Causal Inference in Spatial Analysis
Preface
What is Causal Inference in Spatial Analysis and why do you want it?
In recent years, interest in causality, reproducibility, and open science are transforming how social science is done. Broadly speaking, these trends have reinforced the importance of research design, causal inference, and model criticism in the social and environmental sciences. This is an especially pressing concern when research involves geographic processes, since they often require different ways of thinking and doing in order to analyze effectively. Our book is a bridge between contemporary teaching in social science (political science, sociology, economics) and the unique concerns of spatial data in geography and the environmental sciences. It is relevant to social scientists seeking to become familiar with causal research methods from scratch as well as learn the uniqueness of spatial data, and for geographers and environmental scientists seeking to learn cutting-edge causal research design and analysis. The text is split into two sections. First, we cover the core principles, concepts, and ethics of social science research design with no math or code. Second, we discuss common research designs and methods, including code and practical lab materials to learn causal research design and methods through widely-accessible real-world example problems pulled from diverse locations and problem domains across the globe.
Audience Needs
Our book has two core value propositions to cover different audience needs. First, our book supports students in geography and environmental sciences to learn about cutting-edge thinking in causal inference and research design. Second, our book supports social sciences learners who have been trained in causal inference and research design to understand spatial data and models. We outline these two needs below.
Students in Geography have their pick of research methods books: waves of data science, geographic information science, and geocomputation textbooks have come out over the past 40 years with an incredible diversity of approaches and pedagogical features. However, unlike much of the rest of social and environmental sciences, the topic of research design has no core or canonical textbook in geography—instead it’s often taught as a part of research methods. As contemporary trends in causal inference and open science have made research design more important over the past 40 years, geography and environmental sciences need a research design textbook to support teaching and learning about ethical, responsible, and causal research designs. Our textbook will fill this growing market gap as the first core textbook on causal methods and research design in geography and environmental sciences. Thus, for people within geography, we will provide an accessible and clear overview of both the contemporary principles and thinking behind causal research designs and the simple methods that underlie robust causal inferences.
Students in Social Sciences have a wide variety of texts to support learning causal research concepts, designs, and methods. However, there is currently no textbook covering causal inference methods in spatial analysis. This is a common pattern for teaching in social sciences: as thinking about statistics and research design evolves, pedagogical concerns about the unique dimensions of spatial or temporal research questions take time to settle and develop. Our text will be the first to provide this bridge between causal inference and research design in the social sciences and spatial analysis. Thus, for people outside geography, we will present spatial analysis concepts in a familiar manner, language, and setting.
Together, we think these two angles will stand to benefit many students. These approaches to thinking and critical reasoning are also lucrative for nonprofit, policy, and industrial research. Indeed, policy and industry have causal questions they want answered, too—questions students can answer by learning the concepts, designs, and methods in our text. Currently, these research designs and methods are not well-supported by textbooks in geography. Again from the other direction, much of the data we use to answer questions and solve problems in policy and industry involves geographic data; place and time are two fundamental components of big data and data by-products collected by networked digital technologies. But there are no textbooks to help outline how these problems can be solved. Thus, our textbook will also support learners to understand how to design and execute research that uses spatial analysis and causal inference to improve understanding. Overall, this textbook supports learning spatial analysis and causal inference methods needed by students in geography, students in other social and environmental sciences, and practitioners in nonprofit, policy, and industrial sectors.