Residents' knowledge of behavioral pediatrics.

Four groups of residency programs (4433 residents) were compared regarding performance on the In-Training Exam of the American Board of Pediatrics and a 30-item addendum assessing behavioral knowledge: 10 programs funded to provide mandatory training, 6 not funded but requiring training, 4 control programs not requiring training but participating in an evaluation of behavioral training, and all other programs (221) whose residents took the exam. Three measures were calculated for each resident: percent correct on "organic" items; percent correct on "behavioral" items; and a ratio (behavioral to organic). Moderate positive correlations were found between organic and behavioral scores. There were no significant differences in organic performance related to program type. Ratio scores were higher for PL-1-funded residents than for any other PL-1 group. For PL-2s, Funded, Not Funded, and Control residents performed better than the All Other group. By the 3rd year, Funded residents performed better than All Other residents; Not Funded and Control scores fell in between. These data suggest that: (1) behavioral knowledge is only moderately related to organic knowledge; (2) requiring behavioral training increases behavioral knowledge but does not decrease the level of organic knowledge; and (3) funded programs attract residents more knowledgeable about behavior who, during their training, remain more knowledgeable than residents in programs not participating in an evaluation of behavioral training.