认知诊断CAT中项目曝光控制方法的比较

项目曝光率关系到题库建设和测验安全,是计算机化自适应测验(Computerized Adaptive Testing, CAT)需要考虑的重要问题。在认知诊断 CAT 情形下,首先基于传统 CAT 中 a-分层方法的思想提出按项目信息量对题库分层的分层多阶段(Stratified Multistage, SM)选题方法;然后将 SM 方法与项目合格(Item Eligibility, IE)方法相结合得到SMIE方法。在此基础上,开展模拟研究比较SM、IE、SMIE、最大修正优先指标(Maximum Modified Priority Index, MMPI)方法、限制阈值(Restrictive Threshold, RT)方法和限制进度(Restrictive Progressive, RPG)方法的选题表现。总体上,它们的测量精度从高到低依次为IE、SM、SMIE、RT、RPG和MMPI方法;项目曝光分布均匀性的优劣次序为MMPI、RPG、SMIE、RT、SM和IE方法;SMIE和RT方法能较好地平衡测量精度和项目曝光均匀性要求。     

二参数逻辑斯蒂模型项目参数的估计精度

项目参数的估计精度对于测验的编制尤其是题库的建立十分重要。目前,国内外对项目参数估计精度的研究,大部分是基于在已知项目参数真值的情况下,运用各种参数估计方法产生新的估计值,再和真值进行偏度(BIAS)和均方根差(RMSE)的比较,从而说明该种估计方法的有效性。但是这种方法不能提供不同的参数真值之间的估计误差的变化规律。为了弥补这一缺陷,本文尝试从项目参数估计信息函数的角度出发研究项目参数的估计精度问题。本研究以二参数Logistic模型作为研究对象,首先定义了项目参数的估计信息函数,然后基于完全随机实验设计,通过模拟研究的方法探索影响项目参数的估计精度的因素,实验共设计了(2×3×2)种情形。研究结果表明：(1)项目参数(a,b)的估计精度均随着被试样本量的增大而提高;(2)被试的能力分布对难度参数的估计精度影响较大,对区分度参数的估计精度影响相对较小;(3)难度参数和区分度参数的估计精度都分别受到参数a和参数b的共同作用。     

无领导小组讨论的多侧面Rasch模型应用*

采用项目反应理论(IRT)的多侧面Rasch模型(MFRM),分析评价中心技术中无领导小组讨论(LGD)的测评结果,探讨被试能力水平、评委评分宽严度、评分内部一致性、维度难度和评定等级等问题,进而讨论各种偏差。通过 MFRM 分析人事测评结果,可深入了解被试能力的真实差异、甑别维度难度、探查测评误差源,从而完善测评试题编制、评估或诊断评委合格性、提高测评维度与测评目的匹配性,为拓展项目反应理论在人事测评中的应用提供独特视角。     

Assessing the Verbal Behavior of Conduct Problem Children During Mother-Child Interactions:

In this study we compared the verbal behavior of 4 to 8-year-old conduct problem and normal children in interactions with their mothers in the Child-Directed Interaction phase of the Dyadic Parent-Child Interaction Coding System (Eyberg & Robinson, 1983). It was predicted that conduct problem children would use less praise, ask fewer questions, make more critical statements, and give more commands during interactions with their mothers than normal children. Results support two of these four predictions, with conduct problem children using fewer questions and less praise. Implications for the assessment and treatment of conduct problems as well as how these verbal behaviors may develop from problem parent-child interactions are discussed.     

Relations between a ruminative thinking style and obsessive–compulsive symptoms in non-clinical samples

Abstract

There is tremendous interest in understanding the cognitive processes behind obsessive–compulsive disorder (OCD). Whereas previous research on cognitive OCD models has focused on the dysfunctional content of obsessional thinking, processes and styles of thinking have not yet been investigated. The present study investigated the relationship between a ruminative response style and obsessive–compulsive (OC) symptomatology in two non-clinical samples. In Sample 1, 261 students completed the Ruminative Response Scale, the Padua-Inventory, Revised, and the Beck Depression Inventory. Tendency to ruminate was positively correlated with the severity of OC symptoms and particularly with obsessive rumination, even after controlling for depression. Results were replicated in Sample 2 (211 students). Data indicate that a ruminative response style and obsessive rumination share common processual features. Understanding the interaction between rumination and obsessional thinking might help to further elucidate the role of cognitive vulnerability factors in OCD and to expand cognitive and metacognitive models of OCD.     

Toward a Public Mental Health Approach for Survivors of Terrorism

Abstract

Although most people exposed to bioterrorism or mass casualties will be extremely distressed during the immediate aftermath, only a minority (approximately 30%) will develop clinically significant psychiatric disorders. From a public mental health perspective, the challenge is to provide both preventive programs for the entire population and early detection and intervention for those at greatest risk for PTSD or other post-traumatic psychiatric disorders. Both individual and soci-etal preventive and early intervention approaches are reviewed. Utilization of the media, especially television, is presented as an example of one of many potential community/societal public mental health approaches.     

Miserliness in human cognition: the interaction of detection,override and mindware

ABSTRACT

Humans are cognitive misers because their basic tendency is to default to processing mechanisms of low computational expense. Such a tendency leads to suboptimal outcomes in certain types of hostile environments. The theoretical inferences made from correct and incorrect responding on heuristics and biases tasks have been overly simplified, however. The framework developed here traces the complexities inherent in these tasks by identifying five processing states that are possible in most heuristics and biases tasks. The framework also identifies three possible processing defects: inadequately learned mindware; failure to detect the necessity of overriding the miserly response; and failure to sustain the override process once initiated. An important insight gained from using the framework is that degree of mindware instantiation is strongly related to the probability of successful detection and override. Thus, errors on such tasks cannot be unambiguously attributed to miserly processing – and correct responses are not necessarily the result of computationally expensive cognition.     

The logic sense: exploring the role of executive functioning in belief and logic-based judgments

Abstract

The Default Interventionist account suggests that by default, we often generate belief-based responses when reasoning and find it difficult to draw the logical inference. Recent research, however, shows that in some instances belief judgments take longer, are more prone to error and are more affected by cognitive load. One interpretation is that some logical inferences are available automatically and require intervention in order to respond according to beliefs. In two experiments, we investigate the effortful nature of belief judgments and the automaticity of logical inferences by increasing the inhibitory demands of the task. Participants were instructed to judge conclusion validity, believability and either font colour or font style, to increase the number of competing responses. Results showed that conflict more strongly affects judgments of believability than validity and when inhibitory demands are increased, the validity of an argument impacts more on belief judgments. These findings align with the new Parallel Processing model of belief bias.     

Just more of the same,or different? An integrative theoretical framework for the study of cumulative interruptions at work

We propose a theoretical framework that explores the accumulation of work interruptions and their effects. Most research studies have dealt with interruptions as isolated phenomena, ignoring the simultaneous or sequential occurrence of interruptions common in everyday life. We fill this gap and provide insight into the process of the accumulation of interruptions by mapping deep-level regulation onto an observable sequence of actions. Furthermore, we explain how cumulative interruptions can lead to qualitatively different effects because of the interaction and joint development of isolated effects, identifying some mediating and moderating factors. In doing so, we disclose the relationships between the effects of single interruptions found in laboratory studies and the impacts on health and well-being of multiple interruptions found in applied research.