-
PloS One 2023Calibrating model parameters on heterogeneous data can be challenging and inefficient. This holds especially for likelihood-free methods such as approximate Bayesian...
Calibrating model parameters on heterogeneous data can be challenging and inefficient. This holds especially for likelihood-free methods such as approximate Bayesian computation (ABC), which rely on the comparison of relevant features in simulated and observed data and are popular for otherwise intractable problems. To address this problem, methods have been developed to scale-normalize data, and to derive informative low-dimensional summary statistics using inverse regression models of parameters on data. However, while approaches only correcting for scale can be inefficient on partly uninformative data, the use of summary statistics can lead to information loss and relies on the accuracy of employed methods. In this work, we first show that the combination of adaptive scale normalization with regression-based summary statistics is advantageous on heterogeneous parameter scales. Second, we present an approach employing regression models not to transform data, but to inform sensitivity weights quantifying data informativeness. Third, we discuss problems for regression models under non-identifiability, and present a solution using target augmentation. We demonstrate improved accuracy and efficiency of the presented approach on various problems, in particular robustness and wide applicability of the sensitivity weights. Our findings demonstrate the potential of the adaptive approach. The developed algorithms have been made available in the open-source Python toolbox pyABC.
Topics: Computer Simulation; Bayes Theorem; Algorithms
PubMed: 37216372
DOI: 10.1371/journal.pone.0285836 -
Memory (Hove, England) Mar 2023Previous research has shown that neural activity elicited by informative prestimulus cues during encoding differ with respect to subsequent memory outcomes. These...
Previous research has shown that neural activity elicited by informative prestimulus cues during encoding differ with respect to subsequent memory outcomes. These findings indicate prestimulus cues create a "brain state" associated with subsequent memory that, potentially, also has downstream effects benefitting processes associated with successful encoding and subsequent memory performance. However, previous studies have not included the conditions necessary to appropriately test this latter assumption. The present study examines how informative and uninformative prestimulus encoding cues affect memory accuracy for upcoming stimuli compared to a no cue condition. At encoding, participants made one of two semantic judgments on words preceded by an informative prestimulus cue that identified the upcoming semantic judgment, an uninformative prestimulus cue that signalled an upcoming trial but no information about the semantic judgment, or no cue. Dual process estimates of familiarity, but not recollection, demonstrated a graded pattern with the informativeness of the prestimulus cues (i.e., informative > uninformative > no cues). Moreover, both informative and uninformative prestimulus cues enhanced subsequent source memory accuracy for the encoding task compared to the no cue condition. These findings suggest that prestimulus cues can strengthen the processes that support successful memory encoding and benefit subsequent familiarity and source memory.
Topics: Humans; Recognition, Psychology; Cues; Brain; Semantics; Cognition
PubMed: 36546483
DOI: 10.1080/09658211.2022.2156546 -
Entropy (Basel, Switzerland) Dec 2020Information geometry has offered a way to formally study the efficacy of scientific models by quantifying the impact of model parameters on the predicted effects....
Information geometry has offered a way to formally study the efficacy of scientific models by quantifying the impact of model parameters on the predicted effects. However, there has been little formal investigation of causation in this framework, despite causal models being a fundamental part of science and explanation. Here, we introduce causal geometry, which formalizes not only how outcomes are impacted by parameters, but also how the parameters of a model can be intervened upon. Therefore, we introduce a geometric version of "effective information"-a known measure of the informativeness of a causal relationship. We show that it is given by the matching between the space of effects and the space of interventions, in the form of their geometric congruence. Therefore, given a fixed intervention capability, an effective causal model is one that is well matched to those interventions. This is a consequence of "causal emergence," wherein macroscopic causal relationships may carry more information than "fundamental" microscopic ones. We thus argue that a coarse-grained model may, paradoxically, be more informative than the microscopic one, especially when it better matches the scale of accessible interventions-as we illustrate on toy examples.
PubMed: 33375321
DOI: 10.3390/e23010024 -
The British Journal of Developmental... Mar 2022By roughly 6 years of age, children acquire the stereotype that men are more competent than women in science, technology, engineering, and mathematics (STEM),...
By roughly 6 years of age, children acquire the stereotype that men are more competent than women in science, technology, engineering, and mathematics (STEM), potentially leading to greater trust in scientific information provided by men. This study tested whether 3- to 8-year-old children differentially endorsed conflicting information about science and toys presented by male and female informants depicted as a 'man' and 'woman' (Exp1) or 'scientists' (Exp2). Children were expected to endorse toy testimony from gender-matched informants; thus, the key question concerned endorsement of science testimony. In Exp1 (N = 149), boys and girls showed a same-gender informant preference for toy testimony; however, girls endorsed the male informant's testimony more for science than for toys - but only when tested by a male experimenter. In Exp2 (N = 264), boys and girls showed a same-gender preference, irrespective of content. Findings suggest that STEM-related gender stereotypes might lead girls to trust scientific information presented by men over women in certain contexts.
Topics: Child; Child, Preschool; Female; Gender Identity; Humans; Judgment; Male; Mathematics; Play and Playthings; Stereotyping; Trust
PubMed: 34651320
DOI: 10.1111/bjdp.12397 -
PLoS Computational Biology Feb 2023Models of evolution of simple languages have typically assumed full alignment of the speaker and listeners interests, with perfect understanding representing the optimal...
Models of evolution of simple languages have typically assumed full alignment of the speaker and listeners interests, with perfect understanding representing the optimal outcome for both parties. In more realistic settings, communicating individuals will often desire different outcomes from one another. Previous work has shown that misalignment of speaker-listener interests reduces the maximum informativeness among Nash-equilibrium languages, and that multiple equilibrium languages (with different degrees of informativeness) are supported. We study the stochastic evolutionary dynamics of signaling games in which the alignment of speaker-listener interests can vary. We find that increased misalignment of speaker-listener interests is associated with a decrease in information transmission. Moreover, the most common languages to evolve are typically the most informative languages supportable as static Nash equilibria, suggesting a solution to the 'equilibrium selection problem'. In addition, our dynamics reveal the mechanism by which less informative languages evolve: words that previously signaled intense states come to be used hyperbolically for less intense states, with listeners' interpretation of these newly-ambiguous words evolving downward in response. We ground our results in linguistic data on intensifiers such as so and very, words which have unique dynamics-with constant recycling and innovation that match our theoretical results well.
Topics: Humans; Non-alcoholic Fatty Liver Disease; Language; Linguistics; Signal Transduction; Speech Perception
PubMed: 36821533
DOI: 10.1371/journal.pcbi.1010872 -
Current Topics in Behavioral... 2018Schizophrenia is considered to develop as a consequence of genetic and environmental factors impacting on brain neural systems and circuits during vulnerable... (Review)
Review
Schizophrenia is considered to develop as a consequence of genetic and environmental factors impacting on brain neural systems and circuits during vulnerable neurodevelopmental periods, thereby resulting in symptoms in early adulthood. Understanding of the impact of schizophrenia risk factors on brain biology and behaviour can help in identifying biologically relevant pathways that are attractive for informing clinical studies and biomarker development. In this chapter, we emphasize the importance of adopting a reciprocal forward and reverse translation approach that is iteratively updated when additional new information is gained, either preclinically or clinically, for offering the greatest opportunity for discovering panels of biomarkers for the diagnosis, prognosis and treatment of schizophrenia. Importantly, biomarkers for identifying those at risk may inform early intervention strategies prior to the development of schizophrenia.Given the emerging nature of this approach in the field, this review will highlight recent research of preclinical biomarkers in schizophrenia that show the most promise for informing clinical needs with an emphasis on relevant imaging, electrophysiological, cognitive behavioural and biochemical modalities. The implementation of this reciprocal translational approach is exemplified firstly by the production and characterization of preclinical models based on the glutamate hypofunction hypothesis, genetic and environmental risk factors for schizophrenia (reverse translation), and then the recent clinical recognition of the thalamic reticular thalamus (TRN) as an important locus of brain dysfunction in schizophrenia as informed by preclinical findings (forward translation).
Topics: Biomarkers; Brain; Glutamic Acid; Humans; Schizophrenia
PubMed: 29721851
DOI: 10.1007/7854_2018_48 -
Tobacco Control Apr 2017In reviewing the first set of pictorial warning labels in the USA, the courts equated textual labels with facts and information, and images with emotion. This study...
BACKGROUND
In reviewing the first set of pictorial warning labels in the USA, the courts equated textual labels with facts and information, and images with emotion. This study tested the differences in perceived informativeness and emotion between textual and pictorial cigarette warning labels.
METHODS
An online study with 1838 US adults who were non-smokers (n=764), transitioning smokers (quit smoking in the past 2 years or currently trying to quit, n=505) or current smokers (n=569). Each participant evaluated 9 out of 81 text and pictorial cigarette warning labels. Participants reported to what extent they perceived the label as informative and factual and the negative emotions they felt while looking at each label. We used linear mixed models to account for the nesting of multiple observations within each participant.
RESULTS
There were no significant differences in perceived informativeness between textual (mean 6.15 on a 9-point scale) and pictorial labels (6.14, p=0.80, Cohen's d=0.003). Textual labels evoked slightly less emotion (4.21 on a 9-point scale) than pictorial labels (4.42, p<0.001, Cohen's d=0.08). Perceived informativeness and emotion were strongly correlated (Pearson r=0.53, p<0.001).
CONCLUSION
Our findings contradict courts' conclusions that pictorial messages are emotional and not factual. Pictorial labels are rated as informative and factual, textual labels evoke emotion, and emotionality and informativeness are strongly correlated. These findings serve as evidence for the Food and Drug Administration (FDA) to counteract the claim that pictorial warning labels, by definition, are not 'purely factual and uncontroversial'.
PubMed: 28428337
DOI: 10.1136/tobaccocontrol-2016-053563 -
Postgraduate Medicine Apr 2017To review the literature evaluating the psychometric properties of parent and teacher informants relative to a gold-standard ADHD diagnosis in pediatric populations. (Review)
Review
OBJECTIVE
To review the literature evaluating the psychometric properties of parent and teacher informants relative to a gold-standard ADHD diagnosis in pediatric populations.
METHOD
We included studies that included both a parent and teacher informant, a gold-standard diagnosis, and diagnostic accuracy metrics. Potential confounds were evaluated. We also assessed the 'OR' and the 'AND' rules for combining informant reports.
RESULTS
Eight articles met inclusion criteria. The diagnostic accuracy for predicting gold standard ADHD diagnoses did not differ between parents and teachers. Sample size, sample type, participant drop-out, participant age, participant gender, geographic area of the study, and date of study publication were assessed as potential confounds.
CONCLUSION
Parent and teachers both yielded moderate to good diagnostic accuracy for ADHD diagnoses. Parent reports were statistically indistinguishable from those of teachers. The predictive features of the 'OR' and 'AND' rules are useful in evaluating approaches to better integrating information from these informants.
Topics: Age Factors; Attention Deficit Disorder with Hyperactivity; Child; Humans; Parents; Reproducibility of Results; Residence Characteristics; School Teachers; Sex Factors
PubMed: 28271921
DOI: 10.1080/00325481.2017.1288064