The size hypothesis was not supported. The ignored naive hypotheses and other hypotheses of Ball
and Foster (1982) still continue to serve as alternative hypotheses for the subsequent appeals by Watts and
Zimmerman (1986; 1990). Finally, it can be seen in Table 2 that only 3 of the 9 studies on the agency
hypotheses given by Watts and Zimmerman (1986) yield what is generally accepted as a statistically
significant size hypothesis, even though these studies were given in response to critics, and were later
acclaimed for their empirical 'regularities.'