CORE--BIOSTATISTICAL AND DATA MANAGEMENT

生物统计/数据管理核心的目的是开发、 维护和保证计划项目数据库的完整性和 指导适当的统计分析，以解决科学 节目的问题。数据管理的具体目标和 生物统计学的核心是：1.创建、管理和保护项目的数据 基地；2.与项目科学家合作进行设计和分析 3.发展新的统计方法，例如 对于关于项目假设的有效和有效的推论是必要的。 该数据库将在约翰霍普金斯大学威尔默研究所和 在带有磁带/盒式磁带备份的专用微型计算机上维护。 问卷将标准化，数据录入将得到核实，范围将扩大- 检查；每月摘要报告将分发给调查员； 备份将每周进行一次，并在异地存放一份副本。 统计分析将分三个阶段进行：探索 观测变量的单变量和多变量经验分布； 观测到的响应和响应之间关系的统计模型 解释变量；以及将考虑到的形式推论 潜变量和测量误差。第一步和第二步将是 采用S和SAS统计软件进行统计分析。最后一步将是 使用程序LISREL进行连续测量，但需要 方法论和计算机软件开发。

The purpose of the Biostatistics/Data Management core is to develop, maintain and guarantee the integrity of the program project data base and to direct appropriate statistical analyses to address the scientific questions of the program. The specific aims of the data management and biostatistics core are: 1. to create, manage and secure the project's data base; 2. to collaborate with project scientists in the design and analysis of the component studies; and 3. to develop new statistical methodology as necessary for valid and efficient inferences about project hypotheses. The data base will be created at the Johns Hopkins Wilmer Institute and maintained on a dedicated microcomputer with tape/cartridge backup. Questionnaires will be standardized, data entry will be verified and range- checked; monthly summary reports will be distributed to investigators; backups will be made weekly and one copy housed off-site. The statistical analysis will be conducted in three stages: exploration of univariate and multivariate empirical distributions of observed variables; statistical modelling of the relationship between observed responses and explanatory variables; and formal inferences which will take account of latent variables and measurement error. The first and second steps will be performed using the statistical packages S and SAS. The final step will use the program LISREL for continuous measures but will require methodologic and computer software development more generally.