A Framework for the Analysis of Data: Conceptualization, Measurement and Aggregation (From Conceptualizing and measuring Democracy by Munk & Verkuilen, 2002, pg 8)
Challenge
|
Task
|
Standard of Assessment
|
Conceptualization
|
Identification of attributes
Vertical organization of attributes by level of abstraction
|
Concept Specification: Avoid maximalist definition (the inclusion of theoretically irrelevant attributes) or minimalist definition (the exclusion of theoretically relevant attributes)
Conceptual logic: Isolate the “leaves” of the concept tree and avoid the problems of redundancy and conflation
|
Measurement
|
Selection of indicators
Selection of measurement level
Recording and publicizing of coding rules, coding process, and disaggregate data
|
Validity: Use multiple indicators and establish the cross-system equivalence of these indicators; use indicators that minimize measurement error and can be crosschecked through multiple sources
Reliability
Validity: Maximize homogeneity within measurement classes with the minimum number of necessary distinctions
Reliability
Replicability
|
Aggregation
|
Selection of level of Aggregation
Selection of aggregation rule
Recording and publicizing of aggregation rules and aggregate data
|
Validity: Balance the goal of parsimony with the concern underlying dimensionality and differentiation
Validity: Ensure the correspondence between the theory of the relationship between attributes and the selected rule of aggregation
Robustness of aggregate data
Replicability
|
No comments:
Post a Comment