论文部分内容阅读
本文对1985～1986年“中国农村及少数民族地区神经系统疾病流行病学调查”所得到的631对脑血管病的1∶1配对资料,利用计算机进行了进一步的多因素的统计分析;筛选了危险因素,配合了Logistic回归模型,所得出的各危险因素的回归系数及标准化的回归系数,可用以估计各危险因素各自的及联合的相对危险度,对模型贡献的大小,本文也计算了理论频数、概率与实际频数、频率,并列表加以比较,以评估模型的拟合情况。本文还分别对出血性脑血管病和缺血性脑血管病及其危险因素配合了类似的Logistic回归模型。全部计算工作在VAx/750计算机上完成。
In this paper, from 1985 to 1986, “China Epidemiological Survey of Nervous System Diseases in Rural and Minority Areas” obtained 631 pairs of cerebrovascular disease 1: 1 matching data, the use of computer for further multi-factor statistical analysis; screening Risk factors, combined with the Logistic regression model, the regression coefficient of each risk factor obtained and the standardized regression coefficients can be used to estimate the relative risk of each risk factor and joint, the size of the contribution to the model, this article also calculated the theory Frequency, probability and the actual frequency, frequency, and the list to be compared to assess the fitting of the model. In this paper, a similar Logistic regression model was also applied to hemorrhagic cerebrovascular disease, ischemic cerebrovascular disease and their risk factors respectively. All calculations are done on the VAx / 750 computer.