基于平衡单水平轮换的连续性抽样估计方法研究

内容提要：针对现存的各种单水平轮换模式和估计方法,本文提出一套统一的平衡单水平轮换模式。在此轮换模式下,引入两类相关关系,运用线性无偏估计方法,并通过使不同类型估计量方差的加权总和最小的方法确定最优系数,从而得到最优线性无偏估计量,不仅能够减少甚至消除估计量偏差的影响,还能使得连续性调查的整体抽样误差最小,适合估计各种类型的估计量。     

环境关心的测量：NEP量表在中国的应用评估

本文检视了西方学者（特别是美国学者）自20世纪70年代以来测量公众环境关心的历程,分析了广为使用的NEP量表的提出及修订,并重点基于2003年中国综合社会调查（城市部分）资料,对修订过的NEP量表在中国的应用效果进行了评估。作者指出,不加改造地在中国应用修订过的NEP量表,效果并不理想。但是,经过适当改造之后,该量表可以作为测量公众环境关心的重要工具。     

国际合作计划生育结合项目效果初探—以居民意识问卷调查结果为依据

根据计划生育结合项目 3个周期居民意识问卷调查结果 ,对 90年代以来的结合项目实施效果进行了分析。通过和对照乡镇的对比可以看出 :项目乡镇在计划生育、妇幼保健、寄生虫预防、健康教育等几个方面都有较大的变化 ,项目获得了预期的效果 ,对当地的计划生育工作有很大的促进作用     

农村集体成员权制度运行状况的实证分析——基于全国12省36县的实地调查

通过对全国12省36县进行田野调查,对农村集体成员权制度在农村社会运行状况进行统计分析,以弄清其实际运行状况,并结合我国相关法律制度与访谈材料做深入研究.结果显示:户籍确认标准已与社会现实之间存在着一定程度的张力;成员资格确认标准的严格程度与集体经济发达水平之间呈正相关;农户对成员权应然层面的主观认知与集体经济发达水平之间具有交互性影响;成员权的实际享有程度较高,但集体利益诉讼权的享有还有待强化;农户对法律明确规定成员资格丧失情形的认可度很高,且乡土社会内在逻辑深深影响着农户对成员资格丧失之应然情形的主观认知.     

Preference for caesarean and attitudes toward birth in a Chilean sample of young adults

BackgroundLittle empirical research exists about what motivates birth mode preferences, and even less about this topic in Latin America, where obstetric interventions and caesareans are some of the highest worldwide.AimTo identify factors associated with caesarean preference among Chilean men and women who plan to have children and to inform childbirth education and informed consent procedures.MethodsAn online cross-sectional survey measuring attitudes toward birth was administered to graduate students at a large public university in Chile. Eligible students were under the age of 40 and had no children but intended to have children. Logistic regression modelling was used to determine which sociodemographic factors, knowledge and beliefs were associated with caesarean preference.FindingsAmong eligible students, 730 responded and 664 provided complete answers to the variables of interest. Respondents had a mean age of 28.8; 38% were male and 62% female. Positive attitude toward technological intervention (Odds Ratio 7.4, 95% Confidence Interval 3.9–14.0), high risk perception of vaginal birth (Odds Ratio 1.8, 95% Confidence Interval 1.1–2.8), family history of caesarean (Odds Ratio 1.9, 95% Confidence Interval 1.0–3.8) and high fear of birth (Odds Ratio 3.7, 95% Confidence Interval 2.0–6.8) were associated with caesarean preference.DiscussionPreference for caesarean birth was highly associated with positive attitudes toward technological intervention and may be related to a lack of knowledge about the realities of caesarean and vaginal birth.ConclusionsPatient-centered education on the relative benefits and risks of birth modes has the potential to influence preferences toward vaginal birth.     

Midwives’ views of changing to a Continuity of Midwifery Care (CMC) model in Scotland: A baseline survey

BackgroundThere is good evidence that Continuity of Midwifery Care (CMC) is associated with improved clinical outcomes, greater maternal satisfaction, and improved work experiences for midwives. Changes made to the organisation require careful implementation, with on-going evaluation to monitor progress.AimTo develop a survey tool that incorporates several validated scales, which was used to collect baseline data prior to implementing a high-quality Continuity of Midwifery Care (CMC) model in Scotland (Hewitt et al., 2019). This tool gathered data about midwives’ personal and professional wellbeing prior to service reorganisation, with a longitudinal study intended to measure change in midwives’ reportage across time. This paper reports the baseline data-collection.MethodsAn on-line survey was shared with practising midwives (n = 321) in Scotland via the NHS intranet, verbally, email, and paper. The survey elicited midwives views about Continuity of Midwifery Care (CMC); values and philosophies of care; attitudes towards their professional role; personal and professional demographics; quality of life and wellbeing. Psychometric attitudinal scales were scored and free text comments themed according to positive/negative opinions of the new Continuity of Midwifery Care (CMC) model to highlight key concerns to be addressed and identify change barriers or facilitators.FindingsThe majority of midwives indicated support for philosophies underpinning Continuity of Midwifery Care (CMC), which includes physiological birth and providing autonomous midwifery care. Participants also indicated positive attitudes towards their current role and organisation, with some worrying about how the organisation was going to implement the changes required. Worries included, receiving an overburdening workload, being deskilled in certain areas of midwifery practice, and lack of support were litigation to arise.ConclusionMidwives support the values and philosophies that underpin Continuity of Midwifery Care (CMC), yet worry about organisational change involved in evolving systems of care. Hence, management require to implement strategies to reduce fears. For example, delivering accurate and honest information, enabling midwives to plan, design and implement changes themselves, and providing emotional and material help.     

有序响应变量的贝叶斯模型选择及其在COPD疾病防治中的应用

慢性阻塞性肺病(COPD)是一种发病率、死亡率都非常高的疾病，且COPD的诊断和严重程度分级依赖于肺功能的检查，但是由于肺功能检查仪器价格昂贵，使得这项检查在很多经济欠发达地区尤其是农村基层医院并没有普及。本文基于有序响应变量模型致力于研究一种便于基层和社区使用的可以初步判别COPD病情的模型，以期提高我国基层和社区的COPD 防治水平。利用贝叶斯变量选择方法和数据增强的潜变量策略得到了易于实施的Gibbs后验抽样算法。数值模拟分析进一步说明了本文提出的有序响应变量贝叶斯模型选择方法的有效性，实例分析得到了易于判别COPD严重程度的稀疏模型。     

大规模数据下基于充分降维的Leverage重要性抽样方法

随着信息技术的飞速发展，大规模数据在短时间内搜集并储存下来，为分析决策提供了巨大的信息量，也给统计建模带来了一定难度。对于样本容量大、变量个数少的数据，Leverage重要性抽样是一个简便可行的方法。本文发现，该方法中度量样本重要性的Leverage分数与因变量无关，而且在维度较大的情形下对样本没有区分程度，使得估计结果较差。为了同时考虑因变量和维度的影响，本文提出了基于充分降维的Leverage重要性抽样方法。该方法以不损失信息为前提，在充分降维的空间内重新计算Leverage分数，使得抽样更具有代表性。模拟数据分析显示，在样本容量较大的复杂数据中，相比于原始的Leverage重要性抽样方法，本文提出的方法可以降低估计的均方误差。三个实际数据也证实了该方法的可行性和有效性。     

Discriminant analysis of survey data

We consider the problem of the effect of sample designs on discriminant analysis. The selection of the learning sample is assumed to depend on the population values of auxiliary variables. Under a superpopulation model with a multivariate normal distribution, unbiasedness and consistency are examined for the conventional estimators (derived under the assumptions of simple random sampling), maximum likelihood estimators, probability-weighted estimators and conditionally unbiased estimators of parameters. Four corresponding sampled linear discriminant functions are examined. The rates of misclassification of these four discriminant functions and the effect of sample design on these four rates of misclassification are discussed. The performances of these four discriminant functions are assessed in a simulation study.