Search for: All records

According to NASEM (2018), data science has foundations in computing, mathematics, and statistics. However, at the K-12 level, these foundations are usually taught as standalone courses that are unconnected with each other. Students may struggle to see their connections. We proposed a framework unifying those foundations using mathematical logic. A core concept in mathematical logic is function. A general function has one or more possibly non-number inputs and an output. Data science motivates a comprehensive understanding of functions and provides extensive culturally relevant, real-world, and data-rich problems and applications for students to practice their understanding. It is interesting to know how well students understand functions. We developed a six-lesson online module with more than 100 in-lesson questions. Initial analysis of the students’ answers to the questions shows that students can understand the basics of the general functions but have more difficulties in involved applications of functions. 

Data science is revolutionizing academia and industry, creating a high demand for a workforce fluent in this field. While the availability of data science courses has increased recently, few curricula rigorously build on mathematical logic. The LogicDS Project addresses this gap by engaging high school students from rural communities in an online data science course integrating mathematics, statistics, and programming into a unified framework based on logic and reasoning. A one-week course, consisting of six lessons, was developed and 110 participants were recruited. Pre- and post-intervention data, along with students' LMS activity logs, were collected to analyze engagement. Results indicate that the Logic-Based framework effectively engages students from diverse backgrounds, with participants finding the course valuable for learning data science skills. Notably, entropy analysis of student activity logs correlated with other mixed methods analyses, providing insights into engaging K-12 students in data science education. 

Abstract To date, many AI initiatives (eg, AI4K12, CS for All) developed standards and frameworks as guidance for educators to create accessible and engaging Artificial Intelligence (AI) learning experiences for K‐12 students. These efforts revealed a significant need to prepare youth to gain a fundamental understanding of how intelligence is created, applied, and its potential to perpetuate bias and unfairness. This study contributes to the growing interest in K‐12 AI education by examining student learning of modelling real‐world text data. Four students from an Advanced Placement computer science classroom at a public high school participated in this study. Our qualitative analysis reveals that the students developed nuanced and in‐depth understandings of how text classification models—a type of AI application—are trained. Specifically, we found that in modelling texts, students: (1) drew on their social experiences and cultural knowledge to create predictive features, (2) engineered predictive features to address model errors, (3) described model learning patterns from training data and (4) reasoned about noisy features when comparing models. This study contributes to an initial understanding of student learning of modelling unstructured data and offers implications for scaffolding in‐depth reasoning about model decision making. Practitioner notesWhat is already known about this topicScholarly attention has turned to examining Artificial Intelligence (AI) literacy in K‐12 to help students understand the working mechanism of AI technologies and critically evaluate automated decisions made by computer models.While efforts have been made to engage students in understanding AI through building machine learning models with data, few of them go in‐depth into teaching and learning of feature engineering, a critical concept in modelling data.There is a need for research to examine students' data modelling processes, particularly in the little‐researched realm of unstructured data.What this paper addsResults show that students developed nuanced understandings of models learning patterns in data for automated decision making.Results demonstrate that students drew on prior experience and knowledge in creating features from unstructured data in the learning task of building text classification models.Students needed support in performing feature engineering practices, reasoning about noisy features and exploring features in rich social contexts that the data set is situated in.Implications for practice and/or policyIt is important for schools to provide hands‐on model building experiences for students to understand and evaluate automated decisions from AI technologies.Students should be empowered to draw on their cultural and social backgrounds as they create models and evaluate data sources.To extend this work, educators should consider opportunities to integrate AI learning in other disciplinary subjects (ie, outside of computer science classes). 

Abstract Understanding the effects of intensification of Amazon basin hydrological cycling—manifest as increasingly frequent floods and droughts—on water and energy cycles of tropical forests is essential to meeting the challenge of predicting ecosystem responses to climate change, including forest “tipping points”. Here, we investigated the impacts of hydrological extremes on forest function using 12+ years of observations (between 2001–2020) of water and energy fluxes from eddy covariance, along with associated ecological dynamics from biometry, at the Tapajós National Forest. Measurements encompass the strong 2015–2016 El Niño drought and La Niña 2008–2009 wet events. We found that the forest responded strongly to El Niño‐Southern Oscillation (ENSO): Drought reduced water availability for evapotranspiration (ET) leading to large increases in sensible heat fluxes (H). PartitioningETby an approach that assumes transpiration (T) is proportional to photosynthesis, we found that water stress‐induced reductions in canopy conductance (G_s) droveTdeclines partly compensated by higher evaporation (E). By contrast, the abnormally wet La Niña period gave higherTand lowerE, with little change in seasonalET. Both El Niño‐Southern Oscillation (ENSO) events resulted in changes in forest structure, manifested as lower wet‐season leaf area index. However, only during El Niño 2015–2016, we observed a breakdown in the strong meteorological control of transpiration fluxes (via energy availability and atmospheric demand) because of slowing vegetation functions (via shutdown ofG_sand significant leaf shedding). Drought‐reducedTandG_s, higherHandE, amplified by feedbacks with higher temperatures and vapor pressure deficits, signaled that forest function had crossed a threshold, from which it recovered slowly, with delay, post‐drought. Identifying such tipping point onsets (beyond which future irreversible processes may occur) at local scale is crucial for predicting basin‐scale threshold‐crossing changes in forest energy and water cycling, leading to slow‐down in forest function, potentially resulting in Amazon forests shifting into alternate degraded states.