skip to main content

Search for: All records

Creators/Authors contains: "Williams, G."

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Abstract Background The Learning Assistant (LA) model with its subsequent support and training has evidenced significant gains for undergraduate STEM learning and persistence, especially in high-stakes courses like Calculus. Yet, when a swift and unexpected transition occurs from face-to-face to online, remote learning of the LA environment, it is unknown how LAs are able to maintain their motivation (competence, autonomy, and relatedness), adapt to these new challenges, and sustain their student-centered efforts. This study used Self-Determination Theory (SDT) to model theoretical aspects of LAs’ motivations (persistence and performance) both before and after changes were made in delivery of a Calculus II course at Texas Tech University due to COVID-19 interruptions. Results Analysis of weekly written reflections, a focus group session, and a post-course questionnaire of 13 Calculus II LAs throughout Spring semester of 2020 showed that LAs’ reports of competence proportionally decreased when they transitioned online, which was followed by a moderate proportional increase in reports of autonomy (actions they took to adapt to distance instruction) and a dramatic proportional increase in reports of relatedness (to build structures for maintaining communication and building community with undergraduate students). Conclusions Relatedness emerged as the most salient factor from SDT to maintain LAmore »self-determination due to the COVID-19 facilitated interruption to course delivery in a high-stakes undergraduate STEM course. Given that online learning continues during the pandemic and is likely to continue after, this research provides an understanding to how LAs responded to this event and the mounting importance of relatedness when LAs are working with undergraduate STEM learners. Programmatic recommendations are given for enhancing LA preparation including selecting LAs for autonomy and relatedness factors (in addition to competence), modeling mentoring for remote learners, and coaching in best practices for online instruction.« less
    Free, publicly-accessible full text available December 1, 2022
  2. In higher education, Learning Assistants (LAs)—a relatively recent evolution grounded in peer mentorship models—are gaining popularity in classrooms as universities strive to meet the needs of undergraduate learners. Unlike Teaching Assistants, LAs are undergraduate students who receive continuous training from faculty mentors in content-area coaching and pedagogical skills. As near-peers, they assist assigned groups of undergraduates (students) during class. Research on LAs suggests that they are significant in mitigating high Drop-Fail-Withdrawal rates of large enrollment undergraduate science, technology, engineering, mathematics, and medical (STEMM) courses. However, there is a dearth of description regarding the learning between LAs and STEMM faculty mentors. This paper reports on perspectives of faculty mentors and their cooperating LAs in regard to their learning relationships during a Calculus II at a research-oriented university during Spring of 2020. Using an exploratory-descriptive qualitative design, faculty (oral responses) and LAs (written responses) reflected on their relationship. Content analysis (coding) resulted in four salient categories (by faculty and LA percentages, respectively) in: Showing Care and Fostering Relationships (47%, 23%); Honing Pedagogical Skills (27%, 36%); Being Prepared for Class and Students (23%, 28%); and Developing Content Knowledge in Calculus (3%, 13%). Benefits of LAs to faculty and ways to commence LA programsmore »at institutions are also discussed.« less
  3. The widespread growth of additive manufacturing, a field with a complex informatic “digital thread”, has helped fuel the creation of design repositories, where multiple users can upload distribute, and download a variety of candidate designs for a variety of situations. Additionally, advancements in additive manufacturing process development, design frameworks, and simulation are increasing what is possible to fabricate with AM, further growing the richness of such repositories. Machine learning offers new opportunities to combine these design repository components’ rich geometric data with their associated process and performance data to train predictive models capable of automatically assessing build metrics related to AM part manufacturability. Although design repositories that can be used to train these machine learning constructs are expanding, our understanding of what makes a particular design repository useful as a machine learning training dataset is minimal. In this study we use a metamodel to predict the extent to which individual design repositories can train accurate convolutional neural networks. To facilitate the creation and refinement of this metamodel, we constructed a large artificial design repository, and subsequently split it into sub-repositories. We then analyzed metadata regarding the size, complexity, and diversity of the sub-repositories for use as independent variables predicting accuracymore »and the required training computational effort for training convolutional neural networks. The networks each predict one of three additive manufacturing build metrics: (1) part mass, (2) support material mass, and (3) build time. Our results suggest that metamodels predicting the convolutional neural network coefficient of determination, as opposed to computational effort, were most accurate. Moreover, the size of a design repository, the average complexity of its constituent designs, and the average and spread of design spatial diversity were the best predictors of convolutional neural network accuracy.« less
  4. Intricate mesostructures in additive manufacturing (AM) designs can offer enhanced strength-to-weight performance. However, complex mesostructures can also hinder designers, often resulting in unpalatably large digital files that are difficult to modify. Similarly, existing methods for defining and representing complex mesostructures are highly variable, which further increases the challenge in realizing such structures for AM. To address these gaps, we propose a standardized framework for designing and representing mesostructured components tailored to AM. Our method uses a parametric language to describe complex patterns, defined by a combination of macrostructural, mesostructural, and vector field information. We show how various mesostructures, ranging from simple rectilinear patterns to complex, vector field-driven cellular cutouts can be represented using few parameters (unit cell dimensions, orientation, and spacing). Our proposed framework has the potential to significantly reduce file size, while its extensible nature enables it to be expanded in the future.
  5. ABSTRACT We present photometry, spectra, and spectropolarimetry of supernova (SN) 2014ab, obtained through ∼200 d after peak brightness. SN 2014ab was a luminous Type IIn SN (MV < −19.14 mag) discovered after peak brightness near the nucleus of its host galaxy, VV 306c. Pre-discovery upper limits constrain the time of explosion to within 200 d prior to discovery. While SN 2014ab declined by ∼1 mag over the course of our observations, the observed spectrum remained remarkably unchanged. Spectra exhibit an asymmetric emission-line profile with a consistently stronger blueshifted component, suggesting the presence of dust or a lack of symmetry between the far side and near side of the SN. The Pa β emission line shows a profile very similar to that of H α, implying that this stronger blueshifted component is caused either through obscuration by large dust grains, occultation by optically thick material, or a lack of symmetry between the far side and near side of the interaction region. Despite these asymmetric line profiles, our spectropolarimetric data show that SN 2014ab has little detected polarization after accounting for the interstellar polarization. We are likely seeing emission from a photosphere that has only small deviation from circular symmetry in the plane normal to our line of sight, but with eithermore »large-grain dust or significant asymmetry in the density of circumstellar material or SN ejecta along our line of sight. We suggest that SN 2014ab and SN 2010jl (as well as other SNe IIn) may be events with similar geometry viewed from different directions.« less