We appreciate Professor Gorman's interest in our research. The sample described in Lonczak et al1 was constituted in the fall of 1985 when panel participants entered the fifth grade, as described in detail in Hawkins et al.2 The full-intervention condition was created by assigning those participants remaining in schools that had participated in social development interventions as part of a prior experimental study to receive intervention in grades 5 and 6. Criteria for the full-intervention group are exposure to the intervention for at least 1 semester in grades 1 through 4 and for at least 1 semester in grades 5 or 6. Professor Gorman notes that in the prior experimental study, the intervention group included 285 participants in grade 2. He suggests that this should have been the full-intervention group sample. This is incorrect. Students who left participating schools prior to grade 5 did not receive at least 1 semester of intervention in grades 5 or 6 and, therefore, did not qualify for the full-intervention condition. They were not eligible for the study sample constituted at grade 5. Of the consenting fifth-grade students assigned to the full-intervention condition at the beginning of this study, 144 (92.3%) were successfully interviewed at age 21 years. Thus, subject attrition was not high, as suggested by Professor Gorman.
Hawkins JD, Lonczak HS, Abbott RD, Kosterman R, Catalano RF. Overstating the Behavioral Effects of the Seattle Social Development Project. Arch Pediatr Adolesc Med. 2002;156(11):1155-1156. doi: