List vs Endorsement Experiments Converge in Afghanistan: New Statistical Methods Validate Findings

This paper compares list experiments (LEs) and endorsement experiments (EEs), two indirect survey techniques for sensitive questions. We show these methods yield compatible findings even when studying delicate topics like support for ISAF among Pashtun respondents in challenging environments such as Afghanistan.

Our approach involves developing a statistical test and multivariate regression models to combine LE results with EE data effectively. This demonstrates that both approaches can produce accurate measurements of sensitive political attitudes.

🌟 Statistical Approach 🌟

We propose novel methods for integrating findings from different indirect question formats:

• New statistical tests comparing experiment outcomes
• Multivariate regression models combining list and endorsement responses

💼 Afghan Context 💼

Our study uniquely demonstrates methodological convergence in one of the most difficult survey environments globally.

🔍 Key Findings 🔍

Both LEs and EEs reveal consistent patterns regarding:

• Support for International Security Assistance Force (ISAF)
• Measurement validity across sensitive political topics
• Potential to enhance research credibility through multiple methods

📣 Takeaway 📣

This work confirms that combining different indirect question methodologies strengthens empirical conclusions about public opinion in conflict zones. The proposed techniques improve measurement reliability without sacrificing scientific rigor.