Is signal detection theory fundamentally flawed? A response to Balakrishnan (1998a, 1998b, 1999)