SCORE is a robust scientific assessment tool that is carefully calibrated to each context to investigate societal dynamics and guide evidence-based policy and programme design for enhancing social cohesion. For a more detailed explanation of SCORE methodology, please read the short methodology paper here.
The SCORE in Bosnia and Herzegovina was first implemented in 2014 by USAID in partnership with SeeD. This was conducted between December 2013 and April 2014, with a representative sample of 2,000 respondents, largely made up of the three constituent ethnic groups of BiH. Data collection was conducted by Prism Research. The overarching aims of the 2014 SCORE BiH were to assess social cohesion, with a focus on intergroup relations and an outcome of readiness for political compromise and the ability to envision a shared future with other ethnic groups.
The SCORE in Bosnia and Herzegovina was repeated in 2019 in partnership between United States Agency for International Development, Office of Transition Initiatives (USAID/OTI), International Organisation for Migration (IOM) and The Centre for Sustainable Peace and Democratic Development (SeeD), with the primary aim of supporting the efforts of USAID/OTI and IOM’s Bosnia and Herzegovina Resilience Initiative (BHRI). The overall objective of the BHRI is to strengthen community resilience and to address the threat of violent extremism, by strengthening alternative narratives that challenge division and the ability of institutional and community actors to mitigate and respond to escalatory violence.
As a result of the emerging priorities of the 2019 SCORE partnership and in line with developments in the SCORE methodology between the two time points, 2019 data contains new indicators that were not measured in 2014 (e.g. Ethnonationalism, Exposure to Ethnonationalist Narratives). Additional research tools were employed in the 2019 SCORE, namely, SeeD’s Resilience Assessment Framework, alongside existing tools, such as the Intergroup Comparison Framework which was used in both project cycles. Due to the renewed scope and programmatic significance of the 2019 SCORE, additional geographical regions were sampled in this latter wave.
Although the two iterations of the SCORE in BiH did not survey the same respondents, certain measurement indicators were preserved across time points, allowing for the inference of temporal trends. These should be interpreted with caution, as complete comparability was often not possible to maintain.
For the SCORE BiH 2019, data was collected by Prism Research between December 2019 and March 2020, achieving a total sample of 4,570 respondents. The majority of the sample was demographically and geographically representative of BiH. A proportion of the sample was used to survey priority groups, namely, youth under the age of 35 and respondents living in the beneficiary areas of the BHRI programme.
The major aims of the SCORE BiH 2019 were:
The indicators of the 2014 and 2019 SCORE Bosnia and Herzegovina, as well as the relationships between them can be explored and disaggregated interactively on this platform. For information on how to use the platform, you can watch the short video on our Facebook page here or read the How to Read SCORE manual here.
The first editions of our methodology were developed in 2012 in Cyprus in a partnership between UNDP and USAID. Our approach is based on participatory research and mixed methods, in which multi-level stakeholder consultations, focus groups and interviews inform the design and calibration of the questionnaire, which draws from our extensive library of measurement instruments and indicators. The figure below illustrates our Process Framework.
Figure 1. Process Cycle
Our participatory and consultative approach cultivates:
The Process Framework is underpinned by a Content Framework, which helps us align and calibrate research objectives with the specific desirable outcomes. The Content Framework focuses on dimensions of societal functioning which can either contribute to stable, prosperous and resilient societies, or, if they remain unattended, undermine social cohesion.
Figure 2. Content Framework
Indicators measure a particular phenomenon (e.g. economic security, active citizenship, level of education, tolerance to corruption etc.), the definition of which can be found in the glossary search box. Indicators are generally measured using a minimum of 3 questionnaire items and their validity confirmed using statistical reliability tests to ensure that the different dynamics underlying the indicator are well-captured. A score from 0 to 10 is calculated for each indicator. 0 means the phenomenon the indicator is measuring is not observed in the context at all, and 10 means that the phenomenon is prevalent. For example, if we want to denote the extent to which people feel safe from violence in daily life, a score of 0 would mean that no one feels secure, while 10 would signify that every person feels secure.
Heatmaps show how indicators are represented across different geographical areas, illustrating regional differences to identify areas of concern or priority to tailor policies and programmes and to improve resource allocation more precisely.
Path Analyses (Predictive Models) represent relationships between indicators based on advanced statistical analysis (e.g. regression, network analysis and structural equation modelling). In models, the relationships are directional, and they should be read from left to right. They have predictive power and are used to identify key drivers of change in society. Models reveal what influences an indicator or what this indicator influences itself. Indicators can be “drivers” as they positively or negatively predict the other indicators they are linked to. In a model, the indicator that the drivers are predicting is called an ‘outcome’. Outcomes are at the right end of the model, and they are usually the end goals that we want to influence. Red connecting lines in models represent a negative relationship and blue connecting lines represent a positive relationship between indicators. The thickness of arrows indicates the strength of the relationship between the indicators. Models should not be confused with correlations, where lines represent associations, but are not directional.
Resilience Analyses are used in contexts where there is a need to nurture a population’s resilience against adversities or risks. These analyses identify which personal assets and community resources will most effectively interrupt pathways from risk exposure to detrimental outcomes. Resilience analyses are best suited for data collected from the same respondents at two different time points. The statistical methodologies relevant for resilience assessments include structural equation modelling, to detect the pathways from exposure to outcome, followed by moderation analyses, which identify the sources of resilience that interrupt specific risk-to-outcome pathways. Once the sources of resilience have been identified, these can be investigated as outcomes themselves in order to identify which capabilities, skills, contextual factors or community assets lead to higher levels of resilience. Intergroup Comparison Frameworks are used in contexts where there is a need to investigate the relationships between groups, and to compare between groups. Intergroup comparisons identify group divisions in society, either to encourage cooperation and reconciliation or to give insights on specific group needs. Statistical methodologies used in intergroup comparisons include latent profile analysis, multivariate analysis of variance and complex network analyses. The results of other analyses, such as path or resilience analyses, can be disaggregated for each group to identify the extent to which social dynamics differ between the groups in question.