Measuring Risk Aversion: An Alternative Approach