The relationship between well-being and commuting revisited: Does the choice of methodology matter?