UCLA

We propose an attention-based theory for how humans extract semantic information from objects located in our peripheral visual field. We build upon the scale invariance for extraction of semantic information from objects in center gaze. Our theory for peripheral vision hypothesizes that a vision attention vector runs parallel to the brain's main visual pathway. Information from this vision attention vector is able to combine with our brain's representation of our visual field, such that a peripheral target is able to be remapped and represented at center, in a downstream stage of the brain's visual processing. We connect our theory to previous observations of peripheral vision impairments and documented cases of specific types of letter confusions. We simulate these past findings by modeling peripheral vision object identification failures as an off-target vision attention vector. Our own crowding experiments show support for an active, attention-based mechanism for extraction of peripheral semantic information. We show that a fully crowded visual environment degrades peripheral vision abilities by over twice that of a locally peripherally crowded visual environment ($P = 2.02 \times 10^{-10}$). Additionally, we find instances where a letter on the outside of a peripheral cluster yields a smaller threshold letter height, as compared to a letter on the inside of that same peripheral cluster ($P = 1.20 \times 10^{-2}$). Such combination of simulation and experiment results offer support for our vision attention vector theory, which is, insofar as we are aware, the first comprehensive theory for how humans can extract a single semantic representation for objects anywhere in the visual field.

Additionally, we review our physics for life sciences (IPLS) laboratory revisions that have positively impacted over 4,850 undergraduate students at UCLA. To achieve learning outcomes of improved critical thinking and problem-solving persistence, we used a combination of ``flipped" pre-laboratory assignments, inquiry-based in-lab activities, and peer-based learning with undergraduate learning assistants. As a result of our revisions, an increased number of students reports pursuing their own scientific questions during physics experiments. On average, students show a pre/post quarter attitude shift of 0.50 Likert Levels ($P = 4.1 \times 10^{-12}$), as shown by E-CLASS assessment analysis. We additionally show that, after a second round of revisions, a decreased number of students reports immediately asking an expert like the instructor for help when facing a challenge during an experiment. (pre/post quarter attitude shift of -0.23 Likert Levels ($P = 1.0 \times 10^{-4}$)). Therefore, our evidence-based revisions of undergraduate physics labs show an increased development of important physics skills in our next generation of healthcare professionals. Our revisions may serve as an example to support other institutions in nation-wide efforts to optimize undergraduate physics education.