The null hypothesis (a,b) are that, for large and small hospitals, there is no difference between the distributions of their separate efficiency measures and the distributions of their pooled efficiency measures. Results in Table 3 indicate that these two null hypothesis cannot be rejected at the 95% level of confidence. These suggest that both large and small hospitals have their own separate frontiers that cannot be distinguishable from the pooled frontier. One likely explanation is that the two hospital groups have access to and practice the same production technology