The Quest for Signals in Noise: Leveraging Experiential Variation to Identify Bilingual Phenotypes

Increasing evidence suggests that bilingualism does not, in itself, result in a particular pattern of response, revealing instead a complex and multidimensional construct that is shaped by evolutionary and ecological sources of variability. Despite growing recognition of the need for a richer characterization of bilingual speakers and of the different contexts of language use, we understand relatively little about the boundary conditions of putative “bilingualism” effects. Here, we review recent findings that demonstrate how variability in the language experiences of bilingual speakers, and also in the ability of bilingual speakers to adapt to the distinct demands of different interactional contexts, impact interactions between language use, language processing, and cognitive control processes generally. Given these findings, our position is that systematic variation in bilingual language experience gives rise to a variety of phenotypes that have different patterns of associations across language processing and cognitive outcomes. The goal of this paper is thus to illustrate how focusing on systematic variation through the identification of bilingual phenotypes can provide crucial insights into a variety of performance patterns, in a manner that has implications for previous and future research.

DIAGNOSTIC STRATEGIES FOR ENDEMIC CORONAVIRUS DISEASE 2019 (COVID-19): RAPID ANTIGEN TESTS, REPEAT TESTING, AND PREVALENCE BOUNDARIES

Context. – Coronavirus disease 2019 (COVID-19) rapid antigen tests generate intrinsically fast, inherently spatial, and immediately actionable results. They quickly confirm COVID-19, but weakly rule out infection. Test performance depends on prevalence and testing protocol. Both affect predictive values. Objectives. – To use original mathematics and visual logistics for interpreting COVID-19 rapid antigen test performance patterns, gauge the influence of prevalence, and evaluate repeated testing. Design. – Mathematica and open access software helped graph relationships, perform recursive computations, and compare performance patterns. PubMed retrieved articles addressing endemic COVID-19. Results. – Tiered sensitivity/specificity comprise: T1) 90%/95%; T2) 95%/97.5%; and T3) 100%/≥99%, respectively. Performance of self- and home antigen tests with Food and Drug Administration Emergency Use Authorization peaks in low prevalence. Fall-off in performance appears with increasing prevalence because suboptimal sensitivity creates false negatives. The rate of false omissions limits clinical use because of prevalence boundaries based on tolerance for risk. Mathematical analysis supports testing twice to improve predictive values and extend prevalence boundaries nearly to levels of herd immunity. Conclusions. – COVID-19 is quickly becoming endemic. Suboptimal sensitivity of rapid antigen tests limits performance in high prevalence. Risk of contagion in packed spaces (e.g., airplanes) might be avoided with dual testing 36 hours apart, allowing time for viral load to increase. Awareness of community prevalence and proof of improved performance with repeated testing will help manage COVID-19 risk, while meeting rapid decision-making needs for highly contagious and new variants (e.g., Delta). New COVID-19 variants call for high quality, low cost, readily accessible, fast, user friendly, and ubiquitous point-of-care testing.

A parallelization model for performance characterization of Spark Big Data jobs on Hadoop clusters

This article proposes a new parallel performance model for different workloads of Spark Big Data applications running on Hadoop clusters. The proposed model can predict the runtime for generic workloads as a function of the number of executors, without necessarily knowing how the algorithms were implemented. For a certain problem size, it is shown that a model based on serial boundaries for a 2D arrangement of executors can fit the empirical data for various workloads. The empirical data was obtained from a real Hadoop cluster, using Spark and HiBench. The workloads used in this work were included WordCount, SVM, Kmeans, PageRank and Graph (Nweight). A particular runtime pattern emerged when adding more executors to run a job. For some workloads, the runtime was longer with more executors added. This phenomenon is predicted with the new model of parallelisation. The resulting equation from the model explains certain performance patterns that do not fit Amdahl’s law predictions, nor Gustafson’s equation. The results show that the proposed model achieved the best fit with all workloads and most of the data sizes, using the R-squared metric for the accuracy of the fitting of empirical data. The proposed model has advantages over machine learning models due to its simplicity, requiring a smaller number of experiments to fit the data. This is very useful to practitioners in the area of Big Data because they can predict runtime of specific applications by analysing the logs. In this work, the model is limited to changes in the number of executors for a fixed problem size.

Childhood Academic Performance: A Potential Marker of Genetic Liability to Autism

Background: Autism spectrum disorder (ASD) is a highly heritable, genetically complex neurodevelopmental disorder. Genetic liability is often expressed among relatives through subclinical, genetically meaningful traits, or endophenotypes. Studies of parents of individuals with ASD suggest important differences from controls in language-related skills in particular, including evidence that such differences may emerge in childhood, that may serve as early markers of genetic liability to ASD. This study investigated whether developmental academic profiles may be evident among clinically unaffected siblings of individuals with ASD, and possibly constitute developmental endophenotypic markers of ASD genetic risk among relatives. Methods: Longitudinal, archival academic testing records were studied to characterize developmental profiles in the domains of language, reading, and math, among clinically unaffected siblings of individuals with ASD. Relationships were explored between siblings’ childhood academic profiles and subclinical ASD-related traits, and the familiality of such traits. Results: Results revealed relatively lower performance in language-related academic skills among siblings of individuals with ASD, mirroring patterns previously reported among parents. Relationships were detected between siblings’ academic performance patterns and subclinical ASD-related traits in themselves and their parents, and with symptom severity in their sibling with ASD. Language phenotypes were associated in mother-sibling dyads, and rigidity and math performance associated in father-child dyads. Limitations: Data from this study represent a relatively small and racially homogenous group of siblings of individuals with ASD, and as such, replication in a larger more diverse sample should be completed to increase generalizability. Conclusions: Distinctive profiles of academic development were evident in siblings in language-related skills, mirroring prior findings in parents, suggesting specific and subtle phenotypes that may represent early-emerging indicators of genetic liability to ASD and are measurable in first-degree relatives using standardized academic testing. Results also suggest differential intergenerational transmission of ASD-related traits between mothers and fathers.