Research Design Jan Osička Lecture outline • Research typologies • Theory: what is it (good for)? • Research strategies: from data to theory and back • Designing social research • Sample research design Research typologies Which typologies do you know? Scope Parsimony (~ reductionism) • Limited scope and already known structure Holism • Wide scope and yet-to-emerge structure The role of researcher Subjective • Researcher-induced interpretatitions are valid Objective • Only rigorous data interpretations are valid Causality Logic of explaining • Causal relationships • Event „cause“ =(time and mechanism)> event „effect“ => general knowledge Logic of understanding • Understanding of how certain effect has emerged => case-specific knowledge Units • Cases (spatially delimited phenomenon observed at a single point in time or over a period of time; Gerring 2007:19) • Variables (concepts that can have various values; Van Evera 1997:10) The N Qualitative • Small N (up to 10) Quantitative • Large N (100 and more) Direction Induction • Generalizing from available data Deduction • Using general knowledge to interpret data Inductive research strategy I Inductive research strategy II Inductive research strategy III Inductive research strategy IV Deductive research strategy I Deductive research strategy II Theory „A set of interrelated constructs (concepts), definitions, and propositions that present a systematic view of phenomena by specifying relations among variables with the purpose of explaining and predicting phenomena.“ Fred Kerlinger, 1986:9 Two meta-typologies Parsimonious Deductive Objective Variable-oriented Quantitative Holistic Inductive Subjective Case-oriented Qualitative Research design (Blaikie 2011) • Research topic • Research question • Literature review • Theory • Method • Data • Problems and limitations Conducting a research Phase Focus on Preliminary Background knowledge Preparatory Research topic Research question Literature review Theoretical assumptions Methodology Data overview Analysis Data collection Data processing Results interpretation Reporting Research report Preparatory phase • Draws heavily from the substantial knowledge • Continuous process • Results in final version of research design Preparatory phase Preliminary Background knowledge Preparatory Research topic Literature review Research question Theory Method Data Research design Research report • Does not need to follow the structure of the research • Shall include only information directly related to the aim and scope of the research. Research report • Annotation • Introduction (Research topic & Research question) • Literature review • Theory • Method & Data • Analysis • Interpretation • Conclusions Sample research design Ross, M. L. (2001), Does oil hinder democracy? World Politics 53. (http://www.maxwell.syr.edu/uploadedFiles/exed/sites/ldf/Academic/ Ross%20-%20Does%20Oil%20Hinder%20Democracy.pdf) Research topic • The relation between oil rent and the nature of regime. • Common assumption: rising income implies more democracy • However: democratizing effect shrinks or disappears if rising income comes from oil. => Oil as an exception? Research questions • Does oil wealth impede democracy? • If yes, • Does it happen only in the Middle East or is it valid for all oil-rich states? • Is it oil‘s unique property or is it valid also for non-fuel minerals? • If yes, what is the causal mechanism? • Rentier effect • Repression effect • Modernization effect Literature review (Pp 329-337) • Oil – minerals – democracy nexus • The „Rentier state“ concept Theory The Rentier state concept (Mahdavy 1970) • Rentier effect • Repression effect • Modernization effect Method • Statistical analysis (Feasible Generalized Least Squares method) • Dependent variable: Regime • Independent variable: Oil and minerals (fuel/non-fuel) • Control variables: Income, Islam, OECD membership, Regime (t-5) • Dummy variables: 26 for each year (1971-1997) to capture two timespecific effects: the cold war, the „contagion“ of regime transitions Indicators and data (Pp 358-361) • Regime: indexes of Polity98 (Polity III, Polity IV; Freedomhouse) • Oil and Minerals: export value as fractions of GDP (Sachs and Warnes 1995, Leite and Weidmann 1999) • Income: GDP PPP (World Bank, CIA World Factbook) • Islam: % of Muslims in state‘s population (Barrett 1982; CIA World Factbook) • OECD: 1/0 (OECD info) • Regime (t-5): indexes of Polity98 Results Total: 2,183 observations from 113 countries »Oil does hinder democracy »In poor states more »In oil-poor states more »This is not limited to the Middle East »Nonfuel minerals hinder democracy too »Oil wealth affects the regime through all three mechanisms; mineral wealth mostly through rentier effect and not at all through repression effect