Heyman R E, Neidig P H
Department of Psychology, SUNY, Stony Brook 11794-2500, USA.
J Consult Clin Psychol. 1999 Apr;67(2):239-42. doi: 10.1037//0022-006x.67.2.239.
This study compared prevalence rates from reasonably representative U.S. Army (N = 33,762) and civilian (N = 3,044) samples. Age and race were controlled by weighting each sample to 1990 U.S. Census characteristics for married, full-time employed persons. Men's reports of moderate husband-to-wife spousal aggression were not significantly higher, but reports of severe aggression were significantly higher in the standardized Army sample than in the comparable civilian sample (adjusted rates of 2.5% vs. 0.7%, respectively). Thus, controlling for age and race results in reasonably similar prevalence rates. Future studies that more carefully control for any background differences in military and civilian respondents could discern whether military service adds any increased risk for partner violence.
本研究比较了具有合理代表性的美国陆军样本（N = 33,762）和平民样本（N = 3,044）中的患病率。通过将每个样本加权至1990年美国人口普查中已婚全职就业人员的特征，对年龄和种族进行了控制。在标准化的陆军样本中，男性报告的中度夫妻间配偶攻击行为发生率并未显著更高，但严重攻击行为的报告发生率显著高于可比的平民样本（调整后的发生率分别为2.5%和0.7%）。因此，控制年龄和种族后，患病率相当相似。未来的研究若能更仔细地控制军事和民用受访者的任何背景差异，或许就能辨别服兵役是否会增加伴侣暴力的风险。
