Ta. Wills et Sd. Cleary, Peer and adolescent substance use among 6th-9th graders: Latent growth analyses of influence versus selection mechanisms, HEALTH PSYC, 18(5), 1999, pp. 453-463
This study analyzed peer-influence versus peer-selection mechanisms in adol
escent tobacco, alcohol, and marijuana use. Participants were surveyed 3 ti
mes, with I-year intervals, about peers' substance use and their own use; S
ample 1 had 1,190 participants (initial mean age = 12.4 years), Sample 2 ha
d 1,277 participants (initial mean age = 11.5 years). Latent growth analyse
s that were based on composite scores indicated that initial peer use was p
ositively related to rate of change in adolescent use, supporting the influ
ence mechanism; there was little evidence for a selection mechanism. Diffic
ult temperament, poor self-control, and deviance-prone attitudes were relat
ed to initial levels for both peer and adolescent use. It is concluded that
peer influence is the primary mechanism during middle adolescence. Tempera
ment-related attributes may be predisposing to early experimentation and de
viant-peer affiliations.