RR: Workshop on Robust Social and Behavioral Sciences

点击翻译按钮获取中文摘要

Psychological scientists have an impressive arsenal of research methodologies at their disposal.  Many of the currently popular methods are tried and tested and have been in use for 100 years or more.  Recently, however, exciting new methods have been developed that make use of modern computing technologies, advances in statistics, and big data. Because these modern methods often allow researchers to process large amounts of data at once and ask very detailed questions, they often give results that are robust (reliable and replicable). In order to bring these robust research methods to the forefront of psychological and social science, a "Workshop on Robust Social and Behavioral Sciences" will be held, which will be attended by some of the world's foremost experts in this area. The most important "robust" methods are (1) multifaceted randomized designs that enhance generalizability to a wider set of conditions, (2) robust statistics that are less sensitive to common violations of assumptions (including, notably, the assumption of random sample selection), (3) alternatives to classical hypothesis testing such as parameter estimation, process modeling, model selection, and Bayesian inference, and (4) collection of intensive data such as data generated by social media, data collected via Amazon Mechanical Turk, or data output by "many-labs" research consortia.  During the Workshop, attendees will discuss the pros and cons of these research methods that are adapted to the demands and affordances of our field in the present day.  An interdisciplinary panel led by an expert in cultural and behavioral change will evaluate the applicability of these methods to different subfields, identify discipline-specific challenges, and prepare a "road map" document outlining current such challenges for methodologists to address.  Attendees will also produce video lectures on the topic of robust methods, and these lectures will be informed by comments from the interdisciplinary panel.  All products of the Workshop will be made public.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.