多级属性Q矩阵的验证与估计

多级属性是将诊断测验中传统的二值(即两种水平, 通常定义为0和1)属性定义为多值(多个水平可以为0, 1, …), 它不但可以描述学生对于知识属性是否掌握, 而且可以描述学生在属性上的掌握程度, 这样使得诊断测验能提供给被试更丰富的知识掌握详情。本文将适用于二级属性Q矩阵的统计量(S统计量)拓展到多级属性下的Q矩阵验证和估计, 在两种常见的条件下, 设计了两种估计算法：联合估计算法和在线估计算法。模拟实验结果表明：联合估计算法适用于对专家界定的初始Q矩阵进行验证, 当初始Q矩阵中包含较少的错误时, 通过联合估计算法有很大可能恢复正确的Q矩阵; 在线估计算法适用于对“新项目”进行属性向量和项目参数的在线标定, 基于一定数量的“基础项目”, 在线估计算法对于新项目的估计也能达到较满意的成功率。实证数据分析则进一步展示了该方法的使用。     

Model-based recursive partitioning of extended redundancy analysis with an application to nicotine dependence among US adults

Extended redundancy analysis (ERA) is used to reduce multiple sets of predictors to a smaller number of components and examine the effects of these components on a response variable. In various social and behavioural studies, auxiliary covariates (e.g., gender, ethnicity) can often lead to heterogeneous subgroups of observations, each of which involves distinctive relationships between predictor and response variables. ERA is currently unable to consider such covariate-dependent heterogeneity to examine whether the model parameters vary across subgroups differentiated by covariates. To address this issue, we combine ERA with model-based recursive partitioning in a single framework. This combined method, MOB-ERA, aims to partition observations into heterogeneous subgroups recursively based on a set of covariates while fitting a specified ERA model to data. Upon the completion of the partitioning procedure, one can easily examine the difference in the estimated ERA parameters across covariate-dependent subgroups. Moreover, it produces a tree diagram that aids in visualizing a hierarchy of partitioning covariates, as well as interpreting their interactions. In the analysis of public data concerning nicotine dependence among US adults, the method uncovered heterogeneous subgroups characterized by several sociodemographic covariates, each of which yielded different directional relationships between three predictor sets and nicotine dependence.     

认知诊断模型中项目水平模型比较统计量的健壮性

使用模拟研究方法比较了以往研究中提出的基于观察信息矩阵、三明治矩阵的Wald（分别表示为W_Obs、W_Sw）、似然比（Likelihood Ratio）统计量以及新提出的基于经验交叉相乘信息矩阵的Wald统计量（W_XPD）在模型——数据失拟条件下进行项目水平上模型比较时的表现。结果显示：（1）W_Sw的一类错误控制率有很强的健壮性。（2）W_XPD在Q矩阵错误设定的大多数条件下的表现优于W_Sw。结论：模型—数据拟合良好时可以使用W_Sw进行项目水平上的模型比较,当模型与数据失拟时W_XPD可能是更好的选择。     

5～9岁儿童在不同复杂性任务上类比推理的发展特点

采用经典的A:B:C:D范式,设计两类难度不同的图片任务,依据关系-表征复杂性模型从等级复杂性和水平复杂性角度分析任务难度.在此基础上,选取5～9岁115名儿童探讨两类任务上类比推理的特点和发展趋势.表明:(1)儿童类比推理呈现多样化特点,主要有:充分意义上掌握了类比推理、部分完成了类比推理、位置关系类比推理、不能类比推理和不确定情况类比推理,其中部分完成了类比推理是复杂任务上所特有的.(2)简单、复杂两类任务上,随着年龄增长,5-9岁儿童充分意义上掌握类比推理的人次比例不断上升,不能类比推理的人次比例不断下降.此外,复杂任务上,随年龄增长,部分完成类比任务的人次比例降低.结果表明,在简单任务上,前运算阶段儿童已具备了类比推理能力;但在复杂任务上,儿童到了具体运算阶段才具备类比推理能力.     

A latent topic model with Markov transition for process data

We propose a latent topic model with a Markov transition for process data, which consists of time-stamped events recorded in a log file. Such data are becoming more widely available in computer-based educational assessment with complex problem-solving items. The proposed model can be viewed as an extension of the hierarchical Bayesian topic model with a hidden Markov structure to accommodate the underlying evolution of an examinee's latent state. Using topic transition probabilities along with response times enables us to capture examinees' learning trajectories, making clustering/classification more efficient. A forward-backward variational expectation-maximization (FB-VEM) algorithm is developed to tackle the challenging computational problem. Useful theoretical properties are established under certain asymptotic regimes. The proposed method is applied to a complex problem-solving item in the 2012 version of the Programme for International Student Assessment (PISA).     

Dysmorphic concern and behavioural impairment related to body image in adolescents

This study examined a moderated mediation model testing whether dysmorphic concern is related to behaviour altered to change appearance. This relationship is potentially mediated by depressive symptomatology (dysphoria and self-esteem) and ideas of reference about “laughing, commenting” and “attention, appearance,” and each mediated relationship moderated by sex. The sample was made up of 3377 adolescents from 12 to 18 years old (M_age = 14.52; SD = 1.65, 56.5% girls). The results showed that dysphoria and ideas of reference about “laughing, commenting” and “attention, appearance” partially mediated the relationship between dysmorphic concern and behavioural impairment related to body image. The relationship with dysphoria was moderated by sex, such that the mediation effect was stronger in girls than in boys. This result implies that girls who are worried about some characteristic of their appearance and show dysphoria are at greater risk of altered behaviour involving avoidance or controlling their appearance than boys. In addition, a possible risk of body dysmorphic disorder (3.45% of the sample) was found, with very prominent hiding behaviour using clothing or control behaviours, such as frequent weighing and looking at oneself in the mirror too much. Implications of these findings are discussed.     

Dimensions of Religion and Spirituality: A Longitudinal Topic Modeling Approach

In lieu of the traditional text data analysis methods, structural topic modeling was utilized to analyze the text contents of 255 self-report inventories of religion and spirituality (R/S) published from 1929 to 2017. The study had two objectives: (a) to clarify and identify the latent dimensions of R/S inherent in the items of the measures; and (b) to examine and demonstrate the usefulness of a longitudinal topic modeling in the study of R/S. We identified 5,617 unique text terms from the measures and fitted topic models on those terms to extract latent dimensions called topics. We also simultaneously analyzed the longitudinal effect of publication decade (i.e., 1950s–2010s) on the topics. A topic model with three topics was chosen to best support the data: Experience of Transcendence (Topic 1), Engagement in Transcendence (Topic 2), and Essence of Transcendence (Topic 3). In addition, the longitudinal analysis revealed that Topic 1 showed a continual increase over the decades, while Topics 2 and 3 both demonstrated a gradual decrease, in effect matching the general trend of Topic 1's increasing popularity in society and the academia.     

The Creativity Matrix: Spotlights and Blind Spots in Our Understanding of the Phenomenon

In this paper, we organize past and present theories and models of creativity by using a new conceptual framework—the creativity matrix—with the aim of highlighting the dimensions of creativity we know a lot about and those we tend to either ignore or find difficult to study. This matrix is formed by bringing together a developmental model of creativity (the 4 C's) and a structural one (the 5 A's). We start by briefly describing these two conceptual frameworks, and then, we proceed to exploring the matrix itself by describing how the 5 A's are dynamically organized at each “level” of the 4 C's. Importantly, our overview of the matrix is informed by existing models and concepts that address one of more of the C's and the A's. This gives us a unique opportunity to take stock of what has been studied so far and, toward the end, consider new avenues for the development of theory and research agendas within creativity studies.