Abstract
Separation anxiety continues to be implicated as an early risk factor to adult emotional disorder but recent research findings are somewhat contradictory. Inconsistencies in approaches to measuring memories of early separation anxiety may have contributed to this lack of clarity. We report the development of a brief self-report instrument, the Separation Anxiety Symptom Inventory (SASI), which was designed to overcome some of these deficiencies in measurement. The SASI was shown to have a coherent factorial structure, high internal consistency (Cronbach's Alpha>.80) and test-retest reliability over an average of 24 months (Intraclass Correlation Coefficient =. 89), with serial scores not being affected by changes in contemporaneous anxiety levels. Some index of the validity of the measure was achieved by (a) comparing SASI scores of index twins with descriptors of their “insecure’ behaviours in early life provided by corresponding co-twins; (b) comparing SASI scores with retrospective DSM III-R diagnoses of early anxiety disorders obtained by structured interviews; and (c) examining SASI scores in subjects with histories of school refusal. The SASI provides a useful standardised measure which will aid in the further testing of the separation anxiety hypothesis of adult emotional disorder.