Developing a Research Project
The Elements of Research Design
Three Purposes of Research Exploration –
Generally not sufficient for dissertation project
Description – –
This can be the goal if significant new facts or relationships are identified Usually need to at least speculate on explanation
Explanation –
Central goal of most political science research
Getting a Project Started Identify
a topic or area of interest Key transition is from topic to question – – –
Identify puzzle in literature Should be able to state question in 1-2 sentences Clearly define the purpose of the project
Once
you have a question, you need hypotheses about the answer
From Theory to Hypotheses Once
you have a question, turn to various methods of theory development Deduction – formal theory, computational models Induction – building by analogy Theory building often involves both methods
Choose a Unit of Analysis Level
of aggregation at which your hypotheses can be observed Individuals (citizens, leaders, etc.) Groups (parties, social movements) Organizations/Institutions (states, bureaucracies, firms)
Units of Analysis and Threats to Inference Ecological –
Fallacy
Cannot infer individual behavior from collective outcomes
Reductionism –
Cannot infer collective outcomes from individual behavior
Choose a Method of Research Experiments Surveys Case – –
Studies
Field Research Archival Research
Aggregate –
Data Analysis
Existing or Field Collection
Use
Multiple Methods Whenever Possible
Select Observations What
is the population to which you want to generalize? How can you reach that population to draw a sample? If random sampling is impossible or inappropriate, how do you select cases to avoid bias?
Operationalize Variables Translate
theoretical concepts into observable information that can be gathered on units of analysis How valid are your measures? How
reliable are your measures?
Collect the Data Be
clear and systematic about the type of data collected Keep clear files and records of all data collected If using quantitative data, keep “do files” for all data construction and analysis.
Analyze the Data Make
analysis techniques transparent To the greatest extent possible, make analyses easily replicable. Think about presenting results in the most meaningful way possible. Clear and meaningful results are more persuasive and more widely read.
Connect Analyses Back to Puzzle and Hypotheses Always
draw readers back from analyses to their implications for hypotheses and your original puzzle What does it all mean? Try to get back to those one or two core sentences
What is a Good Dissertation? Theory –
Is this alone enough for a dissertation?
Theory – –
Testing
This is the goal of most projects Is it ESSENTIAL for a dissertation?
Policy –
development
Evaluation
Is this any different from theory testing?
What is a Good Dissertation? Historical –
When does this constitute a contribution to social science?
Literature – –
Evaluative/Descriptive
Assessing
A viable strategy for dissertations? When does this constitute a contribution to social science?
What is a Good Dissertation? Nomothetic
vs. Ideographic Explanation Is one mode of explanation preferable to the other? Is one mode of explanation inherently more “scientific?” Can ideographic work be political science?