I M Katarina Swahnberg1, Barbro Wijma. 1. Division of Women's Health, Department of Molecular and Clinical Medicine, Faculty of Health Sciences, Linköping University, Linköping, Sweden. katsw@imk.liu.se
Abstract
BACKGROUND: In the literature about abuse, large variations in prevalence rates exist. Validated research instruments are scarce and are needed urgently. Our aim was to validate the 13 questions concerning the experiences of abuse among women in the NorVold Abuse Questionnaire against an interview and two validated questionnaires. METHOD: Data collection was in two parts. i) The NorVold Abuse Questionnaire was sent to a random sample of 2000 women in Ostergötland. ii) A subsample of 64 women was interviewed, and filled in the Conflict Tactic Scale, the Sexual Abuse Questionnaire, and the NorVold Abuse Questionnaire for a second time. The interview had open questions about abuse and was considered our gold standard. RESULTS: The response rate was 61%. The abuse variables in The NorVold Abuse Questionnaire showed good test-retest reliability (84-95%). Specificity was 98% for all kinds of abuse except physical (85%). Sensitivity ranged from 75% (emotional) to 96% (physical). The likelihood ratio ranged from 38 to 43 for all kinds of abuse except physical (likelihood ratio 6). NorAQ performed better against the interview than against the Sexual Abuse Questionnaire and equally against the Conflict Tactic Scale. High lifetime prevalence rates of abuse were found: emotional 21.4%; physical 36.4%; sexual 16.9%; abuse in the health care 15.6%. Prevalence rates of abuse dropped considerably when a criterion of current suffering was added. CONCLUSIONS: The abuse variables in NorAQ have good reliability and validity.
BACKGROUND: In the literature about abuse, large variations in prevalence rates exist. Validated research instruments are scarce and are needed urgently. Our aim was to validate the 13 questions concerning the experiences of abuse among women in the NorVold Abuse Questionnaire against an interview and two validated questionnaires. METHOD: Data collection was in two parts. i) The NorVold Abuse Questionnaire was sent to a random sample of 2000 women in Ostergötland. ii) A subsample of 64 women was interviewed, and filled in the Conflict Tactic Scale, the Sexual Abuse Questionnaire, and the NorVold Abuse Questionnaire for a second time. The interview had open questions about abuse and was considered our gold standard. RESULTS: The response rate was 61%. The abuse variables in The NorVold Abuse Questionnaire showed good test-retest reliability (84-95%). Specificity was 98% for all kinds of abuse except physical (85%). Sensitivity ranged from 75% (emotional) to 96% (physical). The likelihood ratio ranged from 38 to 43 for all kinds of abuse except physical (likelihood ratio 6). NorAQ performed better against the interview than against the Sexual Abuse Questionnaire and equally against the Conflict Tactic Scale. High lifetime prevalence rates of abuse were found: emotional 21.4%; physical 36.4%; sexual 16.9%; abuse in the health care 15.6%. Prevalence rates of abuse dropped considerably when a criterion of current suffering was added. CONCLUSIONS: The abuse variables in NorAQ have good reliability and validity.
Authors: Anna R Laurie; John Showalter; Toya Pratt; Noel H Ballentine; Vernon M Chinchilli; Jennifer S McCall-Hosenfeld Journal: Women Health Date: 2012