Latent Response Research Articles

Examining Parameter Invariance in a General Diagnostic Classification Model.

The present study aimed at investigating invariance of a diagnostic classification model (DCM) for reading comprehension across gender. In contrast to models with continuous traits, diagnostic classification models inform mastery of a finite set of latent attributes, e.g., vocabulary or syntax in the reading context, and allow to provide fine grained feedback to learners and instructors. The generalized deterministic, noisy “and” gate (G-DINA) model was fit to item responses of 1000 male and female individuals to a high-stakes reading comprehension test. Use of the G-DINA model allowed for minimal assumption on the relationship of latent attribute profiles and item-specific response probabilities, i.e., the G-DINA model can represent compensatory or non-compensatory relationships of latent attributes and response probabilities. Item parameters were compared across the two samples, and only a small number of item parameters were statistically different between the two groups, corroborating the result of a formal measurement invariance test based on the multigroup G-DINA model. Neither correlations between latent attributes were significantly different across the two groups, nor mastery probabilities for any of the attributes. Model selection at item level showed that from among the 18 items that required multiple attributes, 16 items picked different rules (DCMs) across the groups. While this seems to suggest that the relationship among the attributes of reading comprehension differs across the two groups, a closer inspection of the rules picked by the items showed that almost in all cases the relationships were very similar. If a compensatory DCM was suggested by the G-DINA framework for an item in the female group, a model belonging to the same family resulted for the male group.

Knowledge Discovery in Nanophotonics Using Geometric Deep Learning

Herein, a new approach for using the intelligence aspects of artificial intelligence for knowledge discovery rather than device optimization in electromagnetic (EM) nanostructures is presented. This approach uses training data obtained through full‐wave EM simulations of a series of nanostructures to train geometric deep learning algorithms to assess the range of feasible responses as well as the feasibility of a desired response from a class of EM nanostructures. To facilitate the knowledge discovery, this approach combines the dimensionality reduction technique with convex‐hull and one‐class support‐vector‐machine (SVM) algorithms to find the range of the feasible responses in the latent response space of the EM nanostructure. More importantly, the one‐class SVM algorithm can be trained to provide the degree of feasibility of a response from a given nanostructure. This important information can be used to modify the initial structure to an alternative one that can enable an initially unfeasible response. To show the applicability of this approach, it is applied to two important classes of binary metasurfaces (MSs), formed by an array of plasmonic nanostructures, and periodic MSs formed by an array of dielectric nanopillars. These theoretical and experimental results confirm the unique features of this approach for knowledge discovery in EM nanostructures.

A hierarchical latent response model for inferences about examinee engagement in terms of guessing and item-level non-response.

In low-stakes assessments, test performance has few or no consequences for examinees themselves, so that examinees may not be fully engaged when answering the items. Instead of engaging in solution behaviour, disengaged examinees might randomly guess or generate no response at all. When ignored, examinee disengagement poses a severe threat to the validity of results obtained from low-stakes assessments. Statistical modelling approaches in educational measurement have been proposed that account for non-response or for guessing, but do not consider both types of disengaged behaviour simultaneously. We bring together research on modelling examinee engagement and research on missing values and present a hierarchical latent response model for identifying and modelling the processes associated with examinee disengagement jointly with the processes associated with engaged responses. To that end, we employ a mixture model that identifies disengagement at the item-by-examinee level by assuming different data-generating processes underlying item responses and omissions, respectively, as well as response times associated with engaged and disengaged behaviour. By modelling examinee engagement with a latent response framework, the model allows assessing how examinee engagement relates to ability and speed as well as to identify items that are likely to evoke disengaged test-taking behaviour. An illustration of the model by means of an application to real data is presented.

Beliefs and needs of academic teachers: a latent class analysis

In the last few years, academic guidance services of the Italian universities have been increasingly involved in planning and organising training courses for academic teachers to improve the quality of teaching. In such a context, it is important to collect information on the teachers’ opinions about their belief on traditional and innovative approaches to teaching and learning evaluation as well as on their need of support to take effective teaching practices. In this contribution we aim at studying the structure of academic teachers’ population of an Italian university in order to detect groups of teachers that are homogenous in terms of beliefs and needs. As beliefs and needs may be conceptualised in terms of an unobservable (or latent) variable composed of multiple traits related with different aspects of beliefs and needs (e.g., passion for teaching, beliefs about teaching methods, ...), the proposed analysis is based on a multidimensional Latent Class Item Response Theory model. This type of model allows us to classify teachers in latent classes with respect to their beliefs and needs in the academic didactic activities. Moreover, it also allows us to identify specific aspects of teaching with respect to which academic teachers tend to disagree/agree, and to relate beliefs and needs of teachers with their individual characteristics. The study involves a sample of academic teachers coming from the University of Florence (IT) that took part in a survey based on a new questionnaire composed of Likert-type items concerning beliefs and needs on several aspects of teaching.

Facilitating Calibration in High-Speed BCI Spellers via Leveraging Cross-Device Shared Latent Responses.

This paper proposes a novel device-to-device transfer-learning algorithm for reducing the calibration cost in a steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) speller by leveraging electroencephalographic (EEG) data previously acquired by different EEG systems. The transferring is done by projecting the scalp-channel EEG signals onto a shared latent domain across devices. Three spatial filtering techniques, including channel averaging, canonical correlation analysis (CCA), and task-related component analysis (TRCA), were employed to extract the shared responses from different devices. The transferred data were integrated into a template-matching-based algorithm to detect SSVEPs. To evaluate its transferability, this paper conducted two sessions of simulated online BCI experiments with ten subjects using 40 visual stimuli modulated by joint frequency-phase coding method. In each session, two different EEG devices were used: first, the Quick-30 system (Cognionics, Inc.) with dry electrodes, and second, the ActiveTwo system (BioSemi, Inc.) with wet electrodes. The proposed method with CCA- and TRCA-based spatial filters achieved significantly higher classification accuracy compared with the calibration-free standard CCA-based method. This paper validated the feasibility and effectiveness of the proposed method in implementing calibration-free SSVEP-based BCIs. The proposed method has great potentials to enhance practicability and usability of real-world SSVEP-based BCI applications by leveraging user-specific data recorded in previous sessions even with different EEG systems and montages.

Hybrid multiobjective optimization algorithm based on multivariate mean square error and fuzzy decision maker

This work presents the turning process of AISI H13 hardened steel with the PCBN 7025 tool, considering six output variables: tool life, machining total cost, surface roughness, machining force, sound pressure level, and specific cutting energy. Several problems are encountered in engineering processes that have adverse effects on the reliability of complex engineering systems. Hence, the aim of this work is to optimize the hardened steel turning process by applying mathematical methods to reduce dimensionality and eliminate the correlation between the multiple responses. The resultant latent response surfaces and their respective targets constitute the normalized multivariate mean square error (MMSE) function that is minimized by the normal boundary intersection (NBI) method. Furthermore, a fuzzy algorithm is applied to identify the best solution from several feasible solutions of the Pareto frontier that is compared with the performances of normalized normal constraint, arc homotopy length, global criterion method, and desirability method. The results show that NBI-MMSE has a higher performance than the other methods. In addition, NBI-MMSE is tested with benchmark functions to evaluate its effectiveness and robustness. Therefore, NBI-MMSE identifies the dynamics of the turning process of AISI H13 steel by revealing the optimal solutions for the input process parameters.

Using generalizability theory with continuous latent response variables.

In this article, we illustrate ways in which generalizability theory (G-theory) can be used with continuous latent response variables (CLRVs) to address problems of scale coarseness resulting from categorization errors caused by representing ranges of continuous variables by discrete data points and transformation errors caused by unequal interval widths between those data points. The mechanism to address these problems is applying structural equation modeling (SEM) as a tool in deriving variance components needed to estimate indices of score consistency and validity. Illustrations include quantification of multiple sources of measurement error, use of non-nested and nested designs, derivation of indices of consistency for norm- and criterion-referenced interpretation of scores, estimation of effects when changing measurement procedures and designs, and disattenuation of correlation coefficients for measurement error. These illustrations underscore the effectiveness of G-theory with continuous latent response variables in providing stable indices of reliability and validity that are reasonably independent of the number of original scale points used, unevenness of scale intervals, and average degree of item skewness. We discuss general distinctions in reliability estimation within G-theory, SEM, and classical test theory; make specific recommendations for using G-theory on raw score and CLRV metrics; and provide computer code in an online supplement for doing all key analyses demonstrated in the article using R and Mplus. (PsycINFO Database Record (c) 2019 APA, all rights reserved).

Avian responses to aircraft in an airport environment

ABSTRACTUnderstanding how free‐ranging birds react to approaching aircraft can provide the foundation for predicting and mitigating risk of bird strikes. We characterized responses by avian species to aircraft (propeller‐driven, jet, rotorcraft) approach (taxi, takeoffs, landings) at Burke Lakefront Airport, Cleveland, Ohio, USA, from June 2015 through September 2016. Based on opportunistic observations from fixed points on the airfield, we quantified 208 bird‐aircraft interactions across 16 species. However, we focused analyses on 5 species with >10 bird‐aircraft interactions (n = 176): American kestrel (Falco sparvarious), European starling (Sturnus vulgaris), killdeer (Charadrius vociferous), mourning dove (Zenaida macroura), and red‐winged blackbird (Agelaius phoeniceus). We evaluated the common measure of perceived risk, flight‐initiation distance (FID), and the likelihood of response (i.e., a measureable FID), each relative to species, flock size, flock altitude, perch height, aircraft approach category, and airframe type. We also assessed effects of light, temperature, and wind speed. None of our predictors contributed to FID. Across the 5 species, birds exposed to direct aircraft approach were ≥2 times more likely to initiate escape as those approached tangentially. The larger mourning dove was, by a factor ≥2, more likely to initiate escape response. As flock size increased, birds were more likely to initiate escape response. Birds were >2 times more likely to initiate escape when approaches involved jets relative to propeller‐driven airframes. Because smaller species were less likely to initiate escape response, thereby enhancing the possibility of latent responses close to aircraft, we suggest that airport biologists should not disregard management to reduce strike hazards posed by smaller species. Further, bird habitat on airfields and fixtures that concentrate bird use in a direct line of aircraft approach will inherently increase the frequency of bird‐aircraft interactions. Finally, our findings on flock size and airframe type underscore the need for continued efforts to develop methods to enhance avian detection of and response to approaching aircraft. © 2019 The Wildlife Society.

DualFog-IoT: Additional Fog Layer for Solving Blockchain Integration Problem in Internet of Things

Integration of blockchain and Internet of Things (IoT) to build a secure, trusted and robust communication technology is currently of great interest for research communities and industries. But challenge is to identify the appropriate position of blockchain in current settings of IoT with minimal consequences. In this article we propose a blockchain-based DualFog-IoT architecture with three configuration filter of incoming requests at access level, namely: Real Time, Non-Real Time, and Delay Tolerant Blockchain applications. The DualFog-IoT segregate the Fog layer into two: Fog Cloud Cluster and Fog Mining Cluster. Fog Cloud Cluster and the main cloud datacenter work in a tandem similar to existing IoT architecture for real-time and non-real-time application requests, while the additional Fog Mining Cluster is dedicated to deal with only Delay Tolerant Blockchain application requests. The proposed DualFog-IoT is compared with existing centralized datacenter based IoT architecture. Along with the inherited features of blockchain, the proposed model decreases system drop rate, and further offload the cloud datacenter with minimal upgradation in existing IoT ecosystem. The reduced computing load from cloud datacenter doesn't only help in saving the capital and operational expenses, but it is also a huge contribution for saving energy resources and minimizing carbon emission in environment. Furthermore, the proposed DualFog-IoT is also being analyzed for optimization of computing resources at cloud level, the results presented shows the feasibility of proposed architecture under various ratios of incoming RT and NRT requests. However, the integration of blockchain has its footprints in terms of latent response for delay tolerant blockchain applications, but real-time and non-real-time requests are gracefully satisfying the service level agreement.

A Comprehensive IRT Approach for Modeling Binary, Graded, and Continuous Responses With Error in Persons and Items.

Dual item response theory (IRT) models in which items and individuals have different amounts of measurement error have been proposed in the literature. Any developments in these models, however, are feasible only for continuous responses. This article discusses a comprehensive dual modeling approach, based on underlying latent response variables, from which specific models for continuous, graded, and binary responses are obtained. Procedures for (a) calibrating the items, (b) scoring individuals, (c) assessing model appropriateness, and (d) assessing measurement precision are discussed for all the resulting models. Simulation results suggest that the proposal is quite feasible. A practical illustration is given with an empirical example in the personality domain.

Sleep quality mediating the association of personality traits and quality of life among underground workers and surface workers of Chinese coal mine: A multi-group SEM with latent response variable mediation analysis

The aim of this study was to examine whether the association between personality traits and quality of life (QOL) was mediated by sleep quality in coal miners and to explore whether the relationship between surface workers and underground workers was different. The cross-sectional study including 3090 coal mine workers aged 20 to 65 years from Shanxi province of China was conducted. Personality traits, QOL and sleep quality were respectively assessed using Eysenck Personality Questionnaire-Revised Short Scale for Chinese, World Health Organization Quality of Life-BREF, and sleep status questionnaire. The multi-group SEM with latent response variable method was conducted to evaluate the association between personality traits and QOL mediated by sleep quality. The difference of indirect effects between two subgroups was examined by Wald chi-square test. For surface workers, underground workers and overall sample, passive personality traits had a negative impact on QOL through poor sleep quality, whereas the active personality traits acted the opposite. However, the difference of indirect effect between two subgroups was not statistically significant. These results indicated that sleep quality may act as a partial mediator in the relationship between personality traits and QOL, and the relationship may not be affected by working environment.

First-episode psychosis in treatment-resistant schizophrenia: a cross-sectional study of a long-term follow-up cohort

BackgroundApproximately one-third of schizophrenia patients eventually develop treatment-resistant schizophrenia (TRS). Although the time course of TRS development varies from patient to patient, the details of these variations have not been clarified. The present study compared the duration of time required to achieve control of the first-episode psychosis (FEP) between patients who went on to develop TRS and those who did not, in order to determine whether a bifurcation point exists for the transition to TRS.MethodsThe present study included 271 schizophrenia patients. Based on the clinical assessment, each patient was assigned to a TRS (n = 79) or Non-TRS group (n = 182). Clinical factors relating to FEP treatment such as the duration of initial hospital admission and the degree of improvement were retrospectively identified.ResultsThere was no significant difference in the duration of initial hospital admission (defined as the time from treatment introduction to successful discharge) between the two groups (mean of 87.9 days for TRS vs. 53.3 days for Non-TRS). The degree of improvement during initial hospital admission of the TRS group was significantly lower than that of the Non-TRS group (Global Assessment of Functioning (GAF) of 50 points for TRS vs. 61 points for Non-TRS). Approximately half of the TRS patients showed an acute onset pattern and longer hospital admission (mean 169 days) for their FEP. The other half of TRS patients needed no hospital admission, indicating an insidious onset pattern with no clear psychotic episode and treatment introduction without hospital admission.ConclusionsFuture TRS patients can have difficulty in improvement during their FEP. There appear to be two distinct patterns for the development of TRS. One pattern is characterized by refractory positive symptoms and a longer period to control the first psychosis; the other shows latent or insidious onset and poor response to the initial treatment.

Joint longitudinal model development: application to exposure–response modeling of ACR and DAS scores in rheumatoid arthritis patients treated with sirukumab

Exposure-response modeling is important to optimize dose and dosing regimen in clinical drug development. The joint modeling of multiple endpoints is made possible in part by recent progress in latent variable indirect response (IDR) modeling for ordered categorical endpoints. This manuscript presents the results of joint modeling of continuous and ordered categorical endpoints in the latent variable IDR modeling framework through the sharing of model parameters, with an application to the exposure-response modeling of sirukumab. Sirukumab is a human anti- interleukin-6 (IL-6) monoclonal antibody that binds soluble human IL-6 thus blocking IL-6 signaling, which plays a major role in the pathophysiology of rheumatoid arthritis (RA). A phase 2 clinical trial was conducted in patients with active RA despite methotrexate therapy, who received subcutaneous (SC) administration of either placebo or sirukumab of 25, 50 or 100mg every 4weeks (q4w) or 100mg every 2weeks (q2w). Major efficacy endpoints were the 20, 50, and 70% improvement in the American College of Rheumatology (ACR20, ACR50, and ACR70) disease severity criteria, and the 28-joint disease activity score using C-reactive protein (DAS28). The ACR endpoints were treated as ordered categorical and DAS28 as continuous. The results showed that, compared with the common approach of separately modeling the endpoints, the joint model could describe the observed data better with fewer parameters through the sharing of random effects, and thus more precisely characterize the dose-response relationship. The implications on future dose and dosing regimen optimization are discussed in contrast with those from landmark analysis.

Application and Comparison of Multidimensional Latent Class Item Response Theory on Clustering Items in Comprehension Tests

Application and Comparison of Multidimensional Latent Class Item Response Theory on Clustering Items in Comprehension Tests

Supervised Dirichlet Process Mixtures of Principal Component Analysis

Abstract We introduce probabilistic principal component analysis (PPCA) into Dirichlet Process Mixtures of Generalized Linear Models (DPGLM) and propose a new model called Supervised Dirichlet Process Mixtures of Principal Component Analysis (SDPM-PCA). In SDPM-PCA, we assume covariates and response variable are generated separately through the latent variable of PPCA, and nonparametrically modeled using the Dirichlet Process Mixture. By jointly learning the latent variable, cluster label and response variable, SDPM-PCA performs locally dimensionality reduction within each mixture component, and learns a supervised model based on the latent variable. In this way, SDPM-PCA improves the performance of both dimensionality reduction and prediction on high-dimensional data with all advantages of DPGLM. We also develop an inference algorithm for SDPM-PCA based on variational inference, which provides faster training speed and deterministic approximation compared with sampling algorithms based on MCMC method. Finally, we instantiate SDPM-PCA in regression problem with a Bayesian linear regression model. We test it on several real-world datasets and compare the prediction performance with DPGLM and other regular regression model. Experiment results show that by setting properly latent dimension number, SDPM-PCA would provide better prediction performance on high-dimensional regression problem and avoid the curse of dimensionality problem in DPGLM.

Gene selection for microarray gene expression classification using Bayesian Lasso quantile regression

Gene selection has been proven to be an effective way to improve the results of many classification methods. However, existing gene selection techniques in binary classification regression are sensitive to outliers of the data, heteroskedasticity or other anomalies of the latent response. In this paper, we propose a new Bayesian hierarchical model to overcome these problems in a relatively straightforward way. In particular, we propose a new Bayesian Lasso method that employs a skewed Laplace distribution for the errors and a scaled mixture of uniform distribution for the regression parameters, together with Bayesian MCMC estimation. Comprehensive comparisons between our proposed gene selection method and other competitor methods are performed experimentally, depending on four benchmark gene expression datasets. The experimental results prove that the proposed method is very effective for selecting the most relevant genes with high classification accuracy.

Robustness of Parameter Estimation to Assumptions of Normality in the Multidimensional Graded Response Model

ABSTRACTA central assumption that is implicit in estimating item parameters in item response theory (IRT) models is the normality of the latent trait distribution, whereas a similar assumption made in categorical confirmatory factor analysis (CCFA) models is the multivariate normality of the latent response variables. Violation of the normality assumption can lead to biased parameter estimates. Although previous studies have focused primarily on unidimensional IRT models, this study extended the literature by considering a multidimensional IRT model for polytomous responses, namely the multidimensional graded response model. Moreover, this study is one of few studies that specifically compared the performance of full-information maximum likelihood (FIML) estimation versus robust weighted least squares (WLS) estimation when the normality assumption is violated. The research also manipulated the number of nonnormal latent trait dimensions. Results showed that FIML consistently outperformed WLS when there were one or multiple skewed latent trait distributions. More interestingly, the bias of the discrimination parameters was non-ignorable only when the corresponding factor was skewed. Having other skewed factors did not further exacerbate the bias, whereas biases of boundary parameters increased as more nonnormal factors were added. The item parameter standard errors recovered well with both estimation algorithms regardless of the number of nonnormal dimensions.

Modeling and Measuring Scale Attraction Effects: An Application to Charitable Donations

Charities seeking donations typically employ an “appeals scale,” a roster of suggested amounts presented to potential donors, along with an “Other” category. Yet little is known about how the amounts comprising appeals scales affect whether a donation is made and, if so, jointly exert “pull” on its magnitude. Availing of multi-year panel data and a field experiment, we develop a model accounting for individual level donation incidence, amount, and appeals scale attraction effects. The model incorporates heterogeneity across donors in both upward and downward scale point attraction, as well as in donation patterns (e.g., seasonality), and accommodates multiple operationalizations of internal and external referents to summarize the effects of prior donation history and scale points, respectively. Overall results suggest that scale points do exert substantial attraction effects; that these vary markedly across donors; that they are in fact referent-based effects; that donors are more easily persuaded to give less than more; and that, while all scale points exert pull, influence wanes with distance. The modeling framework applies not only in donation contexts, but whenever an ordered categorical scale is used to collect data regarding an underlying latent response.

Chromatin dynamics underlying latent responses to xenobiotics.

Pleiotropic xenobiotics can trigger dynamic alterations in mammalian chromatin structure and function but many of these are likely non-adverse and simply reflect short-term changes in DNA transactions underlying normal homeostatic, adaptive and protective cellular responses. However, it is plausible that a subset of xenobiotic-induced perturbations of somatic tissue or germline epigenomes result in delayed-onset and long-lasting adverse effects, in particular if they occur during critical stages of growth and development. These could include reprogramming, dedifferentiation, uncontrolled growth, and cumulative toxicity effects through molecular memory of prior xenobiotic exposures or altered susceptibility to subsequent xenobiotic exposures. Here we discuss the current evidence for epigenetic mechanisms underlying latent responses to xenobiotics, and the potential for identifying molecular epigenetic changes that are prodromal to overt morphologic or functional toxicity phenotypes.

Generating Multivariate Ordinal Data via Entropy Principles.

When conducting robustness research where the focus of attention is on the impact of non-normality, the marginal skewness and kurtosis are often used to set the degree of non-normality. Monte Carlo methods are commonly applied to conduct this type of research by simulating data from distributions with skewness and kurtosis constrained to pre-specified values. Although several procedures have been proposed to simulate data from distributions with these constraints, no corresponding procedures have been applied for discrete distributions. In this paper, we present two procedures based on the principles of maximum entropy and minimum cross-entropy to estimate the multivariate observed ordinal distributions with constraints on skewness and kurtosis. For these procedures, the correlation matrix of the observed variables is not specified but depends on the relationships between the latent response variables. With the estimated distributions, researchers can study robustness not only focusing on the levels of non-normality but also on the variations in the distribution shapes. A simulation study demonstrates that these procedures yield excellent agreement between specified parameters and those of estimated distributions. A robustness study concerning the effect of distribution shape in the context of confirmatory factor analysis shows that shape can affect the robust [Formula: see text] and robust fit indices, especially when the sample size is small, the data are severely non-normal, and the fitted model is complex.