Revalidating a survey instrument for measuring risk preferences
重新验证了全球偏好调查中测量风险偏好的工具，发现其定量和定性问题能预测中国、美国和德国样本的风险厌恶，且两者结合优于单独使用。
• We revalidate the instrument used for measuring risk preferences in the GPS. • The quantitative and qualitative items predict risk aversion in three samples. • The combination of both items is preferable to using only one item. • Considering quantitative and qualitative items separately can be informative. Survey instruments are often a cost-effective alternative to eliciting preferences via incentivized experiments. We revalidate a survey instrument used for measuring risk preferences in the influential Global Preferences Survey. We find that the instrument consisting of a quantitative and a qualitative item predicts risk aversion in a Chinese, an American and a German sample. For out-of-sample predictions, the combination of both items is preferable compared to using only one of the two items.