|Dr. Sanabria is an associate professor of psychology at Arizona State University and principal investigator at the Basic Behavioral Processes laboratory. He is also affiliated to the neuroscience program in Arizona State University. Dr. Sanabria received his professional degree in psychology from the Universidad de los Andes en Bogotá (Colombia), where he spent a few years as marketing research analyst. He received his doctoral degree in experimental psychology in 2004 from Stony Brook University, where he worked on self-control under the guidance of Dr. Howard Rachlin. He was a postdoctoral research associate under the supervision of Dr. Peter Killeen (2004–2008) and Dr. Janet Neisewander (2008) in Arizona State University. He has published over 40 peer-reviewed publications in the Journal of the Experimental Analysis of Behavior (JEAB), Behavioural Brain Research, Psychopharmacology, and other journals. He is the president-elect of the International Society for Comparative Psychology, board member of the Society for the Quantitative Analysis of Behavior, and associate editor of JEAB and Learning and Behavior. His research is primarily concerned with the development and evaluation of quantitative models of basic behavioral processes (learning, timing, choice, and regulation) in psychiatric disorders (mainly, substance abuse and attention-deficit hyperactivity disorder).|
Quantitative modeling is increasingly common in behavior analysis. Performance on concurrent schedules, timing, delay discounting, behavioral momentum, schedule and stimulus control, variability of inter-response times, and many other aspects of behavior, are often characterized in terms of mathematical equations and computational algorithms. This presentation outlines the advantages, challenges, and pitfalls of a quantitative analysis of behavior. In particular, this presentation is focused on identifying the outcomes that quantitative models may and may not deliver, the assumptions and pre-requisites for quantitative modeling, the risks involved in this strategy, and the tactics that minimize such risks. The presentation will introduce the concepts of model-based inference, parameter estimation, stochastic vs. deterministic models, likelihood vs. probability, parsimony vs. goodness-of-fit, Bayesian modeling, and model selection. This introduction will set the stage for the practical implementation of some of these concepts.
|Learning Objectives: At the end of the presentation, the participant will be able to: (1) identify the advantages and challenges associated with a quantitative analysis of behavior, relative to conventional approaches to behavior analytic research; (2) identify and appropriately use the concepts of model-based inference, parameter estimation, stochastic and deterministic models, likelihood and probability, parsimony and goodness-of-fit, Bayesian modeling, and model selection.|