Linking Theory and Inference
Page 83
Theme: Linking Theory and Inference; emphasis on moving from broad deliberative democracy to testable, middle-range theories; critique of overly broad definitions that harden tests against falsification. Mutz advocates developing and testing middle-range theories related to specific aspects of deliberative democracy, replacing vague, broad concepts with concrete, circumscribed ones and formulating empirically grounded hypotheses about specific relationships.
Good theory and testable hypotheses: A theory should yield specific, testable hypotheses (implications) that can be evaluated in the real world, regardless of whether the study is descriptive or causal. The ability to validate a theory is essential, but broader value comes from generating multiple hypotheses that extend beyond the immediate research question and enrich the scholarly conversation.
Hypothesis definition: A hypothesis is a specific statement derived from theory that can be tested; testable hypotheses help assess a theory’s soundness.
Iterative nature of research: Your question and theory are working, subject to refinement as you learn more. The literature review and theory evolve during the project, not as static components.
Page 84
Drafting a literature review early can help connect literatures, flag contestations, clarify theories in play, and promote hypothesis development. However, defer a full literature review until substantial progress on your project to avoid constraining your theoretical logic prematurely.
Compare theories with observable implications: Incumbents and re-election can be linked to money or to broader resources. Example: Incumbents may win because they have more money (limited observable implications) vs. broader resource-based theory (money, name recognition, campaign workers, casework goodwill, favorable media coverage). A theory with more observable implications is generally broader and more useful.
Incorporating theory into your study: The literature review is the bridge between prior literature and your research question; it is the theoretical justification for your hypotheses, the key variables, and the way you test them.
Page 85
The back-and-forth of literature shapes your working question and theory; iterate as you learn. Your working theory should be viewable as evolving with increased knowledge.
You will cite literature throughout your paper (not just in one section): to justify statements, decisions, theory, research design, results, and conclusions.
Decide whether you will test an existing theory or propose a new explanation/theory for the phenomenon under study.
Thinking about the Literature Review: The purpose is to explain the logic driving your study grounded in prior research (theory). The review reveals the main theories that make your research question worthy of study and identifies plausible rival theories to test alongside your own.
Page 86
Three goals of the literature review (reiterated):
1) Expand the discussion of the research question, linking key problems, theories, and data to justify salience and formulation.
2) Delineate key discussions, debates, and data related to your question and the theory you examine; show where your theory emerges and acknowledge plausible rival theories to grapple with inference.
3) Present your own working answer to the research question (your theory or hypotheses), typically with explicit hypotheses that can be tested. These hypotheses are often framed as if–then or when–then statements.Examples of hypotheses (illustrative):
Women’s estimation of campaign costs is significantly less accurate than men’s.
When men and women are exposed to the same information about running for office, there are no gender differences in terms of interest in pursuing political office.
The final component ties to data and methods: define concepts, determine which factors to include, and consider control variables (demographic information, partisan affiliation, etc.).
Theories guide the data and methods choices and the interpretation of findings.
Page 87
Writing the theory section and variables: Move from broad to specific; identify key concepts, define them, and specify the variables to be analyzed. Consider what data you will need and what you will measure to test your hypotheses.
Example framework (conceptual): If X is true, then Y should occur. When testing a theory about turnout, factors to consider include: (1) population scope (all voting-age citizens vs. eligible voters only), (2) what constitutes a “major scandal,” (3) other factors that explain turnout differences across states (election laws), and (4) additional significant factors from literature (to omit or include justified by evidence).
The last point emphasizes transparency about why certain variables are included or omitted and how this relates to theory.
Page 88
The literature review as a narrative: Though some papers label a separate literature review, others interweave it with theory and findings. The goal remains to show how the literature informs and justifies your study’s design.
The literature review should serve as a funnel: begin with broad problem context and significance, then narrow toward the debates and theories that shape your research question, and finally present your own theory and hypotheses.
The review should reveal the logic of your study: how prior research leads to your framing, concept definitions, and chosen variables, while also acknowledging rival theories and how you will test them.
Practical note: avoid treating the literature as the focal point; instead, present the empirical findings or theoretical insights themselves as the focal points of your sentences (emphasize the ideas, not the author).
Page 89
Box 3.2: The Increasing Prevalence of Co-authorship in Political Science
2006 APSA report: 40% of articles in top journals had more than one author (four times higher than 1950s/60s).
Co-authorship is more common in American politics; about one-third of studies in comparative politics and international relations are co-authored.
Benefits: collaborative exchanges generate new insights not easily found by a single author; diverse expertise enhances study design (e.g., combining political science with economics or psychology).
Authorship norms differ by field: economics often lists authors alphabetically; psychology lists by contribution. Alphabetical ordering can affect career outcomes despite norms.
The ordering of authors can influence tenure and recognition, highlighting the importance of properly attributing contributions.
Page 90
Box 3.3: What is Plagiarism?
Plagiarism is the theft of another’s intellectual work; many institutions offer resources to help distinguish acceptable from unacceptable practices.
Practical advice: take careful notes while reading; do not copy-paste from sources into your working documents; maintain separate notes with complete source citations.
A quoted footnote example emphasizes citing sources to deepen understanding and contribute to the scholarly conversation.
The section also notes that you should cite literature across your paper (not just in one section) to substantiate statements, theory, data choices, methodology, results, and conclusions.
Page 91
Two examples of theory building
Racial Prejudice and Voting for Obama
Post-2008 election: debates about whether whites voted for Obama due to race.
Existing literature indicated whites often oppose minority candidates due to race-related perceptions; however, theory requires mechanisms—why does race affect vote choice, beyond simple prejudice?
Schaffner’s work introduces priming: campaigns/events can make race more or less salient in voters’ minds, altering the effect of racial attitudes on vote choice.
The hypothesized mechanism: whites with high racial prejudice will be least likely to support minority candidates when race is primed; if race is not primed, prejudice may have a weaker or different effect.
Schaffner’s hypothesis and findings: support for the idea that priming moderates the impact of prejudice on Obama vote support.
Citation: Schaffner, Brian F., 2011, Racial salience and the Obama vote, Political Psychology, 32: 963–988.
Page 92
Are Women’s Organizations More Democratic?
Barakso asks whether women-led or women’s groups govern differently from mixed-gender groups.
Barakso builds a theory from literature across psychology, business, sociology, and political science, which finds that women are more cooperative, consensus-seeking, and more likely to seek others’ opinions.
Barakso cites evidence that female corporate managers delegate more, and female committee chairs take more integrative approaches.
Box notes that despite a strong theoretical expectation, Barakso did not find that women’s organizations were more democratically structured than other groups.
This creates a puzzle: theory drawn from multiple disciplines suggested a certain decentralized, democratic governance pattern, but empirical findings did not confirm it in this case.
Citations: Barakso, Maryann, 2007, Is there a ‘woman’s way’ of governing? Assessing the organizational structures of women’s membership associations, Politics & Gender, 3: 201–227.
Page 93
Box 3.4: How Political Science Draws from (and Contributes to) Research in Other Disciplines
The interconnectedness of social science disciplines: political science often cites economics, sociology, psychology, and other fields.
Figure 3.1 (map) shows citation traffic between disciplines; economics is the most influential source for other social sciences, followed by sociology and psychology.
The takeaway: high-quality political science often builds on theories and findings from multiple disciplines, which strengthens causal understanding and interpretation.
Page 94
Figure 3.1 (continued): A visual map of citation traffic among disciplines; depicts how interdisciplinary integration occurs in practice across political science, economics, sociology, psychology, and other fields.
The broader message: effective research in political science routinely integrates insights from multiple disciplines to inform questions, theories, and empirical tests.
Page 95
Taking Alternative Theories Seriously: What to Do When Theories and Hypotheses Do Not Match Findings?
Unexpected results require re-examining assumptions, models, and data for coding or data-entry errors.
If errors aren’t found, do not abandon the study; such results may reveal omitted information or missing theoretical considerations.
Possible explanations for unexpected/null findings:
Failure to draw on a wide enough set of theories to consider alternatives.
Insufficiently well-grounded hypotheses.
Omitted variables that would reframe the interpretation.
The antidote: actively test alternative theories and explanations to reduce the risk of omitted information; relate this to the broader goal of empirical political science: explain phenomena parsimoniously and accurately.
The Barakso case highlights how interdisciplinary literature can inform expectations, yet empirical results may still challenge those expectations; such outcomes prompt further scholarly puzzles.
The chapter emphasizes that “women’s organizations” literature did not confirm a universal democratic advantage, indicating the complexity of empirical generalization.
Page 96
The paradox of finding no relation between women’s representation and policy outcomes: even if women’s presence does not translate into different policy outcomes in a given study, it does not necessarily refute the broader claim that women’s under-representation matters for policy; context, issue areas, and constraints may influence results.
The importance of considering nuanced explanations: even if women and men hold similar policy preferences in some contexts, there may be mechanisms like social norms or pressure to conform that mask potential differences.
The author emphasizes staying open to alternative explanations and ensuring your methodology can capture subtle effects or interactions that might explain unexpected results.
The broader point: unexpected results can be highly informative and contribute to the literature when analyzed carefully.
Page 97
Summing Up: Theory and Inference
Theory-building is essential for making strong causal inferences; it guides expectations, hypotheses, and research design.
Theory helps identify the most relevant variables and the controls necessary to strengthen inference.
Theory clarifies the causal mechanisms behind findings and demonstrates how the research contributes to general knowledge about political processes.
In short, theory is a fundamental building block of political science research.
Key terms introduced (as a quick reference):
concepts, falsifiability, generalizations, hypotheses, literature review, observable implications, theory.
Page 98
A Menu of Approaches (Section II) – Transition to further material.
Page 99
This page intentionally left blank.
Cross-page connections and practical takeaways:
Build theory with testable, concrete hypotheses; avoid vagueness that impedes falsification.
Treat the literature review as a tool to justify your theory, define variables, and map rival theories, while guiding the reader through your research design.
Use the literature to frame the big questions, but emphasize how your study advances theory rather than merely cataloging prior work.
Be mindful of co-authorship and plagiarism issues; maintain rigorous note-taking practices and cite sources appropriately to contribute to a collaborative scholarly conversation.
Recognize that unexpected or null results are not failures but opportunities for deeper theoretical refinement and methodological improvement.
Practical formulaic reminder: hypotheses can often be framed as if–then statements to structure empirical tests and their implications.
Here are the key takeaways from the provided notes:
Theory and Hypotheses: Effective theory building involves formulating specific, testable hypotheses (often as if–then or when–then statements) that can be empirically evaluated. Avoid broad, vague definitions that hinder falsification.
Role of the Literature Review: The literature review serves as a crucial bridge between prior research and your study. It should justify your theory, connect key problems and data, delineate discussions and debates, acknowledge plausible rival theories, and present your own working hypotheses. It's a tool to lead the reader through your research design, not just a catalog of prior work.
Integrating Theory into Research: Theory guides concept definition, variable selection, research design, and the interpretation of findings. It clarifies causal mechanisms and helps select appropriate controls to strengthen causal inference.
Iterative Nature of Research: Your research question, theory, and literature review are dynamic and subject to refinement as you learn more. Be prepared to iterate and evolve your thinking throughout the project.
Handling Unexpected Results: Null or unexpected findings are not failures but valuable opportunities for deeper theoretical refinement and methodological improvement. They prompt re-examination of assumptions, models, and data, and may reveal omitted information or missing theoretical considerations. Actively test alternative theories.
Scholarly Practices: Maintain rigorous note-taking to avoid plagiarism, ensure proper citation throughout your paper (not just in one section), and be mindful of co-authorship norms. Emphasize empirical findings or theoretical insights as focal points, rather than authors.