How Good Are Corruption Measures? A Practical Rubric for Data Quality

🔍 The Problem and Purpose

Political scientists frequently struggle to judge whether measures are valid and reliable enough for substantive research. Stand-alone assessment tools exist, but they are rarely combined in practice, and guidance aimed at data consumers—those who must decide whether to use existing measures—is limited. This article presents a practical, integrated approach for assessing measure quality and demonstrates it on prominent corruption indicators.

🔧 A Three-Part Quality Rubric

A multimethod assessment integrates complementary tools to evaluate measure quality across three targeted dimensions:

• Content validity — whether indicators capture the intended concept fully and appropriately.
• Data-generation validity and reliability — whether the process that produced the data yields consistent, trustworthy measurements.
• Convergent validity — whether the measure aligns with other established indicators that should, in theory, converge.

🧪 Applied to V-Dem Corruption Measures

The rubric is applied to corruption measures from the Varieties of Democracy (V-Dem) project to both illustrate the assessment process and to evaluate those measures in context. The application:

• Demonstrates how each component of the rubric is operationalized using complementary methods.
• Identifies several quality advantages and disadvantages of V-Dem’s corruption measures when compared to other existing corruption measures.

📌 Why This Matters

This integrated approach gives data consumers concrete steps to judge whether existing measures are suitable for substantive analysis and highlights trade-offs among alternative indicators. The rubric is intended to help researchers make better-informed choices about measurement and to foster clearer communication between data producers and users.