A method for simulating non-normal distributions

A method of introducing a controlled degree of skew and kurtosis for Monte Carlo studies was derived. The form of such a transformation on normal deviates [X N(0, 1)] isY =a +bX +cX ² +dX ³. Analytic and empirical validation of the method is demonstrated.This work was done while the author was at the University of Illinois at Champaign-Urbana.     

多水平随机中介效应估计及其比较

本文在综述各类多水平中介模型的基础上, 聚焦于自变量、中介变量、因变量都来自多水平结构中较低水平的多水平随机中介效应模型, 通过蒙特卡洛模拟研究比较该模型与简化的多水平固定中介效应模型、传统中介效应模型的差别, 并考察了目前用于多水平随机中介效应的三种参数估计方法：限制性极大似然、极大似然、最小方差二次无偏估计在不同情况下对随机中介效应估计的优劣。研究结果显示：当数据符合多水平随机中介效应模型时, 使用简化模型将错误估计中介效应及其标准误, 得到不正确的统计检验结果; 使用多水平随机中介效应模型能够实现对中介效应的正确估计和检验, 其中限制性极大似然或极大似然估计方法优于最小方差二次无偏估计方法。     

分类精确性指数Entropy在潜剖面分析中的表现:一项蒙特卡罗模拟研究

本研究通过蒙特卡洛模拟考查了分类精确性指数Entropy及其变式受样本量、潜类别数目、类别距离和指标个数及其组合的影响情况。研究结果表明:(1)尽管Entropy值与分类精确性高相关,但其值随类别数、样本量和指标数的变化而变化,很难确定唯一的临界值;(2)其他条件不变的情况下,样本量越大,Entropy的值越小,分类精确性越差;(3)类别距离对分类精确性的影响具有跨样本量和跨类别数的一致性;(4)小样本(N=50~100)的情况下,指标数越多,Entropy的结果越好;(5)在各种条件下Entropy对分类错误率比其它变式更灵敏。     

How Do Propensity Score Methods Measure Up in the Presence of Measurement Error? A Monte Carlo Study

Considering that the absence of measurement error in research is a rare phenomenon and its effects can be dramatic, we examine the impact of measurement error on propensity score (PS) analysis used to minimize selection bias in behavioral and social observational studies. A Monte Carlo study was conducted to explore the effects of measurement error on the treatment effect and balance estimates in PS analysis across seven different PS conditioning methods. In general, the results indicate that even low levels of measurement error in the covariates lead to substantial bias in estimates of treatment effects and concomitant reduction in confidence interval coverage across all methods of conditioning on the PS.     

R-RUM的参数估计及性能评价研究

认知诊断作为21世纪一种新的测量范式,在国内外越来越受到重视。该文运用MCMC算法实现了R-RUM的参数估计,并采用Monte Carlo模拟方法探讨其性能。研究结果表明:(1)R-RUM参数估计方法可行,估计精度较高;(2)Q矩阵复杂性和模型参数水平对模型参数估计精度有较大影响,随着r_(jk)*值的增大和Q矩阵复杂性的增加,项目参数和被试参数估计精度逐渐下降;(3)在特定情形下,R-RUM具有一定的稳健性。     

三类多层中介效应分析方法比较

比较了贝叶斯法、Monte Carlo法和参数Bootstrap法在2-1-1多层中介分析中的表现。结果发现：1)有先验信息的贝叶斯法的中介效应点估计和区间估计都最准确;2)无先验信息的贝叶斯法、Monte Carlo法、偏差校正和未校正的参数Bootstrap法的中介效应点估计和区间估计表现相当,但Monte Carlo法在第Ⅰ类错误率和区间宽度指标上表现略优于其他三种方法,偏差校正的Bootstrap法在统计检验力上表现略优于其他三种方法,但在第Ⅰ类错误率上表现最差;结果表明,当有先验信息时,推荐使用贝叶斯法;当先验信息不可得时,推荐使用Monte Carlo法。     

抗感染治疗的今天和明天

抗菌药物治疗是临床医学界普遍关切的问题。滥用抗生素的问题引起社会和政府的重视。但解决这问题需要重视调查研究,充分应用科学研究成果,制定科学合理的政策和规定。本文还从医生的角度,对经验用药、社区获得性肺炎治疗中的几个问题,介绍了国外的做法,对应用药代动力学和最低抑菌浓度结合的研究方法,包括Monte Carlo模拟的实验和临床研究,以及防突变浓度（MPC）和突变选择窗（MEw）的概念作了介绍。     

Universality of the reversal transition in growth phenomena of molecular crystals

ABSTRACT

The fundamental origin of the reversal transition during crystal growth has been investigated by Monte Carlo simulations. Evidence is provided supporting the geometrical nature of the transition, arising from the broken translational invariance at the growth front and the interactions of multipoles of opposed parity. The range of interactions, the details of the force field and model are irrelevant to the occurrence and only determine the functional behaviour of the relevant physical quantities like the localisation of the critical point and the morphology of the phase diagram.     

Photic Stimulation of Young Rats

Zajonc, Harrison, and their colleagues have recently conducted a series of studies demonstrating a positive, monotonic relation between frequency of “mere” exposure and liking for stimuli. Other studies have found either the inverted-U relation or a decrease in liking. It was proposed that an emphasis upon mere exposure may be somewhat misleading, and that a concern with degree of stimulus familiarity might be more fruitful. Conflicting forms of the exposure-liking relation seem to be potentially reconcilable if consideration is given to factors that influence the rate at which stimuli become familiar, or capable of being anticipated and represented in memory.

Two experiments using stimuli and procedures taken from Zajonc (23) yielded a positive, monotonic relation between frequency and liking. A third experiment, designed to produce greater attention to the stimuli and thus hasten familiarization, yielded an inverted-U relation. A fourth experiment used simpler verbal stimuli than the first three. This was designed to result in even faster familiarization. As expected, there was a negative relation between liking and exposure frequency.

The “frequency group” of investigators has explained the monotonic exposure effect in terms of the response competition hypothesis. However, the latter is incapable of handling a nonmonotonic relation. A reasonable alternative seems to be the expectancy arousal hypothesis (5): viz., that liking is maximum for stimuli that arouse moderately strong expectancies of either a “molar” or a “molecular” nature. Previous research (5, 6, 7, 8) has supported this hypothesis, although some of the studies are also amenable to the response competition hypothesis. A further experiment was designed to distinguish between predictions from the response competition and the expectancy arousal hypotheses. The results supported the latter.     

Mixed effects modeling of Morris water maze data: Advantages and cautionary notes

Morris water maze data are most commonly analyzed using repeated measures analysis of variance in which daily test sessions are analyzed as an unordered categorical variable. This approach, however, may lack power, relies heavily on post hoc tests of daily performance that can complicate interpretation, and does not target the nonlinear trends evidenced in learning data. The present project used Monte Carlo simulation to compare the relative strengths of the traditional approach with both linear and nonlinear mixed effects modeling that identifies the learning function for each animal and condition. Both trend-based mixed effects modeling approaches showed much greater sensitivity to identifying real effects, and the nonlinear approach provided uniformly better fits of learning trends. The common practice of removing a rat from the maze after 90 s, however, proved more problematic for the nonlinear approach and produced an underestimate of y-axis intercepts.     

基于概化理论的方差分量变异量估计

概化理论广泛应用于心理与教育测量实践中, 方差分量估计是进行概化理论分析的关键。方差分量估计受限于抽样, 需要对其变异量进行探讨。采用蒙特卡洛(Monte Carlo)数据模拟技术, 在正态分布下讨论不同方法对基于概化理论的方差分量变异量估计的影响。结果表明: Jackknife方法在方差分量变异量估计上不足取; 不采取Bootstrap方法的“分而治之”策略, 从总体上看, Traditional方法和有先验信息的MCMC方法在标准误及置信区间这两个变异量估计上优势明显。