Four analysis moments for fuzzy cognitive mapping in participatory research
Fuzzy cognitive mapping (FCM) is a practical tool in participatory research. Its main use is clarifying causal understandings from several knowledge sources. It provides a shared substrate or language for sharing views of causality. This makes it easier for different interest groups to agree what to do next. Each map is a collection of causal relationships with three elements: factors (cause and outcome), arrows linking factors, and weights indicating the perceived influence of each cause on its outcome. Stakeholder maps are soft models of how they see causes of an outcome, such as access to services or systemic racism. Based on a standardized FCM protocol, we present four moments in FCM analysis. (1) Agree shared meaning across maps. (2) Calculate the maximum influence of perceived causes. (3) Simplify the maps for communication. (4) Identify priorities for action. We provide explanations of the four moments in FCM analysis, with examples from five countries. FCM offers a practical means to guide health action. It incorporates local perspectives with transparent and traceable procedures.
Paper Context
Main findings: We present Fuzzy Cognitive Mapping as a shared platform for analyzing diverse knowledge sources and amplifying the voices of stakeholders in understanding, discussing, and acting on health issues.
Added knowledge: This manuscript addresses many of the questions we frequently receive from researchers who want to conduct participatory research globally.
Global health impact for policy and action: Fuzzy Cognitive Mapping is instrumental in the context of an increasing acceptance that evidence-based medicine, health management, and planning must leverage different types of evidence – some formal and scientific, others local with differences between stakeholders.
Fuzzy cognitive mapping (FCM) can clarify the understanding of how causes contribute and interact to influence an outcome [1–3]. Applications include environmental science [5], engineering [6], economics [7,8], organizational behaviour, information technology and, more recently, health care [9,10]. FCM has uses in machine learning, and it offers a computational framework [11,12] for predictive modelling [13–15]. It is also a powerful tool to include voices and knowledge of stakeholders in decision-making, the focus of this paper.
Participatory research uses FCM as soft models of stakeholder causal knowledge [1,16]. From an individual or group perspective, stakeholders map their view of causes of a particular outcome. Views of causality differ from person to person and group to group, and each view is partial and changing over time [17]. FCM provides a shared language for juxtaposing these different views. It allows comparisons and sometimes combination [18,19]. In our practice, FCM focuses on inclusion of under-represented stakeholder knowledge in shared decision-making [20,21]. The output is not a regularity- or variance-based vision of cauzation [18]. Independent of researcher paradigms, FCM portrays knowledges of causes [17] in terms the participants use and understand.
Creating fuzzy cognitive maps is straightforward but requires robust protocols for reproducibility across different settings [16]. Maps have three elements [2,3]. Factors or nodes represent causes and outcomes. Arrows represent relationships or causal links between nodes. Adding fuzzy to cognitive mapping, weights reflecting relative strength of influences between nodes. The influence between each cause on an outcome can be direct or indirect through intermediaries [18]. The maps can also reflect cyclic relationships with loops or two-way interactions. shows a fuzzy cognitive map built by traditional midwives in Guerrero, Mexico, to depict how they see contributing factors in healthy maternity in their communities.
Figure 1. Fuzzy cognitive map of factors contributing to better maternal health according to a group of traditional midwives in Guerrero, Mexico.
Legend. Indigenous traditional midwives in Guerrero, Mexico, drew a fuzzy cognitive map in a four-hour session with the support of a facilitator and two Indigenous intercultural brokers. The map presents protective factors for maternal health according to participants. Traditional midwives disaggregated maternal health into four outcomes: the woman is happy, the woman is strong and brave, the woman can give birth at home, and the woman does not get sick. a) The original map with labels in Spanish, and b) a digitized diagram with labels translated into English. The map has 12 factors (nodes) and 38 relationships (arrows). Appendix 1 includes the tabular representation of the maps as an adjacency matrix and an edge list.
Our four analysis moments draw on three graduate courses on advanced methods for participatory research ], over 200 hours of short courses for international researchers, and participatory research experience in 25 settings [
Our work with birth companions and family physicians in Montreal, Canada, contextualised published literature on unmet care needs of recent immigrant women [23]. In the Mexican state of Guerrero, we explored views of Nancue ñomndaa and Me’phaa traditional midwives on what promotes maternal health [24]. In Nigeria, FCM stimulated discussion about causes of short childbirth interval in Hausa communities [25]. Two examples from Botswana used FCM to map perceived causes of suicide among young men and violence against women [26–28]. We also provide references to other FCM applications that can inform decisions in FCM analysis.
Comments
Post a Comment