Resource - 8 Chapter 5 Community-based coral reef management in small islands: a social capital analysis

3.4 Data Analysis

This study employs both qualitative and quantitative approaches. A qualitative approach emphasizes words rather than numbers, in attempt to accurately describe and interpret the precise meaning of research focus. In contrast, a quantitative approach emphasizes numbers rather than words Bryman 2001:20 in Nurrochmat 2005. The analysis and discussions are relied on qualitative analysis. The quantitative analysis is important to substantiate the discussion, especially to quantify the qualitative data. Different methods of analysis are carried out. The prosess of data analysis is presented in Figure 11. Questions: Quantitative analysis of Multi-Criteria Analysis How is the condition of fishery sustainability? How is the condition of resource use and destructive fishery? Quantitative analysis of index of destructive fishery Methods of analysis: How bonding and bridging social capital influence the way resource is used? Questions: How investment in social capital contribute? Methods of analysis: Quantitative analysis of index of bonding and bridging social capital. Qualitative analysis of building bonding, bridging and linking social capital. Questions: Methods of analysis: How is the condition of coral reef? Questions: Data on percent coverage of live coral Methods of analysis: Resource characteristics How local rules are enforced? To what extent social capital investment shapes local rules and institutions? Qualitative analysis of institutions and collective action to: Overcome destructive fishery Chapter 7. Manage community marine sanctuary Chapter 8. Institution governing resource use Characteristics of resource users

Chapter 5 Resource

use Chapter 6

Chapter 7 - 8 Chapter 5

Figure 11 Research process of institutional analysis of coral reef management. Calculating Index of Destructive Fishery The outcome that intended to examine in this study is the resource use of the resident fishers. Compliance towards coral reef management is the minimal use of fishing using bomb or poison, among other things. Therefore, it is important to measure the rate of bomb and poison fishing. This is one of the challenges during research. Most of all, because the issue is taken in caution by fishers and local community, thus data is not readily and openly available for an external researcher. Thus, the parameterization is done by using available data that collected in 2004 and 2005. An index of destructive fishery is calculated from index of 2004 and of 2005. Index for 2004 is taken from the percentage of fishers using bomb or poison fish. Index for 2005 is calculated from the percentage of responses saying ‘frequent’ use of bomb or poison fishing by resident fishers. The calculation and result of this analysis is presented in Chapter 5. Participatory Multi-Criteria Analysis MCA To assess fishery sustainability, this study uses a formal methodology called multi-criteria analysis MCA. MCA is a general approach that can be used to analyze complex problems involving multi-criteria Mendoza and Prabhu 2003, and have advantages when applied in a complex and stochastic system like fisheries Adrianto et al. 2005. This method is suitable for three reasons. First, It can deal with mixed set of data, quantitative or qualitative, including stakeholders’ opinion. Secondly, it is conveniently structured to enable a collaborative planning and decision-making environment. Finally, it is simple, intuitive, and transparent, hile it has strong technical and theoretical support in its procedures. Following Mendoza and Prabhu ibid, MCA is used as a decision-making tool to analyze and evaluate sustainability under a participatory group decision- making environment. The use of this method can be used for generating criteria and indicators for sustainable resource management, estimating their relative importance, estimating the performance of each indicator relative to its desired condition. The analysis using MCA approach is done into two parts. The first part is to generate a set of sustainability indicators of fisheries. The methods used in this part of analysis are varied, ranging from expert driven and top-down to bottom up, and locally defined Adrianto et al. 2004. This study is following a study done by Adrianto et al. 2005 and using a mixed-method approach, in which it combines expert-driven fisheries sustainability indicators Pitcher 1999 and then these indicators are confirmed to the local stakeholders in order to generate a “local accepted” fishery sustainability indicators. The second part of analysis is evaluating the sustainability indicators in terms of their importance by ranking each indicators using a 5-point scale namely 1 – less important, 3 – moderately important, 5 – extremely important, and 2, 4 – intermediate value. A different scale is proposed by Mendoza and Prabhu 2003 which using 9-point of scale, and Adrianto et al. 2005 used 7-point of scale. However, for reason of simplicity during stakeholder meeting, this study uses 5- point scale. Based on these rankings, relative weight of an indicator is then estimated using a formula as follows Mendoza and Prabhu 2003; Adrianto et al. 2004: j j j a w a = ∑ 3 where j a is the average weight of indicator j and j w is the relative weight of indicator j . The next analysis examined each indicator by judging their current condition relative to their perceived target or desired condition Mendoza and Prabhu 2004; Adrianto et al. 2005. The desired condition was to reflect or represent a sustainable status of fishery sustainability indicators. In this respect, an MCA approach of 5-point scale is applied, following Adrianto et al. 2004, with values 1: extremely weak performance, strongly favorable, 2: poor performance, unfavorable, 3: acceptable, 4: very favorable performance, and 5: state of the art in the region. Then, the sustainability indicator score SIC is calculated using a formula: j j SIC s w = ∑ 4 where SIC is sustainability index of criteria i ecology, economy, social, and institution, j S is the score of indicator jand j w is the relative weight of indicator j Eq. 3. Calculating Index of Bonding and Bridging Social Capital Measuring social capital of the island community can be done by distinguishing their networks into bonding and bridging social capital. Bonding social capital is strong bonds of social relationships which are endorsed among family members or among members of an ethnic group. Bridging social capital is weaker but more cross-cutting ties of social relationships, which can be found in relationships from different ethnic groups or acquaintances Grootaert et al. 2003; Aldridge et al. 1999. Table 13 Variables of bonding and bridging social capital Criteria No Variable Operational definition Attribute Source of data Network 1 Source of the island’s production capital From where production capital i.e. financial of fishing or household enterprise come from Oneself parents; credit union ROSCA; island’s patron; patron in Makassar; outside patron Household survey Q1 2 Marketing of island’s product Where or to whom products of fish or household enterprise fuel, food, etc are sold Island resident i.e. patron, trader, or household; Makassar; outsider 3 Membership in fishing patron-client network Membership in fishing patron-client network Either yes or no Resource user survey Q2 4 Visit out Travel outside island within the last month for household or family reason Either yes or no Group 5 Membership in local group Membership of respondent in local groups: credit union, rotating saving credit associations ROSCA or sport Either yes or no 6 Membership of local group Membership of the local group Island resident; resident of neighboring islands; outsider Note: Q1 was obtained from Household survey April-June 2004, total N = 3,990. Q2 was obtained from Resource user survey July-October 2005, total N = 102. The measurement is done by developing a composite measure or an index. It is carried out by specifying different variables related to the attributes of groups and networks Table 13. Data is derived from the household and the resource user interviews based on structured questionnaires. Groups and networks are maintained by island residents and fishers. These are networks for obtaining production capital for fishing and household enterprise such as kiosk, fuel selling, food stall, house or ship builder, as well as for marketing their products. Each variable of networks and groups is then distinguished to each type of social capital, either bonding or bridging social capital. For example, when the source of production capital is coming from community members in the island or when the market of the products is sold to island residents, then this network is regarded as bonding. On the other hand, when the capital is coming from outside the island or the products are sold outside the island, then it is regarded as bridging social capital. The analysis and results are presented and discussed in Chapter 6. Statistical Analysis Statistical analysis is used to assess the relationships between various dimensions of social capital and reef fishery resource use. The model is proposed as follows: , , Y f SI SO X = 5 where Y is indicator of resource use; SI is indicator of input dimension of social capital, SO is indicator of output dimension of social capital; and X is other indicator. Each of indicator consists of several variables Table 14. Data for analysis is based on population survey and resource user interview that held in 2004 and 2005 respectively. The partial results for statistical analysis are presented in Chapter 6 on sections: community trust; vertical bonding social sapital and Chapter 7 on sections: conservation group; tolerance and fairness; and discount rate. They present some significant correlations between variables that are appropriate for discussion in the respective chapter. The complete result of statistical analysis is discussed in Prasetiamartati et al. 2006a. Most of data collected are nominal or ordinal variables, therefore the analysis is using nonparametric statistical analysis. The methods utilized are the Kruskal-Wallis test, Spearman’s correlation coefficient, and logistic regression. The SPSS Release 11.5.0 software is used to assist in data analysis. Table 14 Variables for each indicator Source of data Population survey 2004 Resource user interview 2005 Independent variables Indicators of resource use ƒ Bombposion fishing ƒ Bomb fishing ƒ Poison fishing ƒ Coral taking Independent variables Indicators of output social capital collective action ƒ Prohibition of destructive fishing ƒ Prohibition of taking coral ƒ Existence of conservation group ƒ Active conservation group Indicators of input social capital ƒ Network of source of capital ƒ Network of marketing ƒ Group membership ƒ Community trust ƒ Mutual help ƒ Community engagement ƒ Participation in community meetng ƒ Tolerance towards destructive fishing ƒ Fairness ƒ Agree on sanctioning destructive fishing Other indicators: 1 Government’s functioning ƒ Law enforcement ƒ Benefit of external assistance 2 Socio-economic condition ƒ Education ƒ Age ƒ Ethnic ƒ Alternative livelihood 3 Resource perception ƒ Crisis perception on fish ƒ Crisis perception on coral reef ƒ Perceived benefit on fish ƒ Perceived benefit on coral reef ƒ Perception on changes on fish size The Kruskal-Wallis H test, an extension of the Mann-Whitney U test, is the nonparametric analog of one-way analysis of variance and detects differences in distribution location. It is used in order to determine the influence of independent variables on dependent variable. The Kruskal-Wallis H test is used for data with several independent samples: 2 12 3 1 1 i i i R H n n n n = − + + ∑ 6 where H is Kruskal-Wallis coefficient, i n is number of observation on sample i , while i is number of data on each sample, and i R is rank on sample i . The Spearman’s correlation coefficient rho is a measure of association between rank orders, which is used to measure the association between dependent variable i Y and independent variables i X .The p-value shown is for testing the null hypothesis that no significant trend exists. If the p-value is 0,05 then the Spearman’s correlation coefficient r s is significant. The Spearman’s correlation coefficient is calculated using a formula: 2 1 3 6 1 n i s di r n n = = − − ∑ 7 where r s is Spearman’s correlation coefficient, i d is the difference between two ranks xi r - yi r , n is number of samples. Finally, in order to develop a model of fishery resource use that incorporated social capital, an analysis using logistic regression is carried out. Logistic regression is useful for situations in which you want to be able to predict the presence or absence of a characteristic or outcome based on values of a set of predictor variables. It is similar to a linear regression model but is suited to models where the dependent variable is dichotomous SPSS Inc. 2002.

Chapter 4 General Conditions of Study Sites