Page 84 - Personality disorders and insecure attachment among adolescents

interviewers were trained to apply the AAI by experienced coders at the Dutch Psychoanalytic Institute in Amsterdam, The Netherlands. Interviews were audio-taped and transcribed for coding. A trained external coder, S. den Hollander, who is reliable since 2001 and trained by D. Pederson & D. Jacobvitz, rated the transcripts using the AAI Scoring and Classification System (Main et al., 1998). The AAI meets stringent psychometric criteria in terms of reliability, discriminant, and predictive validity and it can be used with adolescents (Bakermans-Kranenburg & Van IJzendoorn, 1993a, 2009; Cassidy, 2008; Hesse, 2008; H. Steele & Steele, 2008; van IJzendoorn, 1995). The inter-rater reliability of the Dutch version of the AAI (k = .61) (Bakermans-Kranenburg & Van IJzendoorn, 1993b) qualified as fair (Landis & Koch, 1977). For the purpose of statistical analyses, a continuous scale ranging from one to nine was constructed for both the state-of-mind AAI scales and the experiences toward parents AAI scales.
Procedures
All 67 of the newly admitted adolescents were asked to participate in the study during a four- year period (2008–2012). Following a verbal explanation of the treatment protocol to the subjects, written informed consent was obtained according to legislation, the institution’s policy, and Dutch law (Eurec, 2017). All patients (N = 60) agreed to participate, and, in concordance with the institutional policy, they participated without receiving any incentives or rewards. All procedures in this study were aligned with the 1964 Helsinki declaration and its later amendments, or with comparable ethical guidelines. According to the treatment protocol, the patients completed a set of web-based questionnaires in the first and last weeks of treatment, after which they participated in the SCID-II interview, and, finally, in the AAI interview. This order in the treatment protocol resulted in many missing AAI assessments, mainly because adolescents were not easily committed to a long diagnostic process. In addition, the research process was sometimes obstructed by patient crises. Altogether 60 SCID-II interviews were conducted with patients in combination with the AAI interview, and 33 post- AAI interviews.
Statistical analysis
All analyses were performed using the Statistical Package for the Social Sciences (SPSS), version 23.0 (IBM Corp, 2011). In the first and observational, cross-sectional part of this study, chi- square tests were performed to compare the categorical variables of the AAI in the sample to norm groups. Next, based on the SCID-II, three groups were formed based on the type of a personality
78