.005	.041	0.010
Material well being	.006	.006	.003	.015	0.006
Volunteering	.049	.023	.005	.015	0.006
Valued recognition	.010	.003	.010	.140	0.011
Involvement and engagement	.045	.006	.015	.064	0.017
Family support	.017	.033	.022	.182	0.028
Well-being	.038	.041	.086	.141	0.117
Physical	.010	.003	.013	.016	0.025
Social	.005	.015	.025	.047	0.023
Environmental	.010	.019	.004	.016	0.014
Emotional	.003	.003	.037	.047	0.022
Spiritual	.010	.002	.007	.016	0.009
Information and communication	.230	.103	.168	.050	0.117
Information ground	.077	.074	.010	.030	0.027
Mobile networking	.077	.015	.047	.010	0.038
Internet services	.077	.015	.111	.010	0.029
Safety and security	.309	.336	.384	.254	0.245
Land use safety	.019	.037	.033	.053	0.026
Transportation safety	.067	.037	.096	.027	0.043
Indoor safety	.157	.070	.018	.134	0.062
Outdoor safety	.067	.192	.238	.040	0.070

Criteria and sub-criteria were then arranged in decreasing order of their weights to determine their priorities as shown in Table 7.

Table 7. Priority ranking of criteria and sub-criteria.

Criteria and Sub-criteria	Global weights
Safety and security	0.245
Outdoor safety	0.070
Indoor safety	0.062
Transportation safety	0.043
Land use safety	0.026
Independent living	0.234
Affordability	0.042
Accessibility	0.037
Equipment use	0.027
Transportation options	0.026
Housing options	0.024
Home modifications	0.022
Well being	0.117
Physical	0.025
Social	0.023
Emotional	0.022
Environmental	0.014
Spiritual	0.009
Information and communication provision	0.117
Mobile networking	0.038
Internet services	0.029
Information ground	0.027
Social inclusion	0.093
Family support	0.028
Involvement and engagement	0.017
Valued recognition	0.011
Proximity	0.010
Volunteering	0.006
Material well being	0.006

5. Results and discussion

The priority ranking (Table 7)helps policymakers in determining the priority of action while implementing policies as part of government interventions to make the community more age-friendly. Criteria having higher weights such as ‘safety and security’ and ‘independent living’ ability of people in the community should be considered first on a priority basis for the implementation of relevant and appropriate policies to enhance the age-friendliness of the community. ‘Safety and scurity’ has got the highest priority followed by ‘independent living’, ‘well-being’, ‘information and communication provision’, and ‘social inclusion’. ‘Well-being’ and ‘information and communication provision’ are equally important which signifies that they contribute equally to the age-friendliness of the community. Similarly under each criterion sub-criteria are arranged in decreasing order of their weights to determine their priority. Safety of the outdoor environment is the most important sub-criterion amongst all sub-criteria under ‘safety and security’ criterion. ‘Affordability’ is the most important sub-criterion of ‘independent living’ criterion. ‘Family support’ is the most important sub-criterion of ‘social inclusion’ criterion. ‘Mobile networking’ is the most important sub-criterion of ‘information and communication provision’ criterion.

6. Research findings

The criteria and sub-criteria that contribute maximum to ‘Age-friendliess’, are considered for implementing future policies and guidelines as resources available in developing countries like India are limited. ‘Safety and security of outdoor environment’ is considered in this paper for implementing related policies to improve the safety of the community environment. Friedman’s rank test was applied to rank different policy measures recommended by experts of the community during pilot study, focus group survey, expert opinion survey, and assessment survey. Experts aged 45 and above having in-depth knowledge of Age-friendliness from each of 13 communities of Bhubaneswar Municipal Corporation were asked to rank the different safety and security measures as listed below in Table 8. The age of the experts chosen for this study was 45 as at this age one has several decades of work experience. He or she develops skills of management, leadership, learning ability, and more diverse and professional networks. Usually, younger generations have tendency of switching jobs, however older generations tend to live in the same locality for many years due to which they have profound knowledge about the community they live in (Luke 2021). For this reason, in most of the studies on eldely and ageing-related issues, respondent’s age considered for conducting the survey is 45 and above. In the year 2010, for Longitudinal Ageing Study in India (LASI) pilot survey, 1500 older adults aged 45 and above were interviewed to find out the issues associated with the old age (Agarwal et al. 2016b, 2016b). The sample size for the survey was determined by Cochran’s formula applicable especially for a large population.

Table 8. Ranking of different safety and security measures of the community.

Rank	No. of respondents assigning ranks to different safety and security measures
	Buildings of mixed-use design	Street lighting	CCTV surveillance	Police patrolling	Community boundary with security post
1	66	58	140	62	64
2	50	86	126	68	60
3	80	102	36	104	68
4	110	80	44	76	80
5	84	64	44	80	118
Sum of ranks	1266	1176	896	1214	1298
Mean rank	3.246	3.015	2.297	3.112	3.328

If n is the sample size, i.e, the number of respondents then n =$\frac{z^{2}pq}{e^2}$ (Bartlett-II et al. 2001)

e is the desired level of precision

p is the estimated proportion choosing the attribute

q = 1-p

For 95% confidence level, z = 1.96 and for maximum variability p = 0.5 and q = 0.5. If level of precision is 5%, then n = 385.

Since there are 13 communities, the number of samples from each community taken for the study = 385/13 = 29.7 ≈ 30.

Hence, 30 respondents from each of 13 communities summing a total of 390 were chosen instead of taking 385 as determined by Cochran’s formula.

To implement different policy measures for improving safety of the community, a rating analysis was done using Friedman’s rank test. The null hypothesis chosen for this study is

H₀:

There are no significant differences in the median ratings for different safety and security measures to improve the safety of the community environment.

Friedman rank test (Ramachandran and Tsokos 2020) formula that was used in the study is

${\mathrm{F}}_{\mathrm{r}}=\frac{12}{\mathrm{r}\mathrm{c}\left(\mathrm{c}+1\right)}{\sum }_{j=1}^c\mathrm{R}{\mathrm{j}}^2-3\mathrm{r}\left(\mathrm{c}+1\right)$

r= number of respondents, c = number of groupsRj = sum of ranks of J^th group

R₁ = 1266, R₂ = 1176, R₃ = 896, R₄ = 1214,R₅ = 1298

R₁+ R₂+ R₃+ R₄+ R₅ = 5850

$\frac{\mathrm{r}\mathrm{c}\left(\mathrm{c}+1\right)}{2}=\frac{390\ast 5\left(5+1\right)}{2}=\hspace{0.17em}5850$

$\mathrm{S}\mathrm{o},{\mathrm{R}}_1+{\mathrm{R}}_2+{\mathrm{R}}_3+{\mathrm{R}}_4+{\mathrm{R}}_5=\frac{\mathrm{r}\mathrm{c}\left(\mathrm{c}+1\right)}{2}$

$\begin{array}{rl}& {\mathrm{F}}_{\mathrm{r}}=\frac{12}{\mathrm{r}\mathrm{c}\left(\mathrm{c}+1\right)}{\sum }_{j=1}^c\mathrm{R}{\mathrm{j}}^2-3\mathrm{r}\left(\mathrm{c}+1\right)\\ & =\frac{12}{390\ast 5\left(5+1\right)}\left[{{\mathrm{R}}_1}^2+{{\mathrm{R}}_2}^2+{{\mathrm{R}}_3}^2+{{\mathrm{R}}_4}^2+{{\mathrm{R}}_5}^2\right]-3\mathrm{r}\left(\mathrm{c}+1\right)\\ & =\frac{12}{390\ast 5\left(5+1\right)}\left[{1266}^2+{1176}^2+{896}^2+{1214}^2+{1298}^2\right]\\ & -3\ast 390\left(5+1\right)\\ & =7125.28-7020=105.28\end{array}$

The upper tail critical value for chi-square distribution having 5–1 = 4 degrees of freedom and 0.05 level of significance is 9.488

Here F_r >9.488

So, we reject the null hypothesis at α = 0.05level of significance. We conclude that there are significant differences in the median ratings of five different safety and security measures for a community environment. From the mean ranks of different safety and security measures given in Table 8, we have concluded that community boundaries with security post has got the highest priority for providing safety to the community followed by buildings of mixed use design, police patrolling, street lighting, and CCTV surveillance (Farsani et al. 2019).

CCTV surveillance has the lowest priority according to the obtained result, as there are chances of theft and robbery in spite of presence of CCTV camera. This is because CCTV camera sometimes fails to detect the stealing due to the intelligence of the thief.

Due to the presence of boundary wall with security post Arya village is the safest community among all the communities surveyed. Elderly feel safe and secure while moving in the community according to the respondent’s view. This finding is consistent with the result of Friedman’s rank test supported by Figure 2. During survey, it was observed that most of the plots have their own existing boundaries. Therefore, it would be easy to have a complete boundary around a community by filling the existing gap between two adjacent plots with construction of masonry wall and providing minimum entry and exit points. Deployment of security personnel at these entry and exit points would add safety exponentially to the community.

Figure 2. Security post, boundary wall, and CCTV of Arya Village.

Source: Author.

Similarly, buildings allowing mixed-use activity are preferred over CCTV surveillance according to the results of Table 8 Therefore, Chandrasekharpur community is one of the safest communities as chances of theft and robbery are less due to the occurrence of the mixed-use activity. This is supported by Figure 3. On the other hand, in Sri Vihar community, buildings are not designed to facilitate mixed-use activity. Although CCTV installations are there in most of the newly built apartments and in public places, it faces frequent attacks by thieves and burglars. This is supported by Figure 4. In recent years, emphasis is given to incorporating mixed land use in building byelaws. Apart from providing safety and security, it fulfils changing lifestyle needs of the elderly by ensuring required products and services available in close proximity (The times of India 2019). The mixed-use activity of Chandrasekharpur community is shown in Figure 4. Due to availability of social and physical infrastructures such as primary school, community centre, open space, and healthcare centre in close proximity within community itself, residents are able to maintain their independent lives by availing these services. This is supported by Figures 5 and 6.

Figure 3. CCTV of Sri Vihar.

Source: Author.

Figure 4. Mixed-use activity, Chandrasekharpur.

Source: Author.

Figure 5. Primary school, community centre and open space of Chandrasekharpur.

Source: Author.

Figure 6. Healthcare centre of Chandrasekharpur.

Source: Author.

Police patrolling system of a city also plays important role in preventing or deterring crime and theft by responding to incidents (Zhang and Brown 2013). Regular police patrolling prevents crime and theft as it creates fear in the mind of thieves according to the respondents view. Appropriate patrolling policies such as police district design (Zhang and Brown 2013) or desired technological advancement by use of innovative methods such as network analysis, crime mapping and GIS (Fatih and Bekir 2015), use of various apps and community policing (Delhi Police 2020) are important methods to improve safety of a community.

CCTV installation at different locations in a community has the lowest priority according to the result of the analysis shown in Table 8. This is because it has lot of disadvantages such as high cost, inefficient technology, and intelligence of the criminals to destroy it (Urban company 2017). Therefore, safety of the community by means of CCTV installations could be improved if these drawbacks are taken into consideration and appropriate steps are taken.

As observed from Table 7, affordability contributes maximum to making an individual independent. Independent living ability has a positive relationship with health and well being of older adults (Illario et al. 2016). Affordability of The policy measures to increase affordability of an individual are allowing discount on price of products and services, increasing job opportunities and imparting vocational training. All these measures should be made available to the target beneficiaries such as the elderly, disabled and economically backward as stated by the respondents during survey. However, all other sub-criteria such as ‘Accessibility’, ‘Equipment use’, ‘Transportation option’, ‘Housing option’ are also important to enhance one’s ‘Independent living’ ability as illustrated in results and discussion section.

Physical health of individuals in the community is the prime determinant of one’s well-being. Ferdman states that walkability contributes to one’s objective well-being. It enables opportunities to improve human capacities in various ways such as development of knowledge, creativity and social capacity and develop opportunities enabling driving or commuting (Ferdman 2019). Neighbourhood design affects subjective well-being of residents such as happiness as stated by Pfeiffer and Cloutier. Allowing resident to access open, natural and green spaces may directly increase their happiness. Also designing neighbourhood that creates opportunities for social interaction and safety may foster happiness among its residents (Pfeiffer and Cloutier 2016). Compact city form has positive influence on resident’s subjective well-being such as resident’s personal relationships and health as found out by Mouratidis in a study conducted in Oslo metropolitan area. On the other hand, it has negative influence on resident’s emotional response such as feeling of fear and crime (Mouratidis 2019). Mouratidis in another study concluded that urban planning strategies such as providing access to public spaces, open spaces and nature and communal spaces in residential complexes improve resident’s subjective well-being (Mouratidis 2021). In Chandrasekharpur community physical and social infrastructures such as primary school, open space and healthcare facilities are made available in close vicinity inside the community through municipal government’s intervention during physical planning of the community. Therefore, this type of urban planning has positive influence on resident’s satisfaction in terms of perceived absence of fear and crime, access to services, open spaces and nature.

7. Conclusion

Analytical hierarchy process or Multi Criteria Decision Making Analysis (MCDA) is found suitable for this study since it involves prioritisation of both tangible and intangible criteria based on their weights. The weights of criteria and sub-criteria were the result of their pairwise comparison made by experts during the survey. Unlike other methods such as principal component analysis, neural network etc. AHP is more of deterministic in nature, which means experts find it easier to assign relative importance of a criterion over the other. This prioritisation would help the policymakers to implement policies through governmental interventions. ‘Safety and security’ has got the highest rank followed by ‘independent living’, ‘well-being’, ‘information and communication provision’ and ‘social inclusion’. Similarly, ‘affordability’ that could make one independent has got the highest rank. Affordability of individuals could be increased by different policy measures such as the discounted price of products and services, increased income, and imparting vocational training. Physical well-being has got the highest rank as it contributes maximum to enhancing well-being of the elderly. By remaining physically fit, the elderly could actively participate in social and community life that would enhance age-friendliness of the community. The significance of differences in the mean ranks of different policy measures plays an important role in justifying the benefits of a specific policy in terms of improving the age-friendliness of the community if implemented. Friedman’s rank test could also be applied to other policy measures relating to a particular age-friendly criterion or sub-criterion when it would be a subject of concern. A community boundary with security check post adds maximum security to the community in accordance of the result of Friedman’s rank test. A combination of policy measures having higher priority such as ‘police patrolling’, ‘mixed land-use urban planning’, ‘accessibility to services’ would be the effective solution to increase age-friendliness of the community.

Acknowledgments

Indian Institute of Technology, Kharagpur funded the research work for award of PhD degree to the Author. Therefore, the Author would like to thank the Institute for financial support.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Sashi Bhusan Raut

Sashi Bhusan Raut is an Associate Professor in Piloo Mody College of Architecture at Biju Patnaik University of Technology, Rourkela, Odisha, India. He completed his Ph.D. from the Indian Institute of Technology, Kharagpur, West Bengal, India with a thesis on ‘ Age-friendliness Assessment Model, A Delphi-AHP-Fuzzy Logic-Approach’. Dr. Raut’s main scholarly interest is on the sustainability of cities in developing countries like India and similar environmental settings. His works mainly focus on Age-friendly design, Universal design, and the production of sustainable building materials and products that make the community livable and sustainable. He also works on formulating policies and guidelines for a sustainable community.