Quality of life in Saskatoon: Achieving a healthy, sustainable community.

Williams, Allison M. ; Randall, James E. ; Holden, Bill 等

Resume

Ce papier compare la qualite de vie (QDV) intra-urbaine de trois quartiers de Saskatoon, Saskatchewan. L'objectif de l'etude etait de determiner dans quelle mesure les caracteristiques des differents quartiers ou lieux influencent la qualite de vie de personnes ayant un statut socio-economique (SSE) semblable. Cette recherche s'insere dans un plus grand projet qui, a l'aide d'une approche multi-sectorielle, developpe et utilise les indicateurs de QDV contribuant a une communaute viable et en meilleure sante. Les resultats inclus dans ce papier proviennent d'une enquete de 900 foyers qui demandait aux participants d'evaluer un ensemble de domaines lies a la QDV. Aucune difference statistique n'a ete trouvee entre les habitants de ces trois quartiers au niveau de la perception de leur propre QDV, stress personnel, ou changement au niveau de la satisfaction globale de leur vie malgre des differences significatives au niveau de l'education et du revenu. Pourtant, quand des variables composites ont ete derivees po ur indiquer le niveau de cohesion sociale, la satisfaction avec les relations ainsi que celle avec les habitants de son propre quartier, des differences significatives intra-urbaines ont emerge. En outre, des differences extremement significatives au niveau d'un certain nombre de variables ont ete trouves entre les endroits au SSE le plus bas compares aux endroits avec le SSE moyen ou le plus haut. Ceci suggere que l'ecart croissant entre les riches et les pauvres au Canada se reflete dans la perception de la qualite de vie.

Mots-cles: Qualite de Vie; Indicateurs Sociaux; Saskatoon.

Key words: Quality of Life; Social Indicators; Saskatoon.

Abstract

This paper undertakes an intra-urban comparison of quality of life (QOL) for three sets of neighbourhoods in Saskatoon, Saskatchewan. The objective of the paper is to determine how the characteristics of different neighbourhoods or locales influence QOL for persons of similar socioeconomic status (SES). It is part of a larger project that uses a multi-stakeholder approach to develop and use QOL indicators to achieve a healthier, more sustainable community. The results reported in this paper are based on a comprehensive survey of over 900 households, and asked respondents for their subjective assessment of a set of domains associated with QOL. It was found that there was not a statistically significant difference among residents in the three sets of neighbourhoods in their perception of personal QOL, personal stress, or change in overall life satisfaction, even though education and income were significantly different. However, when composite variables were derived as indicators of level of social cohesion; sat isfaction with relationships and resident's own neighbourhoods, there were significant intra-urban differences. In addition, there were highly significant differences across a wide range of variables between the Low SES areas on the one hand, and the High SES and Medium SES neighbourhoods on the other. This suggests that the growing gap between rich and poor in Canada is being reflected in perception of quality of life.

Introduction

There is a growing interest among academics, the social policy community and governments -- in both Canada and abroad -- to monitor social progress. As a consequence, much activity is focussed on measuring quality of life, often via the development and implementation of social indicators that go beyond the measurement of inflation, interest rates or the gross domestic product (Canadian Council on Social Development. 1996). A recent review of indicator development concluded that there is little likelihood of quality of life indicators being adopted or influential in policy decision-making unless the process has policy relevancy, an interested constituency and, most importantly, stakeholder engagement in indicator development and selection (Hancock et al. 2000). This is in keeping with Friedman (1997), who proposes a holistic scientific strategy for improving the quality of life. Such an integrated approach to quality of life research is crucial to furthering awareness about the impact of social, economic and p olitical decisions and activities, while operating as a guide to decision-makers in both the private and public sectors (Clark 2000). Given that neighbourhood inequality in Canadian cities has been growing since 1980 (Statistics Canada 2000), together with the growth in the number of high-poverty neighbourhoods in Canadian cities between 1980 and 1995 (Lee 2000), the need for policy-relevant intra-city research on quality of life has never been stronger.

The research presented here is part of a larger project that examines the process and results of a multi-stakeholder approach to the development and use of quality of life indicators in achieving a healthy, sustainable community in Saskatoon. Saskatoon is a medium-sized city of 200,000, in the prairie province of Saskatchewan. Both municipal government and community-based organizations (CBOs), together with university-based academics, have shown readiness to address problematic areas and issues in partnership with each other. The envisioned outcome of this larger project is the ongoing sustainability of Saskatoon as a healthy city with an improving and a more equitably distributed quality of life. While determining residents' quality of life, the research examines how evaluations of quality of life differ across three locales (each made up of a number of census tract clusters) representing low, medium and high socioeconomic status (SES). One of the research questions posed by this study is: how do differing n eighbourhood characteristics (locales) influence quality of life for persons in similar SES circumstances? Answering this question not only enriches a general understanding of quality of life outcomes for persons based on their socio-economic status, but also improves the ability of local citizens, organizations and government to target policy and program investments more effectively. The paper begins by highlighting, via a review of the quality of life literature, the need for comparative intra-urban quality of life research. After describing the research methodology, the results of intra-urban comparisons are presented. The paper concludes with a discussion of the research implications.

Quality of Life Research in Geography

Although Myers (1988) has defined four approaches to quality of life analysis, there appear to be two main purposes in carrying out quality of life research: (1) to rate places according to their livability or attractiveness (Nissan 1989, Boyer and Savagneau 1985 and 1989, Pierce 1984, Finlay et al. 1988a and b, Bowman et al. 1981, Conway and Liston 1981, Marlin and Avery 1983, Liu 1976); and (2) to analyse social phenomena in space, for the purpose of "gaining a better understanding of factors which improve or decrease the quality of life and living environments" (Canada Mortgage and Housing Corporation 1991: 15) (Dilks 1996, Vogel 1997, Myers 1987 and 1988, Pacione 1980 and 1986, Helburn 1982, Smith 1973, Bederman 1974, Dickinson et al. 1972). The latter purpose, often called "welfare geography" (Johnston et al. 1986: 437), has a "... basis of concern [with] areal differentiation and regional identification of well-being across space" and has been argued as the more humane of the two types of quality of lif e research (Rogerson 1999). Such research uses an applied geographical framework, where the "conditions at the disadvantaged or deprived end of the quality of life spectrum form a key area of study in contemporary human geography" (Pacione 1986: 1499). Such research has the ability to incorporate a temporal component, but rarely does; the same place(s) can be evaluated over time to describe long-term trends and fluctuations in quality of life. Such information enables decision-makers to better direct positive change through the manipulation and evaluation of plans and programs, and through the guided creation of policy (Myers 1987, Smith 1973).

Territorial indicators, such as the living environment, social order and social belonging (Smith 1973, Schulman and Bond 1978), continue to be popular among geographers as they exemplify a place-based perspective. In his study of quality of life indicators in American metropolitan areas, Liu (1976) used discipline-specific indicators in order to keep intact each discipline's "understanding of how values and ideas should be defined and quantified" (1976: 12). Reid's (1991) study of urban Canada measured quality of life across 12 dimensions defined on the basis of subjective assessments of resident satisfaction. A number of other geographers have studied the perceptual indicators of places (Saarinen 1976, Downs and Stea 1977, Saarinen and Sell 1980, Saarinen 1981). The livability approach is a recent arrival in quality of life research. It, too, is a multi-disciplinary concept, which has a specific planning objective: environmentally sustainable development (Metro Toronto 1991, Gariepy et al. 1990).

Outside of geography, various types of social scientists have contributed to a growing literature on community-level indicators that address the quality of life of communities. Perceived neighbourhood quality is the focus of many studies (Connerly and Marans 1985, Furuseth and Walcott 1990, Olsen et al. 1985, Walter-Busch 1983, Moller and Schlemmer 1983, Mastekaasa and Moun 1984, Tahlin 1990), while others analyse relationships among quality of life and system-level indicators, such as the Children's Stress Index (Schwirian et al. 1995). The Canadian Policy Research Network's "Quality of Life Indicators Project" (2000) has begun to develop a set of national indicators (through a citizenship engagement process) to track Canada's progress in quality of life. Urban planners such as Myers (1987 and 1988) have introduced a community trends approach that focuses on quality of life components and trends within local communities. Although much comparative quality of life work has been conducted (Szalai and Andrews 1 980), few studies of differences in the quality of life of urban sub-groups have been conducted (e.g. Pacione 1980), and little intra-city comparative work has been done. Turksever and Atalik (2000) examine the regional variations in quality of life, based on subjective indicators collected in Instanbul, Turkey, arguing for the importance of differentiating and comparing quality of life across intra-city regions. Examining the quality of life in local metropolitan labour markets in three Italian cities determined the importance of socio-economic differentiations both between and within the cities (Baldazzi et al. 1998). In the case of Saskatoon, two pieces of research have adopted an intra-city approach to describe quality of life. Morton (1999) analysed the differences in the perception of quality of life across groups of census tracts, defined on the basis of a set of variables derived from the Population Census and other secondary sources. Randall (1999) provided a simple intra-urban comparison of quality of life disaggregated by economic, social and physical dimensions, as part of the Atlas of Saskatchewan. Intra-city comparative work is becoming increasingly important given the growing divide between the rich and the poor in Canadian cities (Statistics Canada 2000).

Method and Data

The participatory action research approach (de Koning et al. 1996) reported in this paper ensures the value of the outputs to the stakeholders and the likelihood of their using this research to change policy and programs. In addition to having community representatives working together on the project, a number of forums have been held with the community to both gather feedback on the research process and data-gathering instruments, and to provide a mechanism for community members to contribute ideas and suggestions.

Existing knowledge on the views of dozens of community service organizations concerning factors that enhance and detract from citizen's quality of life were first collected and reviewed using various types of documents produced by these organizations. After the survey draft was critically discussed by community members at the first community forum, it was revised and piloted. The survey instrument asked for subjective assessments of numerous domains, including: personal quality of life, community quality of life, government activities, and government spending and demographics. The telephone survey was conducted in partnership with the city's local newspaper. The Saskatoon StarPhoenix collected the data and soon after, published the highlights. Preliminary analysis of the survey data was presented to the community via a second forum. The sample frame for the telephone survey was chosen to represent the Saskatoon population, as is discussed below.

Sample Frame:

The first step in determining a sample representative of Saskatoon's population was to group neighbourhoods according to socio-economic indicators known to have an impact on quality of life. Selected neighbourhood demographics from the 1996 census were analysed using the Statistical Package for the Social Sciences (SPSS) software program. The variables selected were: percentage of the neighbourhood population that was Aboriginal, median household income, percentage of households that were lone parent families, percentage of housing that was owned, and percentage of the labour force employed. Once standardised, the scores of these variables were submitted to the K-Means Cluster routine. A three-cluster solution was specified to facilitate interpretation of the groups. Summary statistics for each group's component neighbourhoods were obtained with the SPSS report routine and the neighbourhoods were mapped according to group membership.

Summary statistics for the five component variables are shown in Table 1. Figure 1 shows the results of the Cluster technique. These group Saskatoon neighbourhoods according to the typical socio-economic scale from highest -- High socio-economic status (SES) to lowest -- Low socio-economic status (SES). There are 56 neighbourhoods in total within the Saskatoon city limits, with 27 categorised as High SES, 20 categorised as Medium SES and nine categorised as Low SES.

The High SES cluster includes the highest median income neighbourhoods in the city. This group has the highest rate of employment. When combined with the low score for lone-parent families, this suggests that two-income families characterise these neighbourhoods. As shown in Table 2, home ownership is about 15 percentage points above the average for all city neighbourhoods.

The Medium SES cluster represents a diverse middle ground in the socioeconomic and demographic makeup of Saskatoon neighbourhoods. As shown in Table 3, these neighbourhoods have moderate incomes and lower rates of ownership and employment. This reflects the greater presence of multiple unit dwellings and the high rates of single person households in these neighbourhoods.

The Low SES cluster identifies the most disadvantaged neighbourhoods in the city. As shown in Table 4, the group has the lowest income, highest representation of Aboriginals, lowest rate of home ownership, lowest rate of employment, and highest rate of lone-parent households.

The cluster analysis identified neighbourhoods that shared socio-economic characteristics but represented relatively diverse physical and development characteristics. For the purpose of representing the effect of place on quality of life, the sample was focussed on contiguous neighbourhoods within the three clusters. The neighbourhoods from which the survey sample was drawn are represented in Figure 1 by an overlay of different patterns.

The Planning Department of the City of Saskatoon used its Geographic Information System to develop the base for the survey sample. For each cluster, the neighbourhood boundaries of each selected contiguous neighbourhood were used to access Statistics Canada's Postal Code Conversion file. The result was a database of all postal codes in the selected neighbourhoods for each group. The Saskatoon StarPhoenix newspaper then matched this data with a telephone database by postal code. Given that not all residents of the neighbourhoods had a phone, the database biased those who owned telephones. The result was a telephone database for each neighbourhood group. Students were employed to carry out the telephone interviews, attending a training workshop beforehand.

Out of 4,469 households called, 968 responded, giving a response rate of 21.7 per cent. (1) For the multivariate portion of the statistical analysis, 917 complete responses were used, including 303 from the High SES cluster, 332 from the Medium SES cluster, and 282 from the Low SES cluster. The telephone survey was carried out from December 14, 2000 to January 08, 2001 inclusive, except on Christmas Eve, Christmas Day, Boxing Day, New Year's Eve, New Years and January 7, 2001.

Results and Interpretation

In order to describe how differing neighbourhood characteristics influenced quality of life for persons in similar SES circumstances, frequencies were first examined for the following variables: overall quality of life, perception of happiness, change in life satisfaction, satisfaction with own neighbourhood, and satisfaction with city. Tables 5a through 5c present an initial description of the attitudes of the respondents towards their own quality of life in each of the three clusters of neighbourhoods.

At this broad, descriptive level, it can be seen that respondents in the High and Medium SES neighbourhoods have relatively similar perceptions of their own quality of life. Between 66.9 and 70.9 per cent of respondents in the High and Medium SES neighbourhoods described their overall QOL as either excellent or very good, while only 48.7 per cent of respondents in the Low SES neighbourhoods chose these categories. Applying a Chi Square ([chi square]) statistical test to this contingency table, it was found that there are statistically significant differences in the responses across the three neighbourhoods. Although not statistically significant, broader neighbourhood distinctions between the Low SES households and the other two groups are apparent when respondents are asked to describe their feelings of happiness (Table 5b). Here, only 65.6 per cent of the Low SES neighbourhood respondents described themselves as being 'Happy and Interested in Life', compared to 73.5 per cent in the High SES neighbourhoods a nd 73.4 per cent in the Medium SES neighbourhoods. Not all general questions elicited the same distinctiveness across the three clusters of neighbourhoods. When asked about change in personal life satisfaction over the past three years, almost the same proportions of respondents across all three clusters of neighbourhoods indicated that their life satisfaction had improved (49.9, 48.0 and 48.0 per cent), or stayed the same (43.8, 44.9, 41.6 per cent), from High to Low SES respectively. A much higher proportion (10.4%) of those in the Low SES neighbourhoods described their lives as having worsened, compared to 6.3 per cent in the High SES and 7.1 per cent in the Medium SES neighbourhoods.

A comparison of Tables 6a and 6b suggest that much of the degree of satisfaction or dissatisfaction rests not at the level of the city, but rather within the neighbourhood or locale. Although variation across neighbourhoods is statistically significant in both "Satisfaction with Neighbourhood" (Table 6a) and "Satisfaction with City" (Table 6b), there is a much stronger relationship in the former ([chi square] = 134.12) than with the latter ([chi square] = 13.1). Comparing across these two contingency tables, respondents within the High and Medium SES neighbourhoods made virtually no distinction in their degree of satisfaction with their city and neighbourhood. However, respondents within the Low SES neighbourhoods expressed a much lower level of satisfaction with their own neighbourhood than they did with the city as a whole. Although these raw frequency distributions do not yet indicate the underlying factors behind these differences, they do begin to paint a picture of significant place-based differences in attitudes and perceptions, not only in how the residents feel about themselves, but also about the urban places in which they live.

In order to further examine the distinctiveness of the clusters of neighbourhoods, Spearman's Rank Order Correlation Coefficients were calculated between the neighbourhood designation and a wide array of socioeconomic and attitudinal variables drawn from the survey. As was expected, given the sample frame methodology that was used to group urban areas, Table 7 shows that the three clusters of neighbourhoods are distinguished on the basis of employment status, education, income, and income adequacy (2) (significant at [alpha] = .05, two-tailed). An additional set of composite variables was derived from a principal components analysis in an attempt to reflect statistically the level of social cohesion and "satisfaction" respondents have with various aspects of their lives and neighbourhoods. The composite variables or factors were derived initially from a principal components analysis on the 38 survey questions hypothesised to be related to quality of life, where each of the individual questions listed below l oaded highly on these composite variables. Responses to the individual questions were then summed to create a set of summary variables as follows: Social Cohesion (feeling part of the neighbourhood, comfort in participating in neighbourhood projects, calling on neighbours in a crisis, and volunteering for organizations), Satisfaction 1-External Structures (level of satisfaction with neighbourhood, housing, amount of money they have, the city, personal health and leisure activities), Satisfaction 3-Personal Relationships (level of satisfaction with spouse or partner, with rest of family, and with friends), Neighbourhood-Perceptual (attitudes toward the neatness of the neighbourhood, feelings of safety from both violent and property crime in the neighbourhood, friendliness of the neighbourhood and involvement in neighbourhood organizations), Neighbourhood-Social Programs (attitudes regarding social programs, care-giver services, recreation programs, neighbourhood health services and protection services), Neighb ourhood-Amenities (attitudes regarding shops and services in the neighbourhood, religious and spiritual activities, schools, and the condition of public transportation), and Neighbourhood-Physical (attitudes towards the condition of neighbourhood roads, green space, parks, housing, environment and traffic conditions). Table 7 shows all of the significantly correlated derived variables with neighbourhood designation, as well as several additional variables that do not show a statistically significant correlation with neighbourhood cluster. This suggests that respondents within the three clusters of neighbourhoods do have very significant differences in the way they view their own personal and public relationships, their level of social cohesion, and their perception of their own neighbourhoods, including physical and social amenities, social and public programs, safety and crime. (3)

It is perhaps surprising that respondents across the three clusters of neighbourhoods cannot be distinguished on the basis of perception of stress in the lives of respondents ([r.sub.s] = 0.028), feelings about their own personal quality of life ([r.sub.s] 0.018), satisfaction with personal relations ([r.sub.s] = -0.055), change in health ([r.sub.s] = 0.012), or change in overall life satisfaction ([r.sub.s] = 0.030) What this table of correlations does suggest is that there are fairly distinct differences among neighbourhoods both on the basis of the traditional socioeconomic measures that are used in urban household and neighbourhood-based research (not surprising, given the sample frame methodology), as well as with the composite transformed variables of social cohesion, and satisfaction with personal relationships, satisfaction with external structures, and lastly various neighbourhood constructs. However, when asked specifically about their own quality of life and related dimensions (stress, change in h ealth and life satisfaction) there are fewer apparent differences in how respondents across these neighbourhoods perceive themselves. Although there appears to be an inconsistency between this result and the neighbourhood distinctions revealed in Tables 5a and 5b, it is the similarity between the Medium and High SES neighbourhood clusters that is contributing to lower correlation coefficients than might be expected. In other words, most of the distinctiveness in perception of overall quality of life is between the Low SES neighbourhoods on the one hand, and the two other clusters of neighbourhoods on the other. This is confirmed further in the strong correlations found among the composite variables in the Low SES neighbourhoods.

In order to disentangle some of the neighbourhood-level distinctiveness suggested by the correlation coefficients in Table 7, the derived variable with the strongest correlation (Neighbourhood-Perceptual; [r.sub.s] = -0.272) is examined in more detail in Tables 8a to 8e. In these tables, responses to the individual questions that comprise this derived variable are provided for each of the three neighbourhood clusters and for the sample as a whole. Given this strong correlation coefficient, it is not surprising to see these tables showing statistically significant differences across all three sets of neighbourhoods in the attitudes towards various issues. In every table, the calculated [chi square] value is significant where [alpha] is lower than 0.001. The Low SES neighbourhoods are different from the other two areas in every one of these questions. In all cases, residents in these neighbourhoods are much more likely to assess services or conditions as poor. The greatest variations are in the degree of safety f rom both violent (Table 8c) and property crime (Table 8d). In the former question, 22.5 per cent of respondents in the Low SES neighbourhoods rated safety from violent crime as poor, compared to 3.2 and 2.4 per cent in the other two neighbourhoods. While 7.3 and 7.6 per cent of residents in the High and Medium SES neighbourhoods rated their degree of safety from property crime as poor, this is much lower than the 29.3 per cent for this category in the Low SES neighbourhoods.

Table 9 shows the correlations among sets of the composite variables (as described above) across the three clusters of neighbourhoods. What is especially apparent is that the strongest correlations among these composite variables are in the Low SES neighbourhoods. This implies that, even though there may be no significant neighbourhood differences in perception of QOL, as suggested above, there are significant differences among the three neighbourhoods in the interaction between personal satisfaction, i.e., through Social Cohesion, Satisfaction1, Satisfaction2, and Satisfaction3, and perception of place or neighbourhood, i.e., through the composite variables Neighbourhood - Perceptual/Social Programs/Amenities/ Physical. It should be noted that, although many of these correlation coefficients seem weak, they are statistically significant at least partly because of the large sample size.

One example from Table 9 illustrates the distinctiveness in the association between attitudes about place and personal quality of life variables across the three clusters of neighbourhoods. In all three sets of neighbourhoods, there is a statistically significant correlation between level of Social Cohesion (SC) and Neighbourhood-Perceptual (Nperc). However, the strength of those associations ranged from a high of [r.sub.s] = 0.352 in the Low SES neighbourhoods to a low of [r.sub.s] = 0.213 in the High SES neighbourhoods. Although both of these values are significant at [alpha] = 0.01, this major difference in the strength of the correlation coefficient suggests that residents in the Low SES neighbourhoods are much more consistent, both positively and negatively, in evaluation of their Social Cohesion and their attitudes towards the neighbourhoods in which they live. Examining Table 9 as a whole, residents in the High and Medium SES neighbourhoods are much less likely to be statistically coherent or consiste nt in comparing various aspects of their quality of life when compared to the Low SES neighbourhoods. These neighbourhood-level differences are consistent across virtually every analysis conducted on this data set including analyses of variance. In almost every case, residents of the Low SES neighbourhoods showed unique attitudes and perceptions towards their own quality of life and neighbourhoods that were much more consistent within the group, and set themselves apart from the Medium and High SES groups.

Conclusion

The research results illustrate what has become known as the growing gap between the rich and the poor in Canada, shown in the increasing gap between the low-income and high-income neighbourhoods (Statistics Canada, 2000). The clear discrepancy found between the Low SES neighbourhoods and the other two neighbourhood clusters suggest the clear focus needed on implementing appropriate measures for improved quality of life in the Low SES neighbourhoods. A recent (October 2001) community forum, held with more than 100 community members -- including policy-makers and government officials -- refined an Action Plan for Saskatoon based on the research results. Many of the strategies suggested in the Action Plan are focused on the Low SES neighbourhoods. One such example is the development of an intersectoral coalition (CBOs, university, health agencies, business and churches) to lobby and educate government on poverty issues. The targeted outcomes of such efforts include an increased minimum wage and higher quality social welfare programs. All involved are committed to bringing about an enhanced and more equitably distributed quality of life in Saskatoon and to monitor the changes taking place overtime. This last objective will be determined by undertaking a second survey in 2003.

In addition to contributing to the enhancement of Saskatoon's community quality of life, this work informs the research emerging in population health, quality of life, and urban social geography, while providing credibility to social reporting and assisting in the evaluation of program impacts and results. As noted by Dissart and Deller (2000: 144), "... quality of life plays and will increasingly play a significant role in the various dimensions of places ...". It is hoped that this study will prompt numerous other intra-city quality of life studies. Furthermore, discussion of the research process itself will contribute to the development and refinement of culturally and socially appropriate community-based research techniques.

Acknowledgements

Funded in part by the Community-University Institute for Social Research (CUISR) and the Social Sciences and Humanities Research Council of Canada under a Community University Research Alliance grant. We are grateful for the comments by two anonymous reviewers.

Table 1

Summary Statistics for All Neighbourhoods (N = 56)

All Neighbourhoods Range Minimum Maximum Mean Std. Dev

Aboriginal (%) 44.37 0 44.4 8.2 9.64
Median Income ($) 70,522 14,390 84,912 38,954 15,531
Lone Parent (%) 37.7 0 37.7 10.5 6.42
Housing Owned (%) 90.9 9.1 100.0 62.7 21.47
Employed (%) 85.6 11.5 97.1 67.9 15.66
Table 2

Summary Statistics for High SES Cluster Neighbourhoods (N=27)

High SES Cluster
Neighbourhoods Mean Minimum Maximum Std. Dev Kurtosis

Aboriginal (%) 4.32 0 16.4 4.4 1.98
Median Income ($) 51,068 33.785 84,912 12,496 .74
Lone Parent (%) 9.4 0 18.9 4.25 .04
Housing Owned (%) 77.6 56.5 100 13.66 -1.06
Employed (%) 77.9 56.2 97.1 8.03 1.38
Table 3

Summary Statistics for Medium SES Cluster Neighbourhoods (N=20)

High SES Cluster
Neighbourhoods Mean Minimum Maximum Std. Dev Kurtosis

Aboriginal (%) 5.5 0 13.9 4.01 -0.30
Median Income ($) 29,792 17,432 41,540 6,734 -0.92
Lone Parent (%) 7.8 0 16.7 4.53 -0.37
Housing Owned (%) 51.3 12.2 85.7 17.56 0.76
Employed (%) 58.8 11.5 75.9 17.65 1.97
Table 4

Summary Statistics for Low SES Cluster Neighbourhoods (N = 9)

High SES Cluster
Neighbourhoods Mean Minimum Maximum Std.Dev Kurtosis

Aboriginal (%) 25.9 15.7 44.4 10.71 -1.06
Median Income ($) 22,970 14,390 35,290 7,120.00 -0.62
Lone Parent (%) 19.9 11.3 37.7 7.36 5.06
Housing Owned (%) 43.4 9.1 63.1 18.38 -0.23
Employed (%) 58.6 41.0 71.3 9.75 0.12
Table 5a

Overall Quality of Life in Saskatoon Neighbourhoods

Neighbourhood Percent responses by Category (N=950)
Cluster Excellent Very Good Good Fair

High SES 23.1 47.8 23.1 5.1
Medium SES 25.6 41.3 27.2 5.6
Low SES 16.2 32.5 37.9 10.2

TOTAL 21.7 40.5 29.4 6.9

Neighbourhood Percent responses by
 Category (N=950)
Cluster Poor Total

High SES 0.9 100.0
Medium SES 0.3 100.0
Low SES 3.2 100.0

TOTAL 1.5 100.0

[chi square] = 45.49

[alpha] = 0.001
Table 5b

Perception of Happiness in Saskatoon Neighbourhoods

Neighbourhood Percent responses by Category (N-955)
Cluster Happy/ Somewhat Somewhat
 Interested Happy Unhappy
 In Life

High SES 73.5 23.0 3.2
Medium SES 73.4 23.2 2.5
Low SES 65.6 29.2 2.9

TOTAL 71.0 25.1 2.8

Neighbourhood Percent responses by Category (N-955)
Cluster Unhappy/ Life not Total
 Little Interest Worth-
 in Life while

High SES 0.0 0.3 100.0
Medium SES 0.6 0.3 100.0
Low SES 1.3 1.0 100.0

TOTAL 0.6 0.5 100.0

[chi square] = 10.85

[alpha] = 0.210
Table 5c

Change in Life Satisfaction in Saskatoon Neighbourhoods

Neighbourhood Percent responses by Category (N-957)
Cluster Improved Stayed the Same Gotten Worse

High SES 49.9 43.8 6.3
Medium SES 48.0 44.9 7.1
Low SES 48.0 41.6 10.4

TOTAL 48.6 43.5 7.9

Neighbourhood Percent
 responses
 by
 Category
 (N-957)
Cluster Total

High SES 100.0
Medium SES 100.0
Low SES 100.0

TOTAL 100.0

[chi square] = 4.328

[alpha] = 0.363
Table 6a

Satisfaction with Own Neighbourhood

Neighbourhood Percent responses by Category (N = 947)
Cluster Very Somewhat Somewhat
 Satisfied Satisfied Dissatisfied

High SES 68.2 26.1 4.1
Medium SES 71.2 26.9 1.6
Low SES 37.4 38.0 17.3

TOTAL 59.0 30.3 7.6

Neighbourhood Percent responses by
 Category (N = 947)
Cluster Very
 Dissatisfied Total

High SES 1.6 100.0
Medium SES 0.3 100.0
Low SES 7.3 100.0

TOTAL 3.1 100.0

[chi square] = 134.12

[alpha] = 0.0001
Table 6b

Satisfaction with City

Neighbourhood Percent responses by Category (N = 951)
Cluster Very Somewhat Somewhat
 Satisfied Satisfied Dissatisfied

High SES 61.5 34.4 3.5
Medium SES 56.9 38.5 4.0
Low SES 49.4 43.6 4.8

TOTAL 55.9 38.8 4.1

Neighbourhood Percent responses by
 Category (N = 951)
Cluster Very
 Dissatisfied Total

High SES 0.6 100.0
Medium SES 0.6 100.0
Low SES 2.2 100.0

TOTAL 1.2 100.0

[chi sqaure] = 13.1

[alpha] = 0.041
Table 7

Spearman's Rank Order Correlations ([r.sub.s]) with Neighbourhood
Designation

Variables: Correlation Coefficient
 ([r.sub.s]):

Individual Variables

Employment Status 0.148 (*)
Education 0.205 (*)
Income 0.289 (*)
Income Adequacy 0.297 (*)
Perception of Personal Stress 0.028
Personal QOL 0.018
Personal Relations -0.055
Change in Health 0.012
Change in Overall Life Satisfaction 0.030

Composite Variables (*)

Social Cohesion -0.135 (*)
Satisfaction 1-External Structures -0.238 (*)
Satisfaction3-Personal Relationships -0.126 (*)
Neighbourhood-Perceptual -0.272 (*)
Neighbourhood-Social Programs -0.176 (*)
Neighbourhood-Amenities -0.181 (*)
Neighbourhood-Physical -0.200 (*)

(*)Significant at [alpha] = 0.05 (two-tailed)
Table 8a

Rating of Degree of Neighbourhood Neatness

Neighbourhood Percent responses byCatgory (N=959)
Cluster Excellent Very Good Good

High SES 10.9 32.7 44.6
Medium SES 14.5 32.9 42.0
Low SES 7.9 26.6 31.0

TOTAL 11.2 30.8 39.2

Neighbourhood Percent responses byCatgory
 (N=959)
Cluster Fair Poor Total (*)

High SES 10.6 1.3 100.0
Medium SES 7.6 3.0 100.0
Low SES 18.4 16.1 100.0

TOTAL 12.1 6.8 100.0

[chi square] = 95.468

[alpha] = 0.001

(*)Totals may not to 100.0% due to rounding errors.
Table 8b

Rating of Friendliness of the Neighbourhood

Neighbourhood Percent responses byCategory (N=957)
Cluster Excellent Very Good Good

High SES 11.2 30.7 44.7
Medium SES 12.0 29.8 43.4
Low SES 8.3 27.6 40.4

TOTAL 10.6 29.4 42.8

Neighbourhood Percent responses byCategory
 (N=957)
Cluster Fair Poor Total

High SES 10.9 2.6 100.0
Medium SES 14.2 0.6 100.0
Low SES 16.0 7.7 100.0

TOTAL 13.7 3.6 100.0

[chi square] = 30.731

[alpha] = 0.001
Table 8c

Rating of Safety from Violent Crime in the Neighbourhood

Neighbourhood Percent responses byCategory (N=961)
Cluster Excellent Very Good Good

High SES 11.1 34.9 43.2
Medium SES 11.5 34.7 35.6
Low SES 5.4 18.1 34.6

TOTAL 9.4 29.3 37.8

Neighbourhood Percent responses byCategory
 (N=961)
Cluster Fair Poor Total

High SES 7.6 3.2 100.0
Medium SES 15.7 2.4 100.0
Low SES 19.4 22.5 100.0

TOTAL 14.3 9.3 100.0

[chi square] = 137.09

[alpha] = 0.001
Table 8d

Rating of Safety from property Crime in the Neighbourhood

Neighbourhood Percent responses by Category (N=960)
Cluster Excellent Very Good Good

High SES 7.6 25.3 39.2
Medium SES 6.4 24.8 38.8
Low SES 3.2 15.6 30.9

TOTAL 5.7 22.0 36.4

Neighbourhood Percent responses by Category
 (N=960)
Cluster Fair Poor Total

High SES 20.6 7.3 100.0
Medium SES 22.4 7.6 100.0
Low SES 21.0 29.3 100.0

TOTAL 21.4 14.6 100.0

[chi square] = 87.77

[alpha] = 0.001
Table 8e

Rating of Neighbourhood Organizations

Neighbourhood Percent responses by Category (N=763)
Cluster Excellent Very Good Good Fair Poor

High SES 5.9 28.5 44.7 13.8 7.1
Medium SES 3.5 24.2 41.8 16.0 14.5
Low SES 5.1 13.0 37.0 19.7 25.2

TOTAL 4.8 21.9 41.2 16.5 15.6

Neighbourhood Percent
 responses
 by
 Category
 (N=763)
Cluster Total

High SES 100.0
Medium SES 100.0
Low SES 100.0

TOTAL 100.0

[chi square] = 47.82

[alpha] = 0.001
Table 9

Spearman's Rank Order Correlations ([r.sub.s]) Among Composite Variables
by Neighbourhood Cluster

 SC Sat1 Sat2 Sat3

Social Cohesion

 High SES 0.173 (**) 0.074 0.098
 Medium SES 0.184 (**) 0.169 (**) 0.146 (*)
 Low SES 0.337 (**) 0.165 (**) 0.216 (**)

Satisfaction1-External
Structures

 High SES 0.294 (**) 0.309 (**)
 Medium SES 0.337 (**) 0.208 (**)
 Low SES 0.403 (**) 0.239 (**)

Satis2

 High SES 0.202 (**)
 Medium SES 0.125 (*)
 Low SES 0.176 (**)

Satisfaction3-Personal
Relationships

 High SES
 Medium SES
 Low SES

Neighbourhood-Perceptual

 High SES
 Medium SES
 Low SES

Neighbourhood-Social Programs

 High SES
 Medium SES
 Low SES

Neighbourhood-Amenities

 High SES
 Medium SES
 Low SES

 Nperc Nprog Namen

Social Cohesion

 High SES 0.213 (**) 0.183 (**) 0.184 (**)
 Medium SES 0.293 (**) 0.253 (**) 0.246 (**)
 Low SES 0.352 (**) 0.181 (**) 0.187 (**)

Satisfaction1-External
Structures

 High SES 0.296 (**) 0.389 (**) 0.130 (*)
 Medium SES 0.181 (**) 0.254 (**) 0.190 (**)
 Low SES 0.503 (**) 0.302 (**) 0.213 (**)

Satis2

 High SES 0.078 0.223 (**) 0.053
 Medium SES 0.124 (*) 0.180 (**) 0.191 (**)
 Low SES 0.260 (**) 0.220 (**) 0.258 (**)

Satisfaction3-Personal
Relationships

 High SES 0.101 0.227 (**) 0.102
 Medium SES 0.l96 (**) 0.236 (**) 0.120 (*)
 Low SES 0.125 (*) 0.209 (**) 0.176 (**)

Neighbourhood-Perceptual

 High SES 0.358 (**) 0.301 (**)
 Medium SES 0.328 (**) 0.263 (**)
 Low SES 0.390 (**) 0.454 (**)

Neighbourhood-Social Programs

 High SES 0.466 (**)
 Medium SES 0.430 (**)
 Low SES 0.528 (**)

Neighbourhood-Amenities

 High SES
 Medium SES
 Low SES

 Nphys

Social Cohesion

 High SES 0.090
 Medium SES 0.198 (**)
 Low SES 0.195 (**)

Satisfaction1-External
Structures

 High SES 0.251 (**)
 Medium SES 0.209 (**)
 Low SES 0.403 (**)

Satis2

 High SES 0.027
 Medium SES 0.093
 Low SES 0.272 (**)

Satisfaction3-Personal
Relationships

 High SES 0.005
 Medium SES 0.120 (*)
 Low SES 0.026

Neighbourhood-Perceptual

 High SES 0.534 (**)
 Medium SES 0.538 (**)
 Low SES 0.562 (**)

Neighbourhood-Social Programs

 High SES 0.346 (**)
 Medium SES 0.351 (**)
 Low SES 0.305 (**)

Neighbourhood-Amenities

 High SES 0.297 (**)
 Medium SES 0.238 (**)
 Low SES 0.280 (**)

(**)significant at [alpha] = 0.01 (two-tailed)

(*)significant at [alpha] = 0.05 (two-tailed)

Notes

(1.) Differences in the number of responses later in the analysis reflect the fact that cases with missing values were excluded from the multivariate analysis to give an N = 917. Later in the paper, descriptive statistics use all responses to individual questions.

(2.) Income adequacy is derived from reported household income divided by the number of people supported within the household.

(3.) Of these derived variables, only Social Cohesion correlates strongly with gender ([r.sub.s] = 0.106), and even here the significance is primarily a function of the large sample size. Conversely, income and "income adequacy" are very strongly correlated with Satisfaction 1-External Structures, Satisfaction 2-Public Relationships (consisting of satisfaction with treatment by store owners, public employees, satisfaction with main job or activity and work/family balance), Satisfaction 3-Personal Relationships, Neighbourhood-Perceptual, and Neighbourhood-Physical.

References

Abrams, M. 1973. "Subjective Social Indicators" Social Trends 4: 35-50.

Baldazzi, B.; H.C. Corrado, and a. Ferruzza 1998. "The quality of life in some local metropolitan labour market in Italy" Social Indicators Research 44(1): 71-95.

Bederman, S.H. 1974. "The Stratification of the Quality of Life in the Black Community of Atlanta, Georgia" Southeastern Geographer 14: 26-37.

Bowman, T.F., G. A. Giuliani, and M. R. Minge 1981. Finding Your Best Place to Live in America. New York: Warner Books.

Boyer, R. and D. Savagneau 1985. Place Rated Almanac: Your Guide to Finding the Best Places to Live in America. New York: Rand MacNally.

_____. 1989. Place Rated Almanac: Your Guide to Finding the Best Places to Live in America. New York: Rand MacNally.

Canada Mortgage and Housing Corporation 1991. Draft discussion paper. Quality of Life: Issues and Directions. Ottawa: CMHC.

Canadian Council on Social Development 1996. Measuring Well-being: Proceedings from a Symposium on Social Indicators. Ottawa: Canadian Council on Social Development.

Canter, D. 1983. "The Purposive Evaluation of Places: A Facet Approach" Environment and Behavior 15: 659-98.

Clark, N.M. 2000. "Understanding Individual and Collective Capacity to Enhance Quality of Life" Health Education and Behavior 27, 6 December: 699-707.

Connerly, C.E. and R.W. Marans 1985. "Comparing Two Global Measures of Perceived Neighbourhood Quality" Social Indicators Research 17: 29-27.

Conway, H. and L. Listen 1981. The Good Life Index. Atlanta: Conway Publications.

Cutter, S.L. 1985. Rating Places: A Geographer's View on the Quality of Life. Washington: Association of American Geographers Resource Publications in Geography.

De Koning, K. and M. Martin 1996. Participatory Research in Health: Issues and Experiences. London, England: Zed Books.

Dickinson, J.C. III, R.J. Gray, and D.M. Smith 1972. "The Quality of Life in Gainsville, Florida: An Application of Territorial Social Indicators" Southeastern Geographer 12:121-132.

Dilks, D. 1996. "Measuring Urban Sustainability: Canadian Indicators Workshop June 19-21, 1995." Toronto: Prepared for State of the Environment Directorate, Environment Canada and the Centre for Future Studies in Housing and Living Environments, Canada Housing and Mortgage Corporation.

Dissart, J.C. and S.C. Deller 2000. "Quality of life in the planning literature" Journal of Planning Literature 15, 1 August: 135-161.

Downs, R.M. and D. Stea 1977. Maps in Minds: Reflections on Cognitive Mapping. New York: Harper and Row.

Evans, D.R., J.E. Bums, W.E. Robinson, and O.J. Garrett 1985. "The Quality of Life Questionnaire: A Multidimensional Measure" American Journal of Community Psychology 13:305-322.

Findlay, A., R. Rogerson, and A. Morris 1988a. "In What Sense 'Indicators' of Quality of Life" Built Environment 14: 96-106.

Findlay, A., A. Morris, and R. Rogerson 1988b. "Where to Live in Britain in 1988: Quality of Life in British Cities" Cities 5:268-276.

Friedman, M.I. 1997. Improving the quality of life: a holistic scientific strategy. Westport: Praeger.

Furuseth, O. and W.A. Walcott 1990. "Defining Quality of Life in North Carolina" The Social Science Journal 27:75-93.

Gariepy, M., G. Domon, and P. Jacobs 1990. Development viable et environmentale en milieu urbain: essai d'application au cas montrealais. Montreal. Faculte de l'amanagement, Universite de Montreal.

Helburn, N. 1982. "Geography and the quality of life" Annals of the Association of American Geographers 36(4):445-456.

Johnston, R., D. Gregory, and D.M. Smith 1986. The Dictionary of Human Geography, Second Edition. Oxford, UK: Blackwell Reference.

Knox, P. 1976. "Social Well-Being and North Sea Oil: An Application of Subjective Social Indicators" Regional Studies 10:423-432.

Lee, K.2000. Urban Poverty in Canada: A Statistical Profile. Ottawa: Canadian Council on Social Development.

Liu, B. 1976. Quality of Life Indicators in US. Metropolitan Areas: A Statistical Analysis. New York: Praeger.

Lyon, L. 1987. The Community in Urban Society. Philadelphia: Temple University Press.

Marlin, J.T. and J.S. Avery 1983. The Book of American City Rankings. New York: Facts on File.

Mastekaasa, A. and T. Moun. 1984. "The Perceived Quality of Life in Norway: Regional Variations and Contextual Effects" Social Indicators Research 14:385-419.

Metro Toronto 1991. Towards a Liveable Metropolis. Toronto: Metropolitan Planning Department, Metropolitan Plan Review Report No. 13.

Moller, V. and L. Schlemmer 1983. "Quality of Life in South Africa: Towards an Instrument for the Assessment of Quality of Life and Basic Needs" Social Indicators Research 12:225-279.

Morton, P. 1999. Quality of Life in Saskatoon 1991 and 1996: A Geographical Perspective. Unpublished Master's Thesis, University of Saskatchewan, Department of Geography.

Myers, D. 1987. "Community-Relevant Measurement of Quality of Life: A Focus on Local Trends" Urban Affairs Quarterly 23: 108-125.

_____. 1988. "Building Knowledge about Quality of Life for Urban Planning" Journal of the American Planning Association 54: 347-358.

Nissan, E. 1989. "An Alternative Index for Quality of Life in Urban Areas" The Review of Regional Studies 19(1): 77-80.

Olsen, M. E., P. Canan, and M. Hennessy 1985. "A Value-Based Community Assessment Process: Integrating Quality of Life and Social Impact Studies" Sociological Methods and Research 13: 325-361.

Pacione, M. 1980. "Differential Quality of Life in a Metropolitan Village" Transactions: Institute of British Geographers 5(2): 185-206.

_____. 1986. "Quality of Life in Glasgow: An Applied Geographical Analysis" Environment and Planning A 18: 1499-1520.

Pierce, R.M. 1984. "Rating America's Cities: A Perceptual Analysis of Objective Measures." Washington: Paper presented at the Annual Meetings of the Association of American Geographers.

Randall, J. 1999. "Quality of Life--Cities and Towns." In Atlas of Saskatchewan, edited by K.I. Fung. Saskatoon: University of Saskatchewan.

Reid, A. 1991. Urban Canada Study. As reported in the Newsletter, Institute of Urban Studies, University of Winnipeg, December, 1991.

Rogerson, R. 1999. "Quality of Life and City Competitiveness." In Urban Studies 36, 5-6:969-985.

Saarinen, T.F. 1976. Environmental Planning: Perception and Behaviour. Boston: Houghton Mifflin.

_____. 1981. "Environmental Perception" Progress in Human Geography 5: 525-547.

Saarinen, T.F. and J.L. Sell 1980. "Environmental Perception" Progress in Human Geography 4: 535-548.

Schulman, N. and W. Bond. 1978. Urban Indicators: Statistical Profiles of Quality of Life for Canadian Cities. Revised edition. Ottawa: Human Environment Directorate, Ministry of the State for Urban Affairs.

Schwirian, K.P., A.L. Nelson, and P.A. Schwirian 1995. "Modelling Urbanism: Economic, Social and Environmental Stress in Cities" Social Indicators Research 35: 201-223.

Smith, D.M. 1973. The Geography of Social Well-Being in the United States: An Introduction to Territorial Social Indicators. New York: McGraw-Hill.

Statistics Canada (2000). Neighbourhood inequality in cities. The Daily (December 13). Ottawa: Statistics Canada.

Szalai, A. and F.M. Andrews (Eds) 1980. The quality of life: comparative studies. London: Sage.

Tahlin, M. 1990. "Politics, Dynamics, and Individualism - the Swedish Approach to Level of Living Research" Social Indicators Research 22:155-180.

Turksever, N.E. and G. Atalik 2000. "Possibilities and Limitations for the Measurement of the Quality of Life Index in Urban Areas" Social Indicators Research 53: 163-187.

Vogel, J. 1997. "The Future Direction of Social Indicator Research" Social Indicators Research 42:103-116.

Walter-Busch, E. 1983. "Subjective and Objective Indicators of Regional Quality of Life in Switzerland" Social Indicators Research 12:337-391.

Wish, N.B. 1986. "Are we Really Measuring the Quality of Life? Well-being has Subjective Dimensions, as well as Objective Ones" American Journal of Economics and Sociology 45:93-99.