基于SAS软件实现随机抽样及应用

本文的目的是帮助读者方便快捷地运用SAS软件中的SURVEYSELECT过程实现随机抽样。首先,对SURVEYSELECT过程及SAS数据集Customers进行了简单介绍。接着,结合简单随机抽样、分层随机抽样和控制排序分层随机抽样,介绍了随机抽样的SAS实现方法。读者只需要修改本文中所呈现的SAS程序中的少量参数,就可很方便地使用SAS软件实现随机抽样任务。事实说明,尽管SAS软件非常难学难用,但借助现成的SAS程序,可以轻松自如地解决很多具体问题。     

复杂抽样调查设计二值资料一水平多重Logistic回归分析

本文目的是介绍复杂抽样调查设计二值资料多重logistic回归分析方法。通过一个实例,利用八种不同的分析策略(不考虑抽样设计和抽样权重、考虑抽样设计不考虑抽样权重、不考虑抽样设计考虑抽样权重、同时考虑抽样设计和抽样权重以及分别不考虑与考虑派生变量)对数据进行建模。对所得结果进行比较得出如下结论:在对复杂抽样设计资料进行统计分析的过程中,同时考虑抽样设计和抽样权重可以得到符合数据内部变量间依赖关系真实情况的结论。此外,本研究还介绍了采用SAS软件中SURVEYLOGISTIC过程对复杂抽样调查数据进行多重Llogistic回归分析的详细步骤。     

复杂抽样调查设计多值名义资料一水平多重Logistic回归分析

本文目的是介绍复杂抽样调查设计多值名义资料一水平多重logistic回归模型构建,并探讨不同策略之间的差异。采用SAS中的LOGISTIC过程和SURVEYLOGISTIC过程,分别按照是否考虑抽样设计与是否考虑抽样权重共4种分析策略对数据构建广义logistic回归模型,并比较结果。不同分析策略所得结果显示,不仅参数估计值、回归系数标准误、OR值及其置信区间的估计值有所差别,而且对纳入模型的解释变量也有影响。因此,在对复杂抽样调查设计多值名义资料构建广义logistics回归模型时,既要考虑抽样设计,又要兼顾抽样权重,否则即使样本量足够大,也会导致错误的推断结论。     

复杂抽样调查设计多值有序资料一水平多重Logistic回归分析

本文目的是比较不同分析策略对复杂抽样调查设计多值有序资料一水平多重logistic回归分析结果的异同。通过实例分析,利用四种不同的分析策略(将复杂抽样视为单纯随机抽样,考虑抽样设计不考虑抽样权重,考虑抽样权重不考虑抽样设计,同时考虑抽样设计和抽样权重)对复杂抽样设计多值有序资料进行建模。在四种不同分析策略的累积logistic回归模型拟合的结果中,自变量的偏回归系数、标准误差及P值均有所不同。在对复杂抽样调查设计的多值有序资料回归建模时,将抽样设计和抽样权重纳入统计分析,会得到更准确、更稳健的分析结果。     

生存资料回归模型分析--Cox比例风险假设的图形检验法

本文目的是介绍目前使用图形检验比例风险的常用方法。经典的Cox比例风险回归模型要求生存资料满足比例风险假设,而在临床资料中,这个假设往往并不成立。鉴于此,本文首先阐述了比例风险假设的概念;然后介绍了一些检验比例风险假设是否成立的常用图示方法,主要包括Kaplan-Meier生存曲线图、ln[-ln(St)]生存时间关系图、缩放Schoenfeld残差与时间的关系图、SAS软件PHREG过程中ACCESS语句的PH和RESAMPLE选项产生的模拟路径图;最后,基于SAS软件并通过实例演示上述方法的实现。     

生存资料回归模型分析--Cox比例风险假设的拟合优度法

本文目的是介绍一些检验比例风险假设的方法。图示法是通过绘图然后由人工进行判断是否符合比例风险假设,因而具有一定的主观性。在图示法的基础上,本文介绍了从客观角度检验比例风险假设的一些常用方法,主要包括两类:一类是基于残差的检验;另一类则是构建协变量与时间的交互项并对其进行检验的方法。首先阐述了上述方法的原理,然后基于SAS软件并通过一个实例介绍上述方法的实现。     

基于R软件实现随机分组及其应用

本文目的是使读者快速掌握用R软件实现几种随机分组的方法。通过借助R软件中实现随机抽样的sample()函数,间接地实现简单随机分组和分层随机分组的目的。事实表明:R软件易于获取、易学易用;R软件功能强大、适用面宽,可以方便快捷地解决试验设计中的随机分组问题。     

简单曲线回归分析及其应用

本文目的是介绍可以直线化的曲线回归分析相关内容及如何使用SAS软件来实现。一般来说,采用回归分析研究专业上确实存在联系的两个定量变量之间的依存关系。如果两定量变量之间是直线关系,那么直接采用直线回归分析即可;但在医学实验中,两定量变量之间的关系常常不是直线关系而是曲线关系,此时就应采用曲线回归分析。本文重点讲述可以直线化的曲线回归分析的种类及其SAS软件实现。     

基于SAS软件实现随机分组及应用

本文的目的是使读者能方便快捷地运用SAS软件中的PLAN过程实现随机分组。首先,对PLAN过程进行了简单介绍。接着,结合单因素设计、随机区组设计、具有重复试验的随机区组设计和拉丁方设计,介绍了随机分组的SAS实现方法。读者只需要修改本文中所呈现的SAS程序中的少量参数,就可很方便地用SAS软件实现自己的随机分组任务。事实说明,尽管SAS软件非常难学难用,但借助现成的SAS程序,可以轻松自如地解决很多具体问题。     

长处与困难问卷（父母版）在中国8个省份22,108名小学生中的调查结果

背景：我国儿童心理问题不断增多,因此能够有效评估儿童行为和情绪问题的筛选工具对于促进我国儿童身心健康和预防儿童精神疾病有重要的意义。 目的：在来源于中国不同地区城市和农村的大样本儿童中对国际公认的长处和困难问卷（SDQ）中文版的使用进行评估,包括情绪症状、品行问题、多动/注意缺陷、同伴交往问题和亲社会行为。 方法：采用分层随机抽样选取8省份22,108名小学生（年龄在5岁至13岁）的法定监护人（父母为主）进行SDQ（父母版）中文版问卷调查。本研究评估了SDQ评分与社会人口学特征之间的关联性,并比较了中国与日本和英国在“异常”,“临界”和“正常”标准的百分位划界分。 结果：SDQ（父母版）5个分量表中有4个内部一致性尚可,但是“同伴关系问题”分量表则较差（alpha=0.22）。监护人反映男孩比女孩的多动/注意力缺陷问题更多,而女孩比男孩的情感症状问题突出。男孩和女孩的多动/注意力缺陷问题都会随着年龄的增长而减少,而同伴交往问题随着年龄增长而增多。农村地区和监护人不是父母（即祖父母或其他亲属）的孩子中,情绪症状、品行问题和同伴交往问题较为普遍。中国儿童中判定为“异常”的90%百分位划界分（19-40）要高于日本和英国报道的儿童划界分。 讨论：此项研究提示,对男孩的多动/注意缺陷问题应在10岁之前开始预防,并且在青春期早期就要开始训练如何减轻与同伴交往带来的压力。进一步的研究需要注重于提高SDQ在中国文化背景下的信效度,并确定SDQ对于识别孩子是否需要心理健康服务的敏感性和特异性。     

Design, data analysis and sampling techniques for clinical research

Statistical analysis is an essential technique that enables a medical research practitioner to draw meaningful inference from their data analysis. Improper application of study design and data analysis may render insufficient and improper results and conclusion. Converting a medical problem into a statistical hypothesis with appropriate methodological and logical design and then back-translating the statistical results into relevant medical knowledge is a real challenge. This article explains various sampling methods that can be appropriately used in medical research with different scenarios and challenges.     

A procedure for the efficient collection and processing of pedigree data suitable for genetic analysis

Accurate and detailed pedigree data are crucial for family-genetic studies of psychiatric disorders. This methodologic paper describes a procedure for systematically collecting and processing pedigree data concerning a specified set of relatives. A flexible and essentially precoded data recording system for use in an interview situation is described and an example of its application is presented. The procedure guarantees that the investigator is able to determine the precise biological relationship between each pair of individuals in the pedigree, and it provides for the collection of other important information such as dates of birth and death. The system incorporates sufficient redundancy so that inconsistencies in the data recorded may be readily detected by computerized checking operations. A method is also described for reorienting the basic pedigree data obtained from the proband to facilitate the subsequent interviewing of additional relatives within a family.     

Factors affecting the foster care placement decision: an analysis of national survey data

With more than 350,000 children in foster care in the United States, this paper explores why some children are removed from their biological parents while others, with similar problems, receive services in the home. A discriminant analysis of children who were placed in foster care and those who were provided supportive services at home suggests that, for children in all age groups, parents' income level is the best predictor of a child's removal from the home.     

The impact of the study design and the sampling procedure on the assessment of mental health services

In this paper the impact of the study design and the sampling procedure on the assessment of the costs of schizophrenia treatment will be analyzed. The study sample consists of 307 patients with the diagnosis of schizophrenia (ICD 10F20) with the age between 18 and 65 years. Study participants were recruited consecutively in inpatient and outpatient treatment settings and in accommodation facilities according to the proportions of patients treated in these settings in Leipzig which have been assessed before. Treatment costs have been assessed by means of the German version of the Client Sociodemographic and Service Receipt Inventory at five follow-ups at six-months intervals. For the whole study period significant differences of the mean service costs due to the recruitment setting have been found. The mean yearly costs amounted to 44.669 DM for participants recruited in sheltered accommodation, 27.975 DM for participants in inpatient treatment and 5691 DM for patients in outpatient facilities. A random-effect regression model confirmed significant effects of the recruitment setting over all five follow-ups. Additionally, the variance of costs over time indicates that cost assessment by means of cross sectional data involves the danger of errors due to random dispersion.     

Psychiatric morbidity in prisoners with intellectual disabilities: analysis of prison survey data for England and Wales

A substantial number of prisoners have intellectual disabilities. We analysed data on a sample drawn from all prisons in England and Wales. Intellectual disability was defined as Quick Test scores equivalent to an IQ of ≤65. We found a significantly higher prevalence of probable psychosis, attempted suicide and cannabis use in prisoners with intellectual disabilities. Presence of intellectual disability was twice as likely to be associated with probable psychosis but the relationship was fully mediated by self-rated health status. It is important to identify this group as early as possible in order to provide timely interventions to cope in adverse environments and manage substance misuse.