Psychological biases in time-intensity scaling