Page 15 - Age of onset of disruptive behavior of residentially treated adolescents -Sjoukje de Boer

Methodological complicating factors 1 There were several complicating factors to be dealt with in the operationalization of the LCP and AL concepts in a clinical context. On average, youngsters are 17 years old when they are admitted to De Fjord, making it hard to distinguish the LCP and AL groups on the presentation of the disruptive behavior at
that time, as both groups were shown to present similar in frequency and seriousness
of offending during adolescence (Moffitt et al., 1996).
In general, clinically admitted individuals are not followed from birth and
therefore the classification had to be made based on information collected retrospectively (e.g., information from files or recollected information). Also, since the course of the participants’ future disruptive behavior was as yet unknown, the terms LCP and AL could not be used. For this reason, in the present thesis the term early- onset (EO) was used for adolescents admitted to the facility that currently displayed severe disruptive behavior, and who also had shown behavioral problems during childhood. They were expected to be at increased risk of heading for the LCP pathway. Adolescents with current severe disruptive behavior who had not displayed disruptive behavior during childhood, were labeled as adolescent-onset (AO). Presumably, they would develop as those in the AL group, even though it is not guaranteed that AO is equivalent to AL. However, it has been found that some retrospective measures (e.g., psychosocial variables) have low levels of agreement with prospective measures, which directly affects the reliability (Henry, Moffitt, Caspi, Langley, & Silva, 1994). It cannot be ruled out that some adolescents classified in the AO group have not exhibited early-onset disruptive behavior (see e.g., Van Domburgh, Vermeiren, Blokland, & Doreleijers, 2009; Vermeiren, 2003). This may be related to the fact that retrospective classification increases the probability of a recall bias. In this light, several studies have questioned the usefulness of the developmental theory for clinical practice (Fairchild, Van Goozen, Calder, & Goodyer, 2013; Fernández-Montalvo, López- Goñi, Illescas, Landa, & Lorea, 2008; Vermeiren, 2003). However, the level of agreement does seem to depend on the type of behavior that is exhibited . The agreement on the age of onset between prospective and retrospective measures of delinquency for instance may vary depending on the seriousness of the offences (Kazemian & Farrington, 2005). In the present thesis, these concerns were obviated to
13