线性回归模型结构稳定性检验方法及其比较

涉及时间序列数据的线性回归模型往往会出现模型的结构稳定性问题,以往学者在建模时大多忽略了模型结构稳定性的检验,这会造成模型的估计不够准确.文章探讨了模型的结构稳定性问题,提出CHOW检验和引入虚拟变量是解决该问题的两种方法,并利用保险业的数据进行了实证分析.虽然CHOW检验更受学者的欢迎,但引入虚拟变量能更好地解决模型的结构稳定性问题.     

“修正的克强指数”与GDP增长率时间序列建模分析

本文基于2008²011年的月度时间序列数据,建立了一元时间序列模型和二元时间序列的动态回归ARIMAX模型,运用ADF检验对各变量数据的平稳性进行检验,采用AIC、SBC准则选择相对最优模型.结果表明,可引入"修正的克强指数"对GDP增长率进行二元时间序列分析,并且建立的ARIMAX模型优于一元时间序列模型.     

详析模式方法及其在教育科学研究中的应用

详析模式是一种因果分析的多变量检验模型,通过引入检验变量来辨别、阐明和揭示原变量问的真实性因果关系。在教育研究中运用详析模式方法,能够检验教育变量间因果关系的真假、揭示变量间相互关系的县体条件、挖掘教育变量间的潜在关系,从而有效避免由对教育调查资料的感性认识而得出错误结论,有助于抓住教育问题的实质,为正确地进行教育决策提供切实可靠的依据。     

企业模型的研究

企业模型的建立是信息工程的核心和基础。是对整个企业进行一个总体的分析抽取企业中存在的稳定数据和数据关系,以便建立一个面向数据的信息系统,使其可以适应数据处理过程和方法的频繁变化,整个企业模型的建立极为复杂,本文在列出建立企业模型中常遇到的一些问题基础上,对构建企业模型的方法进行了探讨。     

奥运奖牌榜影响因素的实证分析——以第31届奥运会为例

本文采用计量经济学的方法对第三十一届奥林匹克运动会奖牌影响因素进行实证分析,以该届的数据为样本,建立多变量的回归模型,并对其进行估计、检验和解释.实证结果表明:一个国家的经济实力对奥运成绩是有着显著正向影响的,人口数量对奥运成绩可能有阻碍作用,社会主义制度对奥运成绩有不太显著的积极促进作用.     

企业模型的研究

企业模型的建立是信息工程的核心和基础.是对整个企业进行一个总体的分析抽取企业中存在的稳定数据和数据关系,以便建立一个面向数据的信息系统,使其可以适应数据处理过程和方法的频繁变化,整个企业模型的建立极为复杂,本文在列出建立企业模型中常遇到的一些问题基础上,对构建企业模型的方法进行了探讨.     

一种防止教育调查资料误用的有效方法——详析模式方法及其在教育科学研究中的应用

运用调查法进行教育问题研究,很容易出现对于调查资料的误用.详析模式是一种因果分析的多变量检验模型,通过引入检验变量来辨别、阐明和揭示原变量间的真实性因果关系.在教育研究中运用详析模式方法,能够检验教育变量间因果关系的真假、揭示变量间相互关系的具体条件、挖掘教育变量间的潜在关系,从而有效避免由对教育调查资料的感性认识而得出错误结论,帮助我们抓住教育问题的实质,为我们正确地进行教育决策提供切实可靠的依据.     

葡萄酒的评价模型

根据给出的数据,采用秩和检验法、T检验法和区分度标准,分析两组评酒员的结果差异和可信度。以酿酒葡萄的理化指标、葡萄酒的质量为参数,建立聚类分析模型,对酿酒葡萄进行分类定级;通过对葡萄和葡萄酒理化指标的相关性分析筛选变量,然后在二者之间进行多元回归拟合;运用同样的方法,得到用葡萄和葡萄酒的理化指标对葡萄酒质量的评价公式,并进行模型检验。     

高中学生数学建模认知特点比较研究

运用口语报告分析方法和专家—新手比较方法,对数学建模成绩优秀的高三学生(简称优生)与数学建模成绩一般的高三学生(简称一般生)数学建模的认知特点进行比较研究,在研究范围内和条件下获得以下结论:优生与一般生在数学建模的问题表征、策略运用、建模思路、解题结果及求解效率等方面表现出不同的认知特点.具体表现为:(1)在数学建模问题表征的方式、广度和方法方面:二者均采用符号表征方式和方法表征方式,但优生更多地采用机理表征方式;优生倾向于进行多元表征,一般生倾向于进行单一表征;优生倾向于运用循环表征方法,一般生倾向于运用单向表征方法.(2)在数学建模策略运用方面:优生倾向于采用平衡性假设策略,一般生倾向于采用精确性假设策略;优生倾向于采取样例类比构建策略,一般生倾向于采取即时生成构建策略;优生倾向于运用即时监控策略,一般生倾向于运用回顾监控策略;优生倾向于运用理论推演检验策略和直觉判断检验策略,一般生倾向于运用数据检验策略;优生倾向于运用假设调整策略和建模方法调整策略,一般生倾向于运用模型求解调整策略.(3)在数学建模的思路、结果及效率方面:优生数学建模口语报告比较简略,语言表达的逻辑性较强,问题分析深入而透彻,思路多元、快捷而灵活,对数学建模方法的使用表现为启发搜索,获得数学建模正确(合理)结果的次数较多,求解效率较高;而一般生数学建模口语报告比较繁杂,语言表达缺乏内在逻辑联系,对问题的分析浅表而模糊,建模思路单一、迟缓而刻板,对数学建模方法的使用表现为盲目搜索,思路定势和错误总次数较多,求解效率较低.     

环太湖地区创新集聚研究——基于环太湖五市的空间计量分析

以环太湖地区2005-2012年的面板数据为样本,运用空间计量方法对环太湖地区的创新集聚进行了实证研究。研究证明:E-G指数、LQ指数都显示环太湖地区的创新程度呈现了逐步扩大的变化趋势;同时,通过以环太湖地区的创新产出为被解释变量,以创新投入和一阶滞后的被解释变量为解释变量,建立动态的空间误差面板模型进行实证检验,检验结果为环太湖地区的创新产出和创新投入存在显著的正向空间相关性。以上结论提供的政策启示是建立环太湖地区协同创新机制。     

A Tutorial in Bayesian Potential Outcomes Mediation Analysis

Statistical mediation analysis is used to investigate intermediate variables in the relation between independent and dependent variables. Causal interpretation of mediation analyses is challenging because randomization of subjects to levels of the independent variable does not rule out the possibility of unmeasured confounders of the mediator to outcome relation. Furthermore, commonly used frequentist methods for mediation analysis compute the probability of the data given the null hypothesis, which is not the probability of a hypothesis given the data as in Bayesian analysis. Under certain assumptions, applying the potential outcomes framework to mediation analysis allows for the computation of causal effects, and statistical mediation in the Bayesian framework gives indirect effects probabilistic interpretations. This tutorial combines causal inference and Bayesian methods for mediation analysis so the indirect and direct effects have both causal and probabilistic interpretations. Steps in Bayesian causal mediation analysis are shown in the application to an empirical example.     

The importance of specifying and studying causal mechanisms in school-based randomised controlled trials: lessons from two studies of cross-age peer tutoring

Based on the experience of evaluating 2 cross-age peer-tutoring interventions, we argue that researchers need to pay greater attention to causal mechanisms within the context of school-based randomised controlled trials. Without studying mechanisms, researchers are less able to explain the underlying causal processes that give rise to results from randomised controlled trials. Studying implementation fidelity is necessary but not sufficient for causal explanation; the study of causal mechanisms through the application of mixed methods is also required. Due to the increasingly complicated nature of many classroom-based innovations that are subject to evaluation, and the potentially distal nature of hypothesised effects, particularly on attainment, programme theory and articulation of mechanisms are essential in enhancing causal explanation and promoting the accumulation of knowledge of what works and why in classroom settings.     

The Principle of the Nomograph in Education

ABSTRACT

Educational researchers frequently study the impact of treatments or interventions on educational outcomes. However, when observational or quasiexperimental data are used for such investigations, selection bias can adversely impact researchers’ abilities to make causal inferences about treatment effects. One way to deal with selection bias is to use propensity score methods. The authors introduce educational researchers to the general principles underlying propensity score methods, describe 2 practical applications of these methods, and discuss their limitations.     

经济增长与能源消费:来自山东的经验证据

以山东省为样本,通过对产出变量和能源消费变量进行单位根检验和协整检验,利用Granger因果关系模型检验了能源消费与经济增长之间的因果关系。构建能源消费与经济增长之间的误差修正模型,发现不同类型的能源消费与产出之间存在不同的协整关系;在具有协整关系的变量中,存在从经济增长到能源消费的单向因果关系;经济结构对能源消费具有显著影响。这表明,山东省为非能源依赖型经济,如果方式合适,节能政策对山东经济增长目标的达成影响不大。为了在2010年实现单位地区生产总值能源消耗降低20%的政策目标,山东必须制定相关政策,促进第三产业的发展,促进第二产业节能技术的进步,激励采用节约电力的生产过程。     

Learning Hierarchies: a critique

Learning hierarchies have received much attention from developmental and instructional psychologists. This article notes that conceptual confusions and methodological deficiencies occur in much of the research so far published. The conceptual confusions concern the terminology used; the ‘likelihood’ or ‘causal’ relationships between elements in the hierarchy; the distinction between ‘prerequisition’ and ‘positive transfer'; the distinction between single pieces of learning and classes of learning; the inclusivity of hierarchical relationships. The methodological deficiencies arise from an inability to measure ‘causal’ relationships; the omission of measurements of ‘positive transfer'; the difficulty of measuring the range of possible relationships within a hierarchy; the need to remove instructional effects from hierarchy validation studies. It is concluded that these confusions and deficiencies preclude data from learning hierarchy studies from being used to diagnose learning failure and in test construction. Suggestions for alternatives to, and improvements on, current methods are made.     

Taking account of time lags in causal models

Although it takes time for a cause to exert an effect, causal models often fail to allow adequately for time lags. In particular, causal models that contain cross-sectional relations (i.e., relations between values of 2 variables at the same time) are unsatisfactory because they omit the values of variables at prior times, they omit effects that variables can have on themselves, and they fail to specify the length of the causal interval that is being studied. These omissions can produce severe biases in estimates of the size of causal effects. Longitudinal models also can fail to take account of time lags properly, and this too can lead to severely biased estimates. The discussion illustrates the biases that can occur in both cross-sectional and longitudinal models, introduces the latent longitudinal approach to causal modeling, and shows how latent longitudinal models can be used to reduce bias by taking account of time lags even when data are available for only 1 point in time.     

Conceptual Understanding of Causal Reasoning in Physics

College students often experience difficulties in solving physics problems. These difficulties largely result from a lack of conceptual understanding of the topic. The processes of conceptual learning reflect the nature of the causal reasoning process. Two major causal reasoning methods are the covariational and the mechanism‐based approaches. This study was to investigate the effects of different causal reasoning methods on facilitating students’ conceptual understanding of physics. 125 college students from an introduction physics class were assigned into covariational group, mechanism‐based group, and control group. The results show that the mechanism‐based group significantly outperformed the other two groups in solving conceptual problems. However, no significant difference was found in all three groups performance on solving computational problems. Speculation on the inconsistent performance of the mechanism‐based group in conceptual and computational problem solving is given. Detailed analyses of the results, findings, and educational implications are discussed     

Class size effects on mathematics achievement in Cyprus: evidence from TIMSS

Class size reduction has been viewed as one school mechanism that can improve student achievement. Nonetheless, the literature has reported mixed findings about class size effects. We used 4th- and 8th-grade data from TIMSS 2003 and 2007 to examine the association between class size and mathematics achievement in public schools in Cyprus. We employ instrumental variables methods, and take advantage of a regression discontinuity design to examine causal effects of class size on mathematics achievement. The results indicate a non-significant relationship between class size and mathematics achievement in 8th grades. However, there is evidence of positive class size effects in 4th grade. The gender gap is significant and favoured males in 4th grade and females in 8th grade. SES indexes such as parental education and items in the home are positively and significantly related to mathematics achievement. Teacher and school variables are not significantly related with mathematics achievement.     

Designing effective supports for causal reasoning

Causal reasoning represents one of the most basic and important cognitive processes that underpin all higher-order activities, such as conceptual understanding and problem solving. Hume called causality the “cement of the universe” [Hume (1739/2000). Causal reasoning is required for making predictions, drawing implications and inferences, and explaining phenomena. Causal relations are usually more complex than learners understand. In order to be able to understand and apply causal relationships, learners must be able to articulate numerous covariational attributes of causal relationships, including direction, valency, probability, duration, responsiveness, as well as mechanistic attributes, including process, conjunctions/disjunctions, and necessity/sufficiency. We describe different methods for supporting causal learning, including influence diagrams, simulations, questions, and different causal modeling tools, including expert systems, systems dynamics tools, and causal modeling tools. Extensive research is needed to validate and contrast these methods for supporting causal reasoning.     

A longitudinal study of children’s social behaviors and their causal relationship to reading growth

This paper aims at investigating the causal effects of social behaviors on subsequent reading growth in elementary school, using the Early Childhood Longitudinal Study—Kindergarten (ECLS-K) data. The sample was 8,869 subjects who provided longitudinal measures of reading IRT scores from kindergarten (1998–1999) to fifth grades (2003–2004) in the United States. To examine the causal relationship, propensity score methods were used to match higher and lower groups in four social behavior domains such as Approaches to learning, Interpersonal skills, Internalizing problem behavior and Externalizing problem behavior. Results showed that the matched sample achieved sufficient pretreatment balance between the two groups. To examine the effects of social behaviors on the reading growth, multilevel growth model (MGM) was employed. Comparisons of the matched samples showed that children in the high groups of pro-social behavior or in the low groups of problem behavior at kindergarten entrance started with higher reading skills and developed reading achievement faster than those who were not. This study suggests that children’s early social behavior is crucial in reading development. Practical implication and direction of future research are also discussed.