When do individuals choose care robots over a human caregiver? Insights from a laboratory experiment on choices under uncertainty