|Home | About | Journals | Submit | Contact Us | Français|
Data from 480 short-term general nonfederal hospitals are analyzed by means of the Automatic Interaction Detector program in an attempt to discover manipulable factors influencing the innovation of programs contributing to comprehensiveness of health care. The data are divided into three sets of independent variables pertaining to contextual (locational) and organizational characteristics of hospitals and to personal characteristics of hospital administrators. The dependent variable is an overall implementation score for each hospital, assigned on the basis of a scale of rankings of a hypothetical service by a panel of experienced hospital personnel. The AID program eliminates variance in overall implementation attributable to the nonmanipulable contextual and organizational variables, permitting identification of residual variance arising from administrator characteristics.