Hypothesis Construction. From Concept to Measurement In order

Hypothesis Construction

From Concept to Measurement In order to empirically test the relationships between concepts, we must achieve some difficult goals All concepts are complex and variable; as such, they must be broken down into differential components and their variability must be made measurable Many concepts are not directly observable; given that measurement is key to testing, we must find a way to utilize observable phenomena that reflect unobservable concepts

Operationalization Stage 1: Breaking Down Concepts Example: We have theoretical reasons to believe that democracy is causally important in many processes. But what, conceptually, is democracy? Linked to our theory, it probably involves: Participatory capacity Insurance of civil liberties Presence of a civic culture Contestability of political office Separation of powers By breaking down concepts, we are able to: Comprehensively identify the specific “underpinnings” of our theoretical expectations Generate initial expectations about what we must measure to gain a comprehensive reflection of democracy

Operationalization Stage 2: Matching Conceptual Components to Observable Phenomena Now that we’ve gotten an idea of what democracy consists of, what might its components look like in the real world? Participatory capacity: The ability of citizens to participate in the selection and workings of their government Insurance of civil liberties: The governmental guarantee of individual freedoms Presence of a civic culture: Difficult – is it the willingness of the citizenry to take advantage to participatory opportunity? To respect the civil liberties of others? To actively agree on the legitimacy of government? All of the above? Contestability of political office: The capacity of any citizen to obtain office Separation of powers: Structural division of governmental responsibilities and limitations on the capacity of any subdivision to encroach on the responsibilities of others Basic translation of conceptual components into generally observable phenomena

Operationalization Stage 3: The Development/Selection of Useable Indicators How are we to empirically evaluate these phenomena? We select measures that are variable, regularly observable, and reflective of as many stages of our conceptualization process as possible Participatory capacity: Presence of laws protecting voting rights Insurance of civil liberties: Presence of laws eliminating (or minimizing) the capacity of the government to regulate or violate individuals Presence of a civic culture: Voter turnout; political association membership; frequency of political assembly Contestability of political office: Absence of unreasonable governmental restrictions on office-holding Separation of powers: Presence of laws mandating and enforcing separation of governmental responsibilities (including oversight) Improvement of generally observable phenomena into classifiable, measureable indicators

Measurement Characteristics Levels of Measurement Nominal: Categorization and numerical assignment for labeling 0=Presidential Democracy, 1=Parliamentary Democracy; no assumption of ranking Ordinal: Categorization and numerical assignment for comparison (greater/lesser) 1= Undemocratic, 2=Quasi-Democratic, 3=Fully Democratic; mathematical rank ordering of outcomes Interval: Intervals between measures represent equal distance across the entire scale Ratio: Includes features of other levels plus an absolute reference point Voter turnout rate; reference to both other levels and overall population

Variable Reliability and Validity We must pay particular attention to the “disconnect” between concepts and phenomena Reliability: Is our operational definition measuring something consistently and dependently? Ascertained through similar findings across different samples internal consistency assessments (splitting samples) Validity: Does our operationalization perfectly “fit” our concept (rare)? Must be aware of measurement error and whether it occurs randomly or systematically