Episode 39: Give-N you Bayes and Backpropagation (Pt. II)

Episode 39: Give-N you Bayes and Backpropagation (Pt. II)

We're back with Part II. We continue our discussion of how to apply Bayesian models to number cognition, but in this segment we talk about another Lee and Sarnecka (2011) paper in which they show how the very same Bayesian model can be used to test two different theories of how children acquire number. We also talk about the strengths and weakness of large and small artificial neural networks, and Deon makes the case for why small models shouldn't be abandoned. We then talk a bit about what a model of the give-N task might look like and what role realism plays in the model. Links Lee, M. D., & Sarnecka, B. W. (2011). Number-knower levels in young children: Insights from Bayesian modeling. Cognition, 120(3), 391-402.Link [https://www.sciencedirect.com/science/article/pii/S0010027710002283?casa_token=CIRzaM6HEdwAAAAA:k8kWVn9Y6td5T-JEwnPTcPt64Pgiw1rqtLV58BCvomCo51GP_36UAc6Wpy3kevkZvN1xiqvY]

Episode 38: Give-N you Bayes and Backpropagation (Pt. I)

For the first time ever, Jenny and Deon recorded a two-part episode! This is Part I! In this episode, Deon introduces the basics of connectionist modeling and answers Jenny’s many questions about how these models work, how information can be represented in these systems (e.g., localist vs. distributed representations), and the role of theory in grounding decisions about which representation to use. They also discuss how connectionist models can model everything from simple problems like the Boolean AND function to developmental problems like the give-N task. Jenny and Deon also spend some time talking about Lee & Sarnecka’s (2010) Bayesian model of children’s give-N performance, Marr’s levels of analysis, why computational-level explanations alone are ultimately not enough, and what algorithmic- and implementational-level explanations provide that computational-level ones might not. Stay tuned for Part II! Links Lee, M. D., & Sarnecka, B. W. (2010). A model of knower‐level behavior in number concept development. Cognitive science, 34(1), 51-67. https://onlinelibrary.wiley.com/doi/full/10.1111/j.1551-6709.2009.01063.x [https://onlinelibrary.wiley.com/doi/full/10.1111/j.1551-6709.2009.01063.x] Rogers, T. T. (2009). Connectionist models. Encyclopaedia of Neuroscience, L. Squire (Ed.), Oxford: Academic Press, 75-82. https://vanderbilt.box.com/s/momlzz9euvrd8bf4tfygozfqqks82on8 [https://vanderbilt.box.com/s/momlzz9euvrd8bf4tfygozfqqks82on8]

Episode 37: From statistics to meaning (with Katie Graf Estes)

This week, we had the pleasure of sitting down with Dr. Katie Graf Estes, a Professor of Psychology at the University of California, Davis. In the first segement, we discussed what type of person young Dr. Graf Estes was along with her journey through academia to where she is now. During the second segement, we dive into Dr. Graf Estes's 2007 Psychological Science paper "Can Infants Map Meaning to Newly Segmented Words?: Statistical Segmentation and Word Learning." Here, we talk about the creativity of the experimental design, along with what statistical learning means in the grand scheme of language and cognitive development. Link to paper: https://journals.sagepub.com/doi/full/10.1111/j.1467-9280.2007.01885.x [https://journals.sagepub.com/doi/full/10.1111/j.1467-9280.2007.01885.x] Special Guest: Katharine Graf Estes.

Episode 36: The goldilocks of predicability (with Viridiana Benitez)

This week we had the pleasure of chatting with Dr. Viridiana Benitez, an Assistant Professor of Psychology at Arizona State University. In the first part of the episode we talked with Dr. Benitez about her academic journey along with how she approaches mentorship. In part two, we turned to her recent paper in Current Biology entitled "Predictable Events Enhance Word Learning in Toddlers." This paper explores how environmental predictability can scaffold novel word learning in toddlers. Towards the end, we discuss the relationship between Dr. Benitez's work in predictability and her work in bilingualism. Special Guest: Viridiana Benitez.

Episode 35: A 𝗰𝗿𝗶𝘁𝗶𝗰𝗮𝗹 take on the cognitive sciences (with Richard Prather)

In this episode, we were joined by Dr. Richard Prather, an associate professor in the human development department at the University of Maryland, College Park. In the first segment, we talk about Richard's academic journey, his approach to applying to college and then to graduate school, and how he landed on one of his major research topics -- numerical development. We also talk a bit about computational modeling at the end of this segment. In the second segment, we talk about how the cognitive sciences would benefit from adopting a more critical approach, what even is meant by "critical", how he'd respond to various potential critiques of the approach, whether this framework is feasible for junior scientists, and how he's applied it in his own work. Links Prather, R. W. (2023). A new path: Why we need critical approaches to cognitive and psychological sciences. Journal of Applied Research in Memory and Cognition, 12(2), 195–198. Link [https://www.pratherlab.org/wp-content/uploads/2023/04/anewpathpreprint.pdf] Special Guest: Richard Prather.