On the need for longitudinal evidence and multiple measures in behavioral-genetic studies of adult personality