From Methodology to Practice Inductive Iteration in Comparative Research

• Published in: Comparative political studies Vol. 48; no. 5; pp. 616 - 644
• Main Author: Yom, Sean
• Format: Journal Article
• Language: English
• Published: Los Angeles, CA SAGE Publications 01.04.2015; SAGE PUBLICATIONS, INC
• Subjects: Causality; Comparative analysis; Comparative studies; Data mining; Historical analysis; Hypothesis; Iterative methods; Methodology; Qualitative analysis; Quantitative analysis; Transparency; qualitative methods; quantitative methods
• ISSN: 0010-4140; 1552-3829
• DOI: 10.1177/0010414014554685

Most methods in comparative politics prescribe a deductive template of research practices that begins with proposing hypotheses, proceeds into analyzing data, and finally concludes with confirmatory tests. In reality, many scholars move back and forth between theory and data in creating causal explanations, beginning not with hypotheses but hunches and constantly revising their propositions in response to unexpected discoveries. Used transparently, such inductive iteration has contributed to causal knowledge in comparative-historical analysis, analytic narratives, and statistical approaches. Encouraging such practices across methodologies not only adds to the toolbox of comparative analysis but also casts light on how much existing work often lacks transparency. Because successful hypothesis testing facilitates publication, yet as registration schemes and mandatory replication do not exist, abusive practices such as data mining and selective reporting find easy cover behind the language of deductive proceduralism. Productive digressions from the deductive paradigm, such as inductive iteration, should not have the stigma associated with such impropriety.