2018 Education Technology Computational Psychometrics Symposium


The Lineup

Dor Abrahamson | UC Berkeley

POWERPOINT: Envisioning Formative Assessment: Eye-Tracking the Emergence of Mathematics

What if teachers could see an idea evolving in a child's mind even before the child knew it? New theories of learning along with new research methods are transforming the way we design and evaluate educational activities. In particular, eye-tracking instruments enable analysts to monitor the evolution of sensorimotor action schemes students develop in solving movement problems engineered to ground mathematical concepts. I will explain the design rationale of the Mathematics Imagery Trainer project, based in embodiment theory, and then present clinical and eye-tracking data evidencing 'attentional anchors,' dynamical perceptual forms children create spontaneously out of selected environmental features to facilitate the enactment of challenging bimanual movements; the children then become conscious of the attentional anchors as things that, through cultural mediation, take on mathematical meanings. As such, we offer empirical support for historical claims from Piaget, Vygotsky, and Varela pertaining to cognitive growth processes. In turn, we have proof-of-concept data to argue for the utility of formative multimodal assessment in mathematics teachers' instructional practice.

Adam Blum | ActNext + OpenEd

POWERPOINT: Integrating with Learning Object Repositories Using the New LTI Resource Search Standard

In this modern era of adaptive and personalized learning, most educational apps need resources (such as videos or games) for each student's learning gaps.  This has been difficult to achieve as each LOR exposed their own unique APIs for search (and some LORs don't even have an API).  There is no consistency on how to query for various types of resources and their alignments to learning objectives.   There is also no consistency in the metadata returned for each educational resource.  This has been a significant stumbling block in allowing resource usage within learning tools.  The new IMS Global LTI Resource Search standard builds on successful standards like LRMI/schema.org to both standardize how to search for resources and how to describe those that match.  It makes it easy to weave instructional resources into the fabric of every learning tool and platform.  It is supported by leading LORs who are IMS members such as ACT OpenEd, Knovation, SAFARI Montage, Pearson and SchoolCity. On final acceptance by IMS Global the spec will be published for all learning tools and LORs to use.

Maria Bolsinova & Benjamin Deonovic | ACTNext

POWERPOINT: Assessment meets Learning: On the relation between Item Response Theory & Bayesian Knowledge Tracing

POWERPOINT: Urnings: NextGen Psychometric Tools for NextGen Assessment & Learning Systems

The Urnings rating system is an algorithm for tracking changes in ability and item difficulty in real time. When in equilibrium, the estimates are unbiased and their measurement error is known, which makes the system suitable for high-stakes testing. Furthermore, it is computationally very efficient, which makes it suitable for large-scale applications. The Urnings system has been successfully used on historic chess data, star ratings of movies by users, and data from ACT academy. However, until recently the applications of the system have been limited by the restrictiveness of the Rasch model that it is based on. In this presentation we describe the Urnings rating system and its properties, and extend it to a wide range of multidimensional models, including IRT models and CDMs.

Sidney D'Mello | Sidney DMello LLC

POWERPOINT: Distributed Cognition in Multimodal Collaborative Learning Environments

I’m interested in socio-affective-cognitive (SAC) processes (e.g., joint attention; turn taking; negotiation; active participation) that arise during collaborations. These processes are multimodal, interact over multiple spatial and temporal scales, and are situated in a dynamic environment. I will discuss projects aimed at: (1) understanding how SAC processes arise and influence collaborative processes and outcomes; (2) developing multimodal computational models of SAC processes; and (3) embedding the computational models in intelligent technologies to trigger interventions aimed at improving subjective and objective collaborative outcomes.

Susan Embretson | Georgia Tech

POWERPOINT: Diagnosing Mathematics Skill deficits in Middle School

Dr. Embretson’s interests span modern psychometric methods (e.g., item response theory), cognitive and intelligence, and quantitative methods. Her main research program has been to integrate cognitive theory into psychometric models and test design. To this goal, she has been developing new item response theory models and conducting empirical research on the cognitive basis of an individual's responses. Recently, this effort has led to the exciting possibility of "tests without items". That is, items are automatically generated by artificial intelligence to target levels and cognitive sources of difficulty to optimally measure each individual examinee during testing. The measurement areas have included fluid reasoning, spatial ability, mathematical reasoning and verbal comprehension.

Ellen Forte | EdCount, LLC

POWERPOINT: A Framework for Addressing State and Local Assessment Needs

State and local education agencies use a variety of products and services to generate the data they need for accountability, evaluation, and instructional decision-making purposes. This session will offer a framework, developed as part of project funded by a current enhanced assessment grant from the US Department of Education, for reviewing assessment options in relation to information needs. This framework can guide educators as well as the vendors who serve them in addressing an array of validity-related issues.

Juan Pablo Hourcade | The University of Iowa

PREZI: Universal Interactions for Children

Computer literacy and access are increasingly critical to satisfy basic needs and rights (e.g., education, voting, transportation, purchases). This means computer software and hardware designers must ensure accessibility for increasingly diverse populations and computing devices. This talk outlines challenges and opportunities in design to achieve that goal, focusing on the principles of deeply engaging with stakeholders, working in interdisciplinary teams, personalizing technology, and making it practical for people’s reality. I will illustrate these principles by discussing research projects on technologies to support children’s creativity, collaboration, and information access

Yuchi Huang | ACTNext

POWERPOINT: Deep Learning for Educational Innovations

In recent years deep learning based artificial intelligence have had a profound impact on various areas. However, many researchers in Education are not familiar with this method. The purpose of this presentation is to make the inner workings of these techniques transparent and to explore technical challenges and needs in education that can be addressed with deep learning approaches. This presentation starts by introducing the motivation of adopting deep learning in a lot of machine learning driven educational applications. The second section briefly covers the basics of deep learning and its applications so that the audience could gain a rough understanding of various deep learning systems, libraries and computational resources. At last, we explore several use cases of deep learning in Education, such as automatic content generation, text summarization and generation of facial expressions in human-agent interactions.

Christine Ortiz | MIT & Station1

We stand at a pivotal moment in time that presents us with both enormous challenges and potential opportunities. Educational inequality exists globally and includes both access to postsecondary education and to high impact educational experiences (e.g. experiential learning, inquiry and research-based learning, internships, global education, creativity-based activities, interdisciplinary collaboration, etc.) – the latter of which is the currency of future, lifelong, adaptable, meaningful, satisfying, and productive careers. There is also an escalating concern that our traditional ways of carrying out scientific research, technological innovation and development, education are building upon and increasingly exacerbating inequality, thus creating damage to our society and planet. Hence, it is an imperative to pursue intentional strategies in higher education which, as a 2015 UNESCO report eloquently states, “foster respect for life and human dignity, equal rights and social justice, cultural and social diversity, sustainability, and a sense of human solidarity and shared responsibility for our common future.” In this presentation, I will discuss the development and ongoing work to build a new model of higher education based on inclusion and equity, learning through frontier project-based inquiry and research, and the integration of science and technology with societal perspective and impact. The role, need, and potential promise and peril of education technology and technological development approaches will be highlighted.

Yigal Rosen | ACTNext

POWERPOINT: Fostering Learning Solutions through Evidence Centered Design and Scalable Technologies

Several key developments challenge the field of learning technologies: An improved understanding of how people develop domain-specific and cross-cutting capabilities, evidence-centered design, and advances in artificial intelligence. According to the evolving framework, learning and assessment are intertwined into engaging and powerful learning experiences aimed to foster knowledge, skills and capabilities of all learners. In this talk, our new ACTNext Learning Solutions group will present leading principles for design and development of innovative learning and assessment technologies and share exemplar demos and prototypes.


Recent News

ACTNext Welcomes First ACTNext Fellow, Dr. Bryan Maddox!

Dr. Bryan Maddox will collaborate with the ACTNext team on projects related to Artificial Intelligence, Machine Learning, and Learning Solutions


The $1.5M competition focused on AI/ML/xR solutions for college students

Image Image