Causal Pluralism and Causal Inference, with Applications to Health and Status

点击翻译按钮获取中文摘要

There are a number of apparently competing philosophical accounts of causality now, each suggesting a particular privileged method for testing for causality; and each of these methods can be found in use across the sciences today. This project argues that these accounts are not genuinely competing. Rather, they are geared to describe different kinds of causal systems operating in different ways or to answer different kinds of causal questions. This NSF Science and Technology Studies Scholars Research Award aims to characterize the kinds of causal systems and question to which a given method is appropriate, focusing particularly on methods that look for invariance of causal relations under intervention. It does so in the special context of the attempt to understand the causal story responsible for the striking correlations between ill health and low socio-economic status. Improvements in methods for causal research lead to more reliable results on which we can base better policy. This is true not only for questions of health and status of the kind studied in the project, but across socioeconomic issues. No one methodological study or handful of studies will make dramatic impact. But a concerted, continuing effort to understand causality and to improve methods for causal inference, to which the proposed research should contribute, should improve research results; and better research results should mean sounder planning and policy.