Predicting marital violence from the Marriage and Relationship Questionnaire: using LISREL to solve an incomplete data problem

Citation
Rjh. Russell et Pa. Wells, Predicting marital violence from the Marriage and Relationship Questionnaire: using LISREL to solve an incomplete data problem, PERS INDIV, 29(3), 2000, pp. 429-440
Citations number
13
Categorie Soggetti
Psycology
Journal title
PERSONALITY AND INDIVIDUAL DIFFERENCES
ISSN journal
01918869 → ACNP
Volume
29
Issue
3
Year of publication
2000
Pages
429 - 440
Database
ISI
SICI code
0191-8869(200009)29:3<429:PMVFTM>2.0.ZU;2-X
Abstract
Violence by husbands and wives directed towards the spouse, as measured by the Straus Conflict Tactics Scale (CTS), is predicted from 24 Marriage and Relationship Questionnaire (MARQ) scales, 12 from each partner. The analyse s use a sample of 80 complete cases. The information from this group of cas es is supplemented by 1211 incomplete cases who completed the MARQ but not the CTS. LISREL 8 combines the information from the two sources, using pseu do values for the missing information in the larger sample of incomplete ca ses. Violence by the husbands is predicted by the husbands' Roles, Love, Se xual Jealousy and Conciliation scales and by the wives' Family Ties, Attrac tiveness and Problems: Relationship scales. Violence by the wives is predic ted by the husbands' Sexual Jealousy and Conciliation scales and by the wiv es' Partnership, Love and Attractiveness scales. These findings are discuss ed and the method of combining partial data from a large sample with a smal ler number of complete cases is commended. (C) 2000 Elsevier Science Ltd. A ll rights reserved.