Assessing Approximate Fit in Categorical Data Analysis

A family of Root Mean Square Error of Approximation (RMSEA) statistics is proposed for assessing the goodness of approximation in discrete multivariate analysis with applications to item response theory (IRT) models. The family includes RMSEAs to assess the approximation up to any level of association of the discrete variables. Two members of this family are RMSEA₂, which uses up to bivariate moments, and the full information RMSEA_n. The RMSEA₂ is estimated using the M₂ statistic of Maydeu-Olivares and Joe (2005, 2006), whereas for maximum likelihood estimation, RMSEA_n is estimated using Pearson's X² statistic. Using IRT models, we provide cutoff criteria of adequate, good, and excellent fit using the RMSEA₂. When the data are ordinal, we find a strong linear relationship between the RMSEA₂ and the Standardized Root Mean Squared Residual goodness-of-fit index. We are unable to offer cutoff criteria for the RMSEA_n as its population values decrease as the number of variables and categories increase.     

Birth of an Abstraction: A Dynamical Systems Account of the Discovery of an Elsewhere Principle in a Category Learning Task

Human participants and recurrent (“connectionist”) neural networks were both trained on a categorization system abstractly similar to natural language systems involving irregular (“strong”) classes and a default class. Both the humans and the networks exhibited staged learning and a generalization pattern reminiscent of the Elsewhere Condition (Kiparsky, 1973). Previous connectionist accounts of related phenomena have often been vague about the nature of the networks’ encoding systems. We analyzed our network using dynamical systems theory, revealing topological and geometric properties that can be directly compared with the mechanisms of non‐connectionist, rule‐based accounts. The results reveal that the networks “contain” structures related to mechanisms posited by rule‐based models, partly vindicating the insights of these models. On the other hand, they support the one mechanism (OM), as opposed to the more than one mechanism (MOM), view of symbolic abstraction by showing how the appearance of MOM behavior can arise emergently from one underlying set of principles. The key new contribution of this study is to show that dynamical systems theory can allow us to explicitly characterize the relationship between the two perspectives in implemented models.     

Noncompliance in hemodialysis patients as measured with the MBHI

Abstract

Fifteen patients beginning maintenance hemodialysis due to renal failure were administered the Millon Behavioral Health Inventory, the Family Environment Scale and Beck Depression Inventory at the outset of treatment. They were then followed regarding their compliance with fluid, diet and medication, using weight change and serum potassium and phosphorus levels. Statistically significant differences were found on independence/dependence conflicts and between patients showing at least moderate noncompliance with weight and/or phosphorus limitations. Significantly higher levels of depression were associated with consistently poorer compliance patterns.     

Risk for Injury in Preschoolers: Relationship to Attention Deficit Hyperactivity Disorder

Parental ratings of preschoolers’ risk for injury, direct assessment of preschoolers’ behavior thought related to risk for injury (e.g., Inattention, impulsivity) and number of documented injuries were examined in preschoolers with Attention Deficit Hyperactivity Disorder (ADHD) and their non-ADHD peers (Control). Of preschoolers with ADHD, 58.3% exhibited behavior which placed them at-risk for physical injury (0% Control), and their performance was significantly poorer on clinic-based tests. Nonetheless, preschoolers with ADHD did not actually sustain significantly more injuries which warranted medical treatment in an emergency department. Although preschoolers with ADHD may be at increased risk for minor injuries, further research is needed to determine whether they more frequently sustain more serious injuries.     

Self-efficacy and affect responses to Sprint Interval Training

ObjectivesTo examine the influence of efficacious beliefs on affective states, enjoyment, and satisfaction after performing a maximal intensity exercise task (i.e. Sprint Interval Training (SIT).DesignAn experimental design was utilized and participants were randomized to a high self-efficacy feedback (HE), low self-efficacy feedback (LE), or no feedback control condition.MethodsCollege-aged men and women (n = 74) performed four 30-second maximal sprints, interspersed with 4-min breaks. Following each sprint interval, participants received either positive, negative or no feedback and SIT self-efficacy, affective states (i.e., positive well-being, psychological distress, fatigue), heart rate (HR), and ratings of perceived exertion (RPE) were assessed. Upon terminating the exercise session, enjoyment and satisfaction were assessed.ResultsThere was group equivalency in HR and RPE among participants of each group. The HE group had significantly less of a decline in SIT self-efficacy from pre to post SIT than both the LE and control groups, and reported significantly higher levels of exercise enjoyment than the other two groups. In addition, the HE and control groups reported significantly higher satisfaction than the LE group.ConclusionsWe recommend that exercisers receive positive feedback statements while performing SIT to minimize the decrease in SIT self-efficacy and to bolster feelings of enjoyment and satisfaction.     

Three dimensions of expertise

Psychologists and philosophers tend to treat expertise as a property of special individuals. These are individuals who have devoted much more time than the general population to the acquisition of their specific expertises. They are often said to pass through stages as they move toward becoming experts, for example, passing from an early stage, in which they follow self-conscious rules, to an expert stage in which skills are executed unconsciously. This approach is ‘one-dimensional’. Here, two extra dimensions are added. They are drawn from the programme known as Studies of Expertise and Experience (SEE) and its ‘Periodic Table of Expertises’. SEE, which is sociological, and/or Wittgensteinian, in inspiration, takes expertise to be the property of groups; there are ‘domains’ of expertise. Under SEE, level of expertise grows with embedding in the society of domain experts; the key is the transmission of domain-specific tacit knowledge. Thus, one extra dimension is degree of exposure to tacit knowledge. Under SEE, domains can be big or small so there can be ‘ubiquitous tacit knowledge’, such as natural-language-speaking or other elements of general social behaviour, which belong to every member of a society. The second extra dimension is, therefore, ‘esotericity’. The resulting three-dimensional ‘expertise-space’ can be explored in a number of ways which reveal the narrowness of the analysis and the mistakes that have been made under the one-dimensional model.     

Observations on the Responsible Development and Use of Computational Models and Simulations

Most previous works on responsible conduct of research have focused on good practices in laboratory experiments. Because computation now rivals experimentation as a mode of scientific research, we sought to identify the responsibilities of researchers who develop or use computational modeling and simulation. We interviewed nineteen experts to collect examples of ethical issues from their experiences in conducting research with computational models. We gathered their recommendations for guidelines for computational research. Informed by these interviews, we describe the respective professional responsibilities of developers and users of computational models in research. In particular, we examine whether developers should disclose the full computational codes, and we explain how developers and users should minimize harms from improper uses of models.     

The role of religion in moderating the impact of life events on material life goals: some evidence in support of terror management theory

How would religion and a life event carrying an existential threat (LEET) jointly impact a person's life goals of becoming wealthy and successful in one's career? Goal reprioritisation, socioemotional selectivity, and gerotranscendence theories predict a shift away from material goals following a LEET, independent of the effect of religion. However, terror management theory (TMT) predicts that the effect of death thoughts depends on one's prevailing cultural values. As religion can be regarded as a culture, it is possible that Christians' and non-believers' material life goals would be differentially altered by LEET. Data from 1259 young Chinese adults reveal no main effect of LEET, but a strong effect of religion. Moreover, there was an interaction effect between LEET and religion on material life goals: LEET weakened material goals for Christians but not for non-believers. These findings suggest that TMT is more suitable than the other theories for predicting life goal changes.