Evaluation of large-scale programs is problematical because of inherent bias in assignment of treatment and control groups, resulting in serious regression artifacts even with the use of analysis of covariance designs. Nonuniformity of program implementation across sites and classrooms is also a problem. (Author/GSK)

A neurological examination has been developed to discover children with physiologically based learning problems who do not manifest asymmetrical functioning. This study attempts to determine the validity of this examination by its accuracy in predicting the performance of children in a summer Head Start program. Validity was determined by…