Measuring Improvement in Quality of Life in Community-based Poverty Reduction Projects in Nigeria.docx

Measuring Improvement in Quality of Life in Community-Based Development Projects in Nigeria Muhammad Zayyanu1*, Zungwenen U. J.2, Johar F.3, Rafee M. Majid4 1, 2, 3, 4 Faculty of Built Environment, Universiti Teknologi Malaysia *zmuhammed1140@gmail.com Abstract The lingering problem of poverty prompted many countries to adopt community-based strategies for improving Quality of Life (QoL) of poor communities. Numerous studies have focused on identifying but ignored to establish the contribution of the factors that influence improvement in QoL by community-based projects. This paper measures, using Structural Equation Modelling, the contribution of the factors that influence QoL in a Community-based Poverty Reduction Project (CPRP) in Nigeria. The model revealed that the measured factors contributed only 36% of the reduction in poverty, which implies that there are other “hidden” factors responsible for the improvement in the quality of life. Keywords: Quality of Life, Community-based Projects, Poverty Reduction, Structural Equation Modelling © 2017. The Authors. Published for AMER ABRA by e-International Publishing House, Ltd., UK. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer–review under responsibility of AMER (Association of Malaysian Environment-Behaviour Researchers), ABRA (Association of Behavioural Researchers on Asians) and cE-Bs (Centre for Environment-Behaviour Studies), Faculty of Architecture, Planning & Surveying, Universiti Teknologi MARA, Malaysia.8pt Italic DOI: please leave empty 1. Introduction The concept of Quality of life (QoL) relates to prosperity and general welfare of individuals (Abdul Karim 2012; Aklanoğlu & Erdoğan 2012; Hanifah & Hashim 2012; Mohit 2013b). Many development experts favored and argued that the adoption of QoL approach for community development would help local communities to improve their quality of life. As supported by Hamdan et al. (2014), the community-based strategy can enhance societal well-being and quality of life. Many countries adopted community-based strategies to reduce the number of people with an income of less than $1 a day (Chamhuri et al. 2012). Following an agreement between Nigeria and the World Bank, the federal government implemented a Community-based development program as a strategy for Poverty reduction in Nigeria. The program established the Kebbi-state Community-based Poverty Reduction Project (CPRP) in 2001 which emphasizes the involvement of poor communities in the planning, execution, and management of community-level projects. The target of the project is to reduce poverty of poor communities which will, in turn, improve their quality of life. This paper aims to measure the contribution of the factors that influence improvement in the quality of life with particular reference to the Kebbi-state Community-based Poverty Reduction Project in Nigeria. Using a conceptual framework for measurement of QoL and Poverty reduction developed from literature review, the paper measured the contribution of factors of participation in the CPRP in Nigeria. The findings of the study will broaden the understanding of the various factors that influence improvement in QoL of community-based development projects in Nigeria. 2. Literature Review The quality of life can be measured using both objective and subjective parameters (Mohit 2013a; Ana-Maria 2015). Various authors employed different approaches for measurement of QoL as there is no single universally accepted method for its measurement (Rybakovas 2014). For instance, while Marans (2003) and McCrea et al. (2006) favored an objective approach that is not influenced by subjective opinion, Veenhoven (2000) argues that QoL should be measure based on individual perceptions. While the objective method assesses the actual circumstances of people, the subjective approach is more concerned with individual’s satisfaction and feelings about QoL (Muslim et al. 2013). The objective approach measures what people consider being essential to societal well-being, while the subjective measures are more concerned with feelings, experiences, and behavior pattern of individuals (Mohit 2013a). Numerous authors employed either the subjective or objective approaches to assess QoL. For instance, using subjective parameters, Noor & Abdullah (2012) investigated Quality of Work Life (QOWL) in multinational firms in Malaysia. Latif et al. (2013) examined the influence of situational factors (subjective) of QoL on recycling behavior in Malaysia. Using objective approach, Mohit (2013a) studied regional variations in the QoL in Malaysia. Despite many studies on the measurement of both subjective and objective quality of life, there is a dearth of studies that empirically test the link between subjective (reflective) satisfaction with the objective improvement (formative) in QoL (McCrea et al. 2011). Michalos (2008) observed that a comprehensive evaluation of the QoL must assess both objective and subjective parameters. As supported by Mohit (2013b), using the two approaches allows the weakness of one approach to be complemented by the strength of the other. Based on the ideas of Michalos (2008) and Mohit (2013b), Rybakovas (2014) expressed the opinion that the overall perceived (subjective) QoL by individuals consists of a set of latent (hidden) variables which are dependent on the measurable variables (objective QoL). Similarly, Maggino & Zumbo (2012) opined that the empirically observable subjective indicators tend to reflect on latent (objective) variables, which are not open to people’s perception and experience. 2.1 Measurement of Poverty There is no universally accepted criteria for measuring poverty. Waheed (2012) identified various approaches for the measurement of poverty. The approaches are poverty gap income shortfall, composite poverty measures, the physical quality of life index (PQOLI), the augmented physical quality of life index (APQLI) and the human development index (HDI). However, the approaches to measuring poverty have undergone refinement, which introduced the Multidimensional Poverty Index (MPI) as an improvement over the previous methods. The MPI has multiple indicators for measuring the multidimensional aspects of poverty and deprivation with regards to the development of individuals, households, and nations (Chamhuri et al. 2012). The multifaceted nature of the MIP is superior over other approaches because it identifies the poor and estimates the extent of poverty of individuals at the household level. It is assessed using indicators that are consistent with the three dimensions of the UNDP Human Development Index of Education, Health and Standard of living. 2.2 Indicators for Measuring Community Participation and Poverty Reduction Many studies have adopted various parameters for measuring community involvement and poverty alleviation. While community involvement is measured using ‘participation in community development’ (PCD), ‘empowerment’ (EMP) and ‘social capital’ (SOC), poverty reduction (PVR), is measured using indicators developed by Oxford Poverty and Human Development Initiative (University of Oxford 2010) (Table 1). Table 1: Constructs and measures for measuring participation and poverty reduction Constructs Variables Source Participation in Community Development (PCD) Membership of Community Organization, Narayan, (1995), CAG Consultants, (2009); Glass, (1979) Implementation of Projects Contribute Finance Provide Materials Provide Labour Empowerment (EMP) Awareness of Project Braathen, (2000); Narayan-Parker, (2002); Samah & Aref, (2009) Involvement in Community Meetings Contribute to Decision Making Supervision of Project Project Maintenance Social Capital (SOC) Solidarity and cooperation Ferragina, Tomlinson, & Walker, (2013); Woolcock & Narayan, (2000) Give/receive community Assistance Enhanced community development Self-actualization Mutual trust Poverty Reduction (PVR) Number of visit to health facility The University of Oxford, (2010) Nutrition improved Children in primary school Children in secondary school Improved housing condition Access to services Asset ownership 3. Methodology From the review of the literature, this paper adopted a conceptual framework for measuring improvement in the living standard of the project beneficiaries. The framework identified three constructs and fifteen variables for measuring community participation and seven indicators for measuring poverty reduction (Fig 1). The adopted variables received recognition by the reviewed literatures as shown in Table 1. The authors selected, using stratified sampling procedure, two micro-projects from each of the nine infrastructure sectors executed under the CPRP. Twenty households are then randomly selected from each community associated with the 18 selected micro-projects. Accordingly, a total of 360 questionnaires were administered using face-to-face delivery. However, availability of functional micro-projects limits the selection of samples for the study. The study contends that the limitation is to allow for measurement of improvement in QoL in communities with operating micro-projects. The data was processed using SPSS and Structural Equation Modelling approach was employed to confirm the model and test the relationships using Amos software version 22. Figure 1: Conceptual framework for measuring poverty reduction 4. Results and Discussion The Structural equation modeling approach revealed results of the study using both measurement and structural models. While the measurement sub-model examined the relationship between the observed indicators and their underlying constructs (factors), the structural component explored the contribution of the factors to the improvement in poverty reduction of the CPRP project beneficiaries. However, in SEM analysis, there is the need to ensure the reliability and validity of the respective indicators in measuring their underlying constructs. 4.1 Exploratory Factor Analysis Exploratory Factor Analysis (EFA) was used to verify the internal reliability and validity of the research questionnaire due to the ability of the method to examine the structure of all the items in the measurement model (Hair et al., 2006). All the latent constructs achieve internal reliability with a Cronbach’s Alpha of greater than 0.700 (Table 2). The factor loadings of the four constructs (PCD, EMP, SOC, and PVR) used in the study shows excellent reliability with all the twenty-two items. Similarly, the analysis shows a Kaiser-Meyer-Olkin (KMO) values of between 80% and 90% measure of sampling adequacy which indicate a common variance among the measured variables. Table 2: Exploratory Factor Analysis (EFA) Construct Items Factor Loading Cronbach’s Alpha Number of Items Internal Reliability Participation in Community Development (PCD) PCD1 0.928 0.940 5 Excellent PCD2 0.925 PCD3 0.921 PCD4 0.925 PCD5 0.930 Empowerment (EMP) EMP1 0.770 0.788 5 Excellent EMP2 0.776 EMP3 0.709 EMP4 0.726 EMP5 0.752 Social Capital (SOC) SOC1 0.850 0.876 5 Excellent SOC2 0.831 SOC3 0.854 SOC4 0.851 SOC5 0.860 Poverty Reduction (PVR) QOL1 0.921 0.932 7 Excellent QOL2 0.922 QOL3 0.923 QOL4 0.917 QOL5 0.921 QOL6 0.918 QOL7 0.924 4.2 Evaluating the Fitness of the Measurement Model Confirmatory Factor Analysis (CFA) is used to assess and validate the measurement model aalidate the measurement model.ity of the respective indicators in measuring their underlying constructnd to test whether the measures of a construct are consistent with the researcher’s understanding of that constructs (Awang 2015). Every measurement model involving latent constructs needs to undergo CFA before modeling into SEM. However, due to the problems discovered when computing CFA separately for the individual constructs, Awang (2015) suggested the use of pooled CFA for all latent constructs simultaneously. Confirmatory Factor Analysis comprises of four stages: (1) defining the individual construct, (2) developing the overall measurement model, (3) designing a study to produce empirical results, and (4) assessing the model validity and reliability. In examining validity, three requirements of validity assessment must be achieved to the required level to achieve the model fit and to proceed to the structural model analysis. These include; convergent validity, Construct Validity, and Discriminant Validity. While, in assessing the reliability of the measurement model, Internal Reliability, Construct Reliability and Average Variance Extracted need to be evaluated.There are several fit indexes for evaluating the fitness of the SEM models. Table 3 shows the recommended fit indexes and their respective acceptable values. Table 3: Categories of Model Fit and their Level of Acceptance Name of Category Name of Index Index Full Name Level of Acceptance Absolute Fit Chi-Square Discrepancy Chi Square P-value >0.05 RMSEA Root Mean Square of Error Approximation <0.08 GFI Goodness of Fit Index >0.90 Incremental Fit AGFI Adjusted Goodness of Fit >0.90 CFI Comparative Fit Index >0.90 TLI Tucker-Lewis Index >0.90 NFI Normed Fit Index >0.90 Parsimonious Fit CMIN (Chisq/df) Chi Square/Degree of Freedom <3.0 The model in figure 2 generated 22 measurement variables that contain 253 sample moments. The degree of freedom for the model is 199 (253-54) and a chi-square goodness-of-fit statistics of 1152. The fit indices in figure 2 show that apart from the CFI (0.901), the other fitness indexes in the pooled CFA do not meet the recommended value of acceptance. The option is to delete or correlate the unnecessary items in the model to achieve validity and reliability. Figure 2: The measurement model The model modification was carried out, and a new specified model was estimated. The modification indices show correlation between e1 (PCD1) and e2 (PCD2); and e14 (SOC4) and e15 (SOC5). As expected, correlating the unnecessary items improved the model leading to achievement of all the fitness indices. Figure 3 shows the model fitness indexes: RMSEA=0.052, GFI=0.932, AGFI=0.914, CFI=0.964, TLI=0.959, NFI=0.947 and Chisq/df=2.878. The Chi-square goodness-of-fit model (572.730) is smaller compared with the Chi-square value of the original model (1152.199). The result of the modified pooled CFA shows a satisfactory fit model that achieved all the fit indexes. The re-specified measurement model meets the requirement of validity and reliability. Figure 3: Modified measurement model 4.3 Assessing the Validity and Reliability of the Measurement Model The factor loadings of the model shown in Table 4 are adequate for each construct. Similarly, the model achieved both “convergent” and “construct” validity adequately as both the Composite Reliability (CR) and the Average Variance Extracted (AVE) are above 0.6 and 0.5 respectively. The model also achieved discriminant validity as indicated in Table 5 because the bold and diagonal are greater than the preceding values in their rows and columns. Table 4: CFA Result for the Construct in the Model Construct Items Factor Loading CR (≥ 0.6) AVE (≥ 0.5) PCD PCD1 0.80 0.935 0.742 PCD2 0.82 PCD3 0.92 PCD4 0.88 PCD5 0.88 EMP EMP1 0.55 0.779 0.501 EMP2 0.52 EMP3 0.79 EMP4 0.77 EMP5 0.63 SOC SOC1 0.79 0.869 0.573 SOC2 0.89 SOC3 0.76 SOC4 0.68 SOC5 0.64 PVR QOL1 0.78 0.934 0.671 QOL2 0.77 QOL3 0.81 QOL4 0.87 QOL5 0.82 QOL6 0.82 QOL7 0.86 Table 5: Summary of Discriminant Validity Index for the Constructs Construct PCD EMP SOC PVR PCD 0.86 EMP 0.24 0.71 SOC 0.06 0.51 0.76 PVR 0.01 0.32 0.60 0.82 4.4 Structural Equation Model (SEM) In figure 4 the structural path model is presented and evaluated. The model explained 36% of the variance accounted for by the combined influence of the predictors (participation in community development, empowerment, and social capital). This result implies that the combined influence of the variables of community involvement in poverty reduction is 36% while 64% does not affect poverty alleviation. However, among the three factors, social capital has a more significant impact (0.59) on the relationship. The influence of social capital on poverty reduction is also buttressed by (Okunamadewa et al. 2005), (Dschang 2009), and (Santini et al. 2012). Figure 4. Model predicting poverty reduction Similarly, only one of the paths (SOC) out of the three linking the independent variables (PCD, EMP, and SOC) to the dependent variable (PVR) is significant at the critical ratio test (>±1.96, p<0.05). The probability of getting a critical ratio as large as 11.721 in absolute value is less than 0.001 (Table 6). In other words, the regression weight for SOC in the prediction of PVR is significantly different from zero at the 0.001 level (two-tailed). Table 6: Regression weights for path estimate Path Estimate (β) C.R P-Value Result PVR <--- PCD -0.028 -0.777 0.437 Not significant PVR<--- EMP 0.037 0.436 0.663 No significant PVR<--- SOC 0.604 11.721 *** Significant Poverty, being a multi-dimensional construct, has multiple cause-effect relationships. The 64% of the poverty reduction that could not be explained by the model is caused by other “hidden” factors other than those associated with community involvement. Therefore, implies that the lingering problems of poverty in developing countries are so complex that they cannot be solved by a community-based poverty reduction program alone. Investment in both physical and social infrastructure is necessary to reduce poverty (Ogun 2010). As observed by Hewett & Montgomery (2001), the inadequate provision of public services can stalemate efforts to alleviate poverty. For instance, lack of adequate water supply and sanitation can cause elevated health risks to households; and small-scale enterprises requiring electricity face higher production costs. 5. Conclusion The paper measured the influence of factors of community participation on poverty reduction towards enhancing the quality of life of the project beneficiaries. The study developed a model of improvement in Quality of Life for a Community-based poverty alleviation project in Nigeria. The finding of the study revealed a complimentary influence of the three dimensions of community involvement (community participation, empowerment, and social capital) in poverty reduction in Kebbi state, Nigeria. However, the findings of the study indicate that community involvement accounted for only 36% of the poverty reduction of the project’s beneficiaries. The authors, therefore, recommend the adoption of other poverty alleviation strategies that address the multidimensional nature of poverty in developing countries. Such strategies may focus on investigating other pro-poor natural sectors of the economy like agriculture to complement community-based development projects in developing countries. The authors contend that because the majority of poor people used agriculture as their primary source of income, focusing poverty reduction on the sector can tremendously reduce poverty.