Modeling protective action decision-making in earthquakes by using explainable machine learning and video data.

Earthquakes pose substantial threats to communities worldwide. Understanding how people respond to the fast-changing environment during earthquakes is crucial for reducing risks and saving lives. This study aims to study people's protective action decision-making in earthquakes by leveraging explainable machine learning and video data. Specifically, this study first collected real-world CCTV footage and video postings from social media platforms, and then identified and annotated changes in the environment and people's behavioral responses during the M7.1 2018 Anchorage earthquake. By using the fully annotated video data, we applied XGBoost, a widely-used machine learning method, to model and forecast people's protective actions (e.g., drop and cover, hold on, and evacuate) during the earthquake. Then, explainable machine learning techniques were used to reveal the complex, nonlinear relationships between different factors and people's choices of protective actions. Modeling results confirm that social and environmental cues played critical roles in affecting the probability of different protective actions. Certain factors, such as the earthquake shaking intensity and number of people shown in the environment, displayed evident nonlinear relationships with the probability of choosing to evacuate. These findings can help emergency managers and policymakers design more effective protective action recommendations during earthquakes.

The many faces of moralized self-control: Puritanical morality is not reducible to cooperation concerns.

Fitouchi et al.'s moral disciplining approach highlights the significant role social evaluations of self-control appear to play in human moral judgment. At the same time, attributing the wide range of puritanical concerns to a singular focus on self-control seems unwarranted. A more pluralistic approach would enrich understanding of moral judgment in all its cultural and historical diversity.

Restricting diet for perceived health benefit: A mixed-methods exploration of peripartum food taboos in rural Cambodia.

Food taboos encompass food restrictions practiced by a group that go beyond individual preferences. During pregnancy and lactation, food taboos may contribute to inadequate nutrition and poor maternal and infant health. Restriction of specific fish, meat, fruits and vegetables is common among peripartum women in many Southeast Asian countries, but data from Cambodia are lacking. In this mixed-methods study, 335 Cambodian mothers were asked open-ended questions regarding dietary behaviours during pregnancy and up to 24 weeks postpartum. Descriptive statistics and content analysis were used to characterize food taboos and multiple logistic regression analyses were conducted to identify predictors of this practice. Participants were 18-44 years of age, all of Khmer ethnicity and 31% were primiparous. Sixty-six per cent of women followed food taboos during the first 2 weeks postpartum, whereas ~20% of women restricted foods during other peripartum periods. Pregnancy taboos were often beneficial, including avoidance of sugar-sweetened beverages, coffee and alcohol. Conversely, postpartum avoidances typically included nutrient-dense foods such as fish, raw vegetables and chicken. Food taboos were generally followed to support maternal and child health. No significant predictors of food taboos during pregnancy were identified. Postpartum, each additional live birth a woman had reduced her odds of following food taboos by 24% (odds ratio [95% confidence interval]: 0.76 [0.61-0.95]). Specific food taboo practices and rationales varied greatly between women, suggesting that food taboos are shaped less by a strict belief system within the Khmer culture and more by individual or household understandings of food and health during pregnancy and postpartum.

Dwell times showcase how goal structure informs preschoolers' analysis of unfolding motion patterns.

Using Hard et al.'s (2011) dwell-time paradigm, 85 preschoolers (aged 2.5-4.5; 43 female; primarily from white families) advanced at their own pace through one of three slideshows. All slideshows depicted an actor reaching toward, grasping, and retrieving a ball. However, motion patterns differed for one slideshow (straight-reach) relative to the other two (arcing-reaches), and one of the arcing-reach slideshows depicted a violation of typical goal-related motion. Preschoolers' knowledge of goal structure systematically modulated attention to event boundaries across slideshows despite surface differences, even when controlling for pixel change (an index of changes in motion). These findings showcase the value of the dwell time paradigm, and illuminate how children deploy attention as goal-related expectations shape their analysis of continuously unfolding activity.

Thiamine supplementation holds neurocognitive benefits for breastfed infants during the first year of life.

Women reliant on mostly rice-based diets can have inadequate thiamine intake, placing breastfed infants at risk of thiamine deficiency and, in turn, physical and cognitive impairments. We investigated the impact of maternal thiamine supplementation doses on infants' cognitive, motor, and language development across the first year. In this double-blind, four-parallel-arm, randomized controlled trial, healthy mothers of exclusively breastfed newborn infants were recruited in Kampong Thom, Cambodia. At 2 weeks postnatal, women (n = 335) were randomized to one of four treatment groups to consume one capsule/day with varying amounts of thiamine for 22 weeks: 0, 1.2, 2.4, and 10 mg. At 2, 12, 24, and 52 weeks of age, infants were assessed with the Mullen Scales of Early Learning (MSEL) and the Caregiver Reported Early Development Instrument (CREDI). Multiple regression and mixed effects modeling suggest that by 6 months of age, the highest maternal thiamine dose (10 mg/day) held significant benefits for infants' language development, but generally not for motor or visual reception development. Despite having achieved standardized scores on the MSEL that approximated U.S. norms by 6 months, infants showed a significant drop relative to these norms in both language domains following trial completion, indicating that nutritional interventions beyond 6 months may be necessary.

Low-dose thiamine supplementation of lactating Cambodian mothers improves human milk thiamine concentrations: a randomized controlled trial.

BACKGROUND: Infantile beriberi-related mortality is still common in South and Southeast Asia. Interventions to increase maternal thiamine intakes, and thus human milk thiamine, are warranted; however, the required dose remains unknown. OBJECTIVES: We sought to estimate the dose at which additional maternal intake of oral thiamine no longer meaningfully increased milk thiamine concentrations in infants at 24 wk postpartum, and to investigate the impact of 4 thiamine supplementation doses on milk and blood thiamine status biomarkers. METHODS: In this double-blind, 4-parallel arm randomized controlled dose-response trial, healthy mothers were recruited in Kampong Thom, Cambodia. At 2 wk postpartum, women were randomly assigned to consume 1 capsule, containing 0, 1.2 (estimated average requirement), 2.4, or 10 mg of thiamine daily from 2 through 24 weeks postpartum. Human milk total thiamine concentrations were measured using HPLC. An Emax curve was plotted, which was estimated using a nonlinear least squares model in an intention-to-treat analysis. Linear mixed-effects models were used to test for differences between treatment groups. Maternal and infant blood thiamine biomarkers were also assessed. RESULTS: In total, each of 335 women was randomly assigned to1 of the following thiamine-dose groups: placebo (n = 83), 1.2 mg (n = 86), 2.4 mg (n = 81), and 10 mg (n = 85). The estimated dose required to reach 90% of the maximum average total thiamine concentration in human milk (191 µg/L) is 2.35 (95% CI: 0.58, 7.01) mg/d. The mean ± SD milk thiamine concentrations were significantly higher in all intervention groups (183 ± 91, 190 ± 105, and 206 ± 89 µg/L for 1.2, 2.4, and 10 mg, respectively) compared with the placebo group (153 ± 85 µg/L; P < 0.0001) and did not significantly differ from each other. CONCLUSIONS: A supplemental thiamine dose of 2.35 mg/d was required to achieve a milk total thiamine concentration of 191 µg/L. However, 1.2 mg/d for 22 wk was sufficient to increase milk thiamine concentrations to similar levels achieved by higher supplementation doses (2.4 and 10 mg/d), and comparable to those of healthy mothers in regions without beriberi. This trial was registered at clinicaltrials.gov as NCT03616288.

Assessment of salt intake to consider salt as a fortification vehicle for thiamine in Cambodia.

Thiamine deficiency is a public health issue in Cambodia. Thiamine fortification of salt has been proposed; however, the salt intake of lactating women, the target population, is currently unknown. We estimated salt intakes among lactating women (<6 months postpartum) using three methods: repeat observed-weighed intake records and 24-h urinary sodium excretions (n = 104), and household salt disappearance (n = 331). Usual salt intake was estimated by adjusting for intraindividual intakes using the National Cancer Institute method, and a thiamine salt fortification scenario was modeled using a modified estimated average requirement (EAR) cut-point method. Unadjusted salt intake from observed intakes was 9.3 (8.3-10.3) g/day, which was not different from estimated salt intake from urinary sodium excretions, 9.0 (8.4-9.7) g/day (P = 0.3). Estimated salt use from household salt disappearance was 11.3 (10.7-11.9) g/person/day. Usual (adjusted) salt intake from all sources was 7.7 (7.4-8.0) g/day. Assuming no stability losses, a modeled fortification dose of 275 mg thiamine/kg salt could increase thiamine intakes from fortified salt to 2.1 (2.0-2.2) mg/day, with even low salt consumers reaching the EAR of 1.2 mg/day from fortified salt alone. These findings, in conjunction with future sensory and stability research, can inform a potential salt fortification program in Cambodia.

How Does the Mind Render Streaming Experience as Events?

Events-the experiences we think we are having and recall having had-are constructed; they are not what actually occurs. What occurs is ongoing dynamic, multidimensional, sensory flow, which is somehow transformed via psychological processes into structured, describable, memorable units of experience. But what is the nature of the redescription processes that fluently render dynamic sensory streams as event representations? How do such processes cope with the ubiquitous novelty and variability that characterize sensory experience? How are event-rendering skills acquired and how do event representations change with development? This review considers emerging answers to these questions, beginning with evidence that an implicit tendency to monitor predictability structure via statistical learning is key to event rendering. That is, one way that the experience of bounded events (e.g., actions within behavior, words within speech) arises is with the detection of "troughs" in sensory predictability. Interestingly, such troughs in predictability are often predictable; these regions of predictable-unpredictability provide articulation points to demarcate one event from another in representations derived from the actual streaming information. In our information-optimization account, a fluent event-processor predicts such troughs and selectively attends to them-while suppressing attention to other regions-as sensory streams unfold. In this way, usage of attentional resources is optimized for efficient sampling of the most relevant, information-rich portions of the unfolding flow of sensation. Such findings point to the development of event-processing fluency-whether in action, language, or other domains-depending crucially on rapid and continual cognitive reorganization. As knowledge of predictability grows, attention is adaptively redeployed. Accordingly, event experiences undergo continuous alteration.

Causal learning in CTC: Adaptive and collaborative.

Osiurak and Reynaud highlight the critical role of technical-reasoning skills in the emergence of human cumulative technological culture (CTC), in contrast to previous accounts foregrounding social-reasoning skills as key to CTC. We question their analysis of the available evidence, yet for other reasons applaud the emphasis on causal understanding as central to the adaptive and collaborative dynamics of CTC.

Thiamine dose response in human milk with supplementation among lactating women in Cambodia: study protocol for a double-blind, four-parallel arm randomised controlled trial.

INTRODUCTION: Thiamine (vitamin B1) deficiency remains a concern in Cambodia where women with low thiamine intake produce thiamine-poor milk, putting their breastfed infants at risk of impaired cognitive development and potentially fatal infantile beriberi. Thiamine fortification of salt is a potentially low-cost, passive means of combating thiamine deficiency; however, both the dose of thiamine required to optimise milk thiamine concentrations as well as usual salt intake of lactating women are unknown. METHODS AND ANALYSIS: In this community-based randomised controlled trial, 320 lactating women from Kampong Thom, Cambodia will be randomised to one of four groups to consume one capsule daily containing 0, 1.2, 2.4 or 10 mg thiamine as thiamine hydrochloride, between 2 and 24 weeks postnatal. The primary objective is to estimate the dose where additional maternal intake of thiamine no longer meaningfully increases infant thiamine diphosphate concentrations 24 weeks postnatally. At 2, 12 and 24 weeks, we will collect sociodemographic, nutrition and health information, a battery of cognitive assessments, maternal (2 and 24 weeks) and infant (24 weeks only) venous blood samples (biomarkers: ThDP and transketolase activity) and human milk samples (also at 4 weeks; biomarker: milk thiamine concentrations). All participants and their families will consume study-provided salt ad libitum throughout the trial, and we will measure salt disappearance each fortnight. Repeat weighed salt intakes and urinary sodium concentrations will be measured among a subset of 100 participants. Parameters of Emax dose-response curves will be estimated using non-linear least squares models with both 'intention to treat' and a secondary 'per-protocol' (capsule compliance ≥80%) analyses. ETHICS AND DISSEMINATION: Ethical approval was obtained in Cambodia (National Ethics Committee for Health Research 112/250NECHR), Canada (Mount Saint Vincent University Research Ethics Board 2017-141) and the USA (University of Oregon Institutional Review Board 07052018.008). Results will be shared with participants' communities, as well as relevant government and scientific stakeholders via presentations, academic manuscripts and consultations. TRIAL REGISTRATION NUMBER: NCT03616288.

Attentional profiles linked to event segmentation are robust to missing information.

Everyday experience consists of rapidly unfolding sensory information that humans redescribe as discrete events. Quick and efficient redescription facilitates remembering, responding to, and learning from the ongoing sensory flux. Segmentation seems key to successful redescription: the extent to which viewers can identify boundaries between event units within continuously unfolding activities predicts both memory and action performance. However, what happens to processing when boundary content is missing? Events occurring in naturalistic situations seldom receive continuous undivided attention. As a consequence, information, including boundary content, is likely sometimes missed. In this research, we systematically explored the influence of missing information by asking participants to advance at their own pace through a series of slideshows. Some slideshows, while otherwise matched in content, contained just half of the slides present in other slideshows. Missing content sometimes occurred at boundaries. As it turned out, patterns of attention during slideshow viewing were strikingly similar across matched slideshows despite missing content, even when missing content occurred at boundaries. Moreover, to the extent that viewers compensated with increased attention, missing content did not significantly undercut event recall. These findings seem to further confirm an information optimization account of event processing: event boundaries receive heightened attention because they forecast unpredictability and thus, optimize the uptake of new information. Missing boundary content sparks little change in patterns of attentional modulation, presumably because the underlying predictability parameters of the unfolding activity itself are unchanged by missing content. Optimizing information, thus, enables event processing and recall to be impressively resilient to missing content.

Attention reorganizes as structure is detected in dynamic action.

Once one sees a pattern, it is challenging to "unsee" it; discovering structure alters processing. Precisely what changes as this happens is unclear, however. We probed this question by tracking changes in attention as viewers discovered statistical patterns within unfolding event sequences. We measured viewers' "dwell times" (e.g., Hard, Recchia, & Tversky, 2011) as they advanced at their own pace through a series of still-frame images depicting a sequence of event segments ("actions") that were discoverable only via sensitivity to statistical regularities among the component motion elements. "Knowledgeable" adults, who had had the opportunity to learn these statistical regularities prior to the slideshow viewing, displayed dwell-time patterns indicative of sensitivity to the statistically defined higher-level segmental structure; "naïve" adults, who lacked the opportunity for prior viewing, did not. These findings clarify that attention reorganizes in conjunction with statistically guided discovery of segmental structure within continuous human activity sequences. As patterns emerge in the mind, attention redistributes selectively to target boundary regions, perhaps because they represent highly informative junctures of "predictable unpredictability."

Attention rapidly reorganizes to naturally occurring structure in a novel activity sequence.

Fluent event processing involves selectively attending to information-rich regions within dynamically unfolding sensory streams (e.g., Newtson, 1973). What counts as information-rich likely depends on numerous factors, however, including overall event novelty and local opportunity for repeated viewing. Using Hard, Recchia, and Tversky's (2011) method, we investigated the extent to which these two variables affected viewers' attentional patterns as events unfolded. Specifically, we recorded viewers' "dwell times" as they advanced through two slideshows depicting distinct methods of shoelace tying varying in novelty but equated on other dimensions. Across two experiments, novelty sparked increased dwelling overall, and viewers' dwelling patterns displayed rapid and systematic reorganization to structure within the activity stream after just one viewing of distinctively novel content. As well, increased dwelling positively predicted memory performance. These findings newly illuminate reorganization in attention as relevant information within novel activity sequences is quickly incorporated to guide event processing and support event memory.

Exploring some edges: Chunk-and-Pass processing at the very beginning, across representations, and on to action.

We identify three "working edges" for fruitful elaboration of the Chunk-and-Pass proposal: (a) accounting for the earliest phases of language acquisition, (b) explaining diversity in the stability and plasticity of different representational types, and

Inferring action structure and causal relationships in continuous sequences of human action.

In the real world, causal variables do not come pre-identified or occur in isolation, but instead are embedded within a continuous temporal stream of events. A challenge faced by both human learners and machine learning algorithms is identifying subsequences that correspond to the appropriate variables for causal inference. A specific instance of this problem is action segmentation: dividing a sequence of observed behavior into meaningful actions, and determining which of those actions lead to effects in the world. Here we present a Bayesian analysis of how statistical and causal cues to segmentation should optimally be combined, as well as four experiments investigating human action segmentation and causal inference. We find that both people and our model are sensitive to statistical regularities and causal structure in continuous action, and are able to combine these sources of information in order to correctly infer both causal relationships and segmentation boundaries.

Shining light on infants' discovery of structure.

Learning and discovery seem often to begin with noting patterns. Human infants are skilled at pattern detection, even patterns only definable at an abstract level, which is key to their acquisition of complex knowledge systems such as language and music. However, research examining infants' abstract rule learning has generated inconsistent results. We propose that apparent domain differences in infants' abstract rule learning may be the result of extraneous stimulus variation and discrepancies in the methodologies employed across studies probing this skill. We discuss how a behavioral methodology indexing infants' online learning would be valuable in furthering understanding of infants' (as well as adults') abstract rule learning and its neurophysiological concomitants. We outline current research aimed at developing such an index, and we propose future research, pairing such techniques with neurophysiological methods, aimed at shining more light on human skill at discovering structure.

Statistical learning of action: the role of conditional probability.

Identification of distinct units within a continuous flow of human action is fundamental to action processing. Such segmentation may rest in part on statistical learning. In a series of four experiments, we examined what types of statistics people can use to segment a continuous stream involving many brief, goal-directed action elements. The results of Experiment 1 showed no evidence for sensitivity to conditional probability, whereas Experiment 2 displayed learning based on joint probability. In Experiment 3, we demonstrated that additional exposure to the input failed to engender sensitivity to conditional probability. However, the results of Experiment 4 showed that a subset of adults-namely, those more successful at identifying actions that had been seen more frequently than comparison sequences-were also successful at learning conditional-probability statistics. These experiments help to clarify the mechanisms subserving processing of intentional action, and they highlight important differences from, as well as similarities to, prior studies of statistical learning in other domains, including language.

Infants recognize similar goals across dissimilar actions involving object manipulation.

We investigated infants' ability to recognize the similarity between observed and implied goals when actions differed in surface-level motion details. In two experiments, 10- to 12-month-olds were habituated to an actor manipulating an object and then shown test actions in which the actor contacted the object with a novel hand configuration that implied a goal either similar or dissimilar to the habituation event. Infants in both experiments looked significantly longer at test actions depicting a novel implied goal, suggesting that infants glossed over some surface-level motion details and compared implied goals.

Social gating and pedagogy: mechanisms for learning and implications for robotics.

It seems self-evident that human responsiveness to social input enhances learning, yet the details of the social forces at play are only beginning to come into focus. Recent research on language and cognitive development in preschoolers and infants illuminates mechanisms such as social gating and natural pedagogy, and specific ways in which they benefit learning. We review such advances and consider implications of this research for designing robotic systems that can harness the power of social forces for learning.

Sources of information for discriminating dynamic human actions.

Despite the importance of action identification and discrimination in action perception and social cognition more broadly, little research has investigated how these processes are achieved. To this end, we sought to identify the extent to which adults capitalize on featural versus configural sources of information when discriminating small-scale human actions such as grasp and place. Results across two experiments indicate adults are sensitive to both sources of information in action discrimination, but selectively attend to featural over configural action information. The findings also parallel what is known regarding face processing: processing of configural information is especially disrupted by inversion, whereas processing of featural information is specifically affected by low-pass filtering.