Impact of Diagnosticity on the Adequacy of Models for Cognitive Diagnosis under a Linear Attribute Structure: A Simulation Study

Compared to unidimensional item response models (IRMs), cognitive diagnostic models (CDMs) based on latent classes represent examinees' knowledge and item requirements using discrete structures. This study systematically examines the viability of retrofitting CDMs to IRM‐based data with a linear attribute structure. The study utilizes a procedure to make the IRM and CDM frameworks comparable and investigates how estimation accuracy is affected by test diagnosticity and the match between the true and fitted models. The study shows that comparable results can be obtained when highly diagnostic IRM data are retrofitted with CDM, and vice versa, retrofitting CDMs to IRM‐based data in some conditions can result in considerable examinee misclassification, and model fit indices provide limited indication of the accuracy of item parameter estimation and attribute classification.     

Causal and predictive-value judgments,but not predictions,are based on cue-outcome contingency

In three experiments, we show that people respond differently when they make predictions as opposed to when they are asked to estimate the causal or the predictive value of cues: Their response to each of those three questions is based on different sets of information. More specifically, we show that prediction judgments depend on the probability of the outcome given the cue, whereas causal and predictivevalue judgments depend on thecue-outcome contingency. Although these results might seem problematic for most associative models in their present form, they can be explained by explicitly assuming the existence of postacquisition processes that modulate participants’ responses in a flexible way.     

Changes in Memory Prediction Accuracy: Age and Performance Effects

Memory performance predictions are subjective estimates of possible memory task performance. The purpose of this study was to examine possible factors related to changes in word list performance predictions made by younger and older adults. Factors included memory self-efficacy, actual performance, and perceptions of performance. The current study examined the age differences in a sample of 59 older adults (M = 76.83 years) and 52 younger adults (M = 21.19 years) on memory predictions and the accuracy of those predictions for both an immediate and delayed word recall task. While memory self-efficacy did not influence predictions for either group, perceived and actual performance at immediate recall was related to accuracy of predictions at delay for both groups. In addition, there was a significant age-by-time interaction such that older adults became more accurate over time whereas younger adults declined slightly in their prediction accuracy. These findings suggest that older adults are able to make accurate adjustments to their predictions based on both their perceived and actual performance on tasks.     

Bayesian approaches to associative learning: From passive to active learning

Traditional associationist models represent an organism's knowledge state by a single strength of association on each associative link. Bayesian models instead represent knowledge by a distribution of graded degrees of belief over a range of candidate hypotheses. Many traditional associationist models assume that the learner is passive, adjusting strengths of association only in reaction to stimuli delivered by the environment. Bayesian models, on the other hand, can describe how the learner should actively probe the environment to learn optimally. The first part of this article reviews two Bayesian accounts of backward blocking, a phenomenon that is challenging for many traditional theories. The broad Bayesian framework, in which these models reside, is also selectively reviewed. The second part focuses on two formalizations of optimal active learning: maximizing either the expected information gain or the probability gain. New analyses of optimal active learning by a Kalman filter and by a noisy-logic gate show that these two Bayesian models make different predictions for some environments. The Kalman filter predictions are disconfirmed in at least one case.     

Distinguishing social and asocial learning using diffusion dynamics

Theoretical models predict that the cumulative number of individuals displaying a socially learned novel behavior will follow an accelerating pattern over time, whereas asocial processes have been associated with linear or decelerating functions. This raises the possibility that the shape of the diffusion curve may reveal something about the learning processes involved. If true, this would be particularly useful for identifying social transmission in observational field studies. Published data are reviewed and are found to provide limited support for this view. The use of accelerating curves as a diagnostic is challenging because (1) alternative theoretical models make similar predictions, (2) clear supporting empirical data are lacking, and (3) practical considerations frequently make accurate construction of the diffusion curve difficult.     

Interactions of gender with predictors of academic achievement

Predictive models of academic achievement are used in various (often high stakes) applications, including selection and study orientation procedures for higher education. Considering the far-reaching consequences of their outcomes, these models should show as little bias for irrelevant factors as possible. While numerous studies have researched the impact of gender on the isolated individual predictors of academic achievement, no studies yet have explored how gender affects program-specific prediction models of academic achievement. As such, the present study examined whether prediction models exhibit gender differences in the accuracy of their predictions, and how such differences relate to the gender balance within a study program. Besides that, we developed gender-specific prediction models of academic achievement in order to examine how these models differ in terms of which predictors are included, and whether they make more accurate predictions. Data was examined from a large sample of first year students across 16 programs in an open access higher education system (N = 5,016). Results revealed interactions between gender and several predictors of academic achievement. While the models exhibited little difference in the accuracy of their predictions for male and female students, analyses showed that using gender-specific models substantially improved our predictions. We also found that male and female models of academic achievement differ greatly in terms of the predictors included in their composition, irrespective of the gender balance in a study program.     

Effect of Student Characteristics on Teachers' Predictions of Student Success

Abstract

To determine teachers' perceptions of the likelihood of students' academic and social success in general education settings, 384 teachers responded to 1 of 32 possible case studies describing a student, in which gender, reading achievement, social behavior, and attentiveness were manipulated experimentally. Results showed that, in some instances, teachers used irrelevant information in their predictions (e.g., social behavior influenced academic predictions, reading achievement influenced social predictions). Furthermore, teachers demonstrated gender bias in some instances and appeared to make negative predictions when an ambiguous label explained students' reading problems. The author discusses implications for the inclusion of students with special needs in general education classes.     

Three- to Four-Year-Old Children Rapidly Adapt Their Predictions and Use Them to Learn Novel Word Meanings

Adults create and update predictions about what speakers will say next. This study asks whether prediction can drive language acquisition, by testing whether 3- to 4-year-old children (n = 45) adapt to recent information when learning novel words. The study used a syntactic context which can precede both nouns and verbs to manipulate children's predictions about what syntactic category will follow. Children for whom the syntactic context predicted verbs were more likely to infer that a novel word appearing in this context referred to an action, than children for whom it predicted nouns. This suggests that children make rapid changes to their predictions, and use this information to learn novel information, supporting the role of prediction in language acquisition.     

Two-Year-Olds Readily Learn Multiple Labels for the Same Basic-Level Category

2 basic frameworks for lexical principles have been proposed (Golinkoff, Mervis, & Hirsh-Pasek; Markman). In many types of situations, these frameworks make the same predictions regarding 2-year-olds' interpretation of novel terms. However, the predictions diverge in some cases. 3 experiments were conducted to address 1 such situation: the case in which the child hears a novel term but already knows a label for each of the objects present. Results of all 3 studies converged on the same outcome: The new term was most likely to be treated as a second basic-level name for the category to which the object belonged. Analyses of individual patterns of responding revealed that, of the 58 subjects, 50 made more basic-level responses than part responses, 1 made equal numbers of basic-level and part responses, and 7 made more part responses than basic-level responses. Implications of these findings for models of early lexical development are discussed.     

Assessing power PC

In a recent theoretical paper, Cheng (1997) presented a new causal model, power PC. She argued that power PC was able to account for data in the literature that raised problems for associative models—notably, the Rescorla-Wagner model (Rescorla & Wagner, 1972). The purpose of the present paper is threefold: (1) to show that, overall, the data in the literature, which Cheng relied on to make her case, do not in fact provide support for power PC, (2) to show that, overall, the experiments reported in the literature since the publication of Cheng, designed specifically to evaluate the predictions of power PC, also do not provide support for power PC, and (3) to suggest that Cheng’s assessment of associative models was too narrowly defined.     

Combining Principal Component Regression and Artificial Neural Network to Predict Chlorophyll-a Concentration of Yuqiao Reservoir’s Outflow

In order to investigate the eutrophication degree of Yuqiao Reservoir, a hybrid method, combining principal component regression (PCR) and artificial neural network (ANN), was adopted to predict chlorophyll-a concentration of Yuqiao Reservoir’s outflow. The data were obtained from two sampling sites, site 1 in the reservoir, and site 2 near the dam. Seven water variables, namely chlorophyll-a concentration of site 2 at time t and that of both sites 10 days before t, total phosphorus(TP), total nitrogen(TN), dissolved oxygen(DO), and temperature from January 2000 to September 2002, were utilized to develop models. To remove the collinearity between the variables, principal components extracted by principal component analysis were employed as predictors for models. The performance of models was assessed by the square of correlation coefficient, mean absolute error (MAE), root mean square error (RMSE) and average absolute relative error (AARE). Results show that the hybrid method has achieved more accurate prediction than PCR or ANN model. Finally, the three models were applied to predicting the chlorophyll-a concentration in 2003. The predictions of the hybrid method were found to be consistent with the observed values all year round, while the results of PCR and ANN models did not fit quite well from July to October.     

含油制冷剂R600a管内两相流动摩擦压降研究

基于文献含油制冷剂R600a两相流动摩擦压降实验数据,采用文献中6种关联式模型对含油制冷剂R600a两相摩擦压降的实验数据进行预测判断。结果表明:油影响因子模型Schlager关联式和Eckel关联式的预测值与实验值平均绝对偏差分别为27.3%和24.4%;混合物性的两相增强因子模型Zurcher关联式和Hu关联式的预测值与实验值的平均绝对偏差分别为33.8%和34.0%,上述关联式预测值普遍低于实验值。     

Children’s Orthographic Knowledge and Their Word Reading Skill: Testing Bidirectional Relations

Prominent models of word reading concur that the development of efficient word reading depends on the establishment of lexical orthographic representations in memory. In turn, word reading skills are conceptualised as supporting the development of these orthographic representations. As such, models of word reading development make clear predictions of bidirectional relations between lexical orthographic knowledge and word reading skill. We test these predictions in a longitudinal study of 112 English-speaking children in Grades 2 and 3. At two time points, we assessed lexical orthographic knowledge and three aspects of word reading skill: word reading accuracy, word reading efficiency, and phonological decoding. Consistent with theoretical predictions, we found that earlier word reading accuracy, word reading efficiency, and phonological decoding predicted gains in lexical orthographic knowledge. Contrary to theoretical predictions, lexical orthographic knowledge did not predict gains in any of our measured word reading skills.     

Young Children Are Wishful Thinkers: The Development of Wishful Thinking in 3- to 10-Year-Old Children

Previously, research on wishful thinking has found that desires bias older children’s and adults’ predictions during probabilistic reasoning tasks. In this article, we explore wishful thinking in children aged 3- to 10-years-old. Do young children learn to be wishful thinkers? Or do they begin with a wishful thinking bias that is gradually overturned during development? Across five experiments, we compare low- and middle-income United States and Peruvian 3- to 10-year-old children (N = 682). Children were asked to make predictions during games of chance. Across experiments, preschool-aged children from all backgrounds consistently displayed a strong wishful thinking bias. However, the bias declined with age.     

Science teacher candidates’ perceptions about roles and nature of scientific models

Background: Scientific models have important roles in science and science education. For scientists, they provide a means for generating new knowledge or function as an accessible summary of scientific studies. In science education, on the other hand, they are accessible representations of abstract concepts, and are also organizational frameworks to teach and learn inaccessible facts. As being indispensable parts of learning and doing science, use of scientific models in science classes should be reinforced. At this point, uncovering pre-service science teachers’ (PSTs) understandings of scientific models are of great importance since they will design and conduct teaching situations for their students. Purpose: The study aimed to provide an answer to the research question: What understandings do PSTs possess about scientific models? Sample: The sample of the study consisted of 14 PSTs enrolled in an Elementary Science Education program in a public university in Ankara, Turkey. Design and methods: Data were collected by using an open-item instrument and semi-structured interviews, and were analyzed by using qualitative data analysis methods. Results: Findings showed that PSTs held fragmented views of models by having informed views in some aspects while having naïve views on others. That is, although they displayed a constructivist orientation by acknowledging the presence of multiple models for the same phenomenon depending on scientists’ perspectives or creativity involved in the production of scientific knowledge, PSTs also expressed logical positivist views by believing that models should be close to the real phenomena that they represent. Findings further revealed that PSTs generally conceptualized models’ materialistic uses, yet they did not think much about their theoretical and conceptual uses. It was observed that roles like reifying and visualizing were overestimated and models were dominantly characterized as three-dimensional representations. Conclusions: It is clear that PSTs, having difficulties in grasping the concept of models, would possibly have problems in planning their lessons effectively and would not develop accurate concepts in their students. These findings apparently support the need for appropriate pedagogic training of PSTs to scientifically reflect on and professionally make use of models in science classes.     

On sabbatical leave

This article argues for the need to differentiate between professional and support careers in fields of science, technology, engineering, mathematics, and medicine (STEMM). Using data from the Longitudinal Study of American Youth (LSAY), the article identifies important differences in the nature of the work and responsibility that differentiate the scientific workforce into a professional component and a support component. Further analysis demonstrates that the kind of training and education needed for STEMM support occupations is generally lower and different than the education needed for entrance into a STEMM profession. The LSAY data indicate that the students planning to enter a STEMM profession tend to make their choice earlier, select appropriate courses and programs to obtain professional credentials, and make a substantially longer commitment to educational preparation than do young adults who enter STEMM support occupations. We suggest some policy implications that may flow from this distinction.     

On indexing in the Web of Science and predicting journal impact factor

We discuss what document types account for the calculation of the journal impact factor (JIF) as published in the Journal Citation Reports (JCR). Based on a brief review of articles discussing how to predict JIFs and taking data differences between the Web of Science (WoS) and the JCR into account, we make our own predictions. Using data by cited-reference searching for Thomson Scientific's WoS, we predict 2007 impact factors (IFs) for several journals, such as Nature, Science, Learned Publishing and some Library and Information Sciences journals. Based on our colleagues' experiences we expect our predictions to be lower bounds for the official journal impact factors. We explain why it is useful to derive one's own journal impact factor.     

Quasi-Bayesian software reliability model with small samples

In traditional Bayesian software reliability models, it was assume that all probabilities are precise. In practical applications the parameters of the probability distributions are often under uncertainty due to strong dependence on subjective information of experts’ judgments on sparse statistical data. In this paper, a quasi-Bayesian software reliability model using interval-valued probabilities to clearly quantify experts’ prior beliefs on possible intervals of the parameters of the probability distributions is presented. The model integrates experts’ judgments with statistical data to obtain more convincible assessments of software reliability with small samples. For some actual data sets, the presented model yields better predictions than the Jelinski-Moranda (JM) model using maximum likelihood (ML). Project supported by the National High-Technology Research and Development Program of China (Grant Nos.2006AA01Z187, 2007AA040605)     

Editorial:On indexing in the Web of Science and predicting journal impact factor

We discuss what document types account for the calculation of the journal impact factor (JIF) as published in the Journal Citation Reports (JCR). Based on a brief review of articles discussing how to predict JIFs and taking data differences between the Web of Science (WoS) and the JCR into account, we make our own predictions. Using data by cited-reference searching for Thomson Scientific's WoS, we predict 2007 impact factors (IFs) for several journals, such as Nature, Science,Learned Publishing and some Library and Information Sciences journals. Based on our colleagues' experiences we expect our predictions to be lower bounds for the official journal impact factors. We explain why it is useful to derive one's own journal impact factor.     

Is Research Possible? A rejoinder to Tooley's ‘On School Choice and Social Class‘

This paper presents a response to some criticisms by James Tooley of our research on parental choice. The paper argues that many of Tooley's criticisms are based on systematic misinterpretations and misreadings of our work, ignorance of the procedures of qualitative research and the use of untenable speculations about sampling, data collection and the interpretation of data. It is suggested that Tooley's criticisms are ideologically driven and that if taken seriously would make qualitative social research impossible.