REGISTRATION FOR THIS WORKSHOP IS CLOSED.

Description: The use of categorical data generated by surveys or yes/no questions is pervasive in social science and psychological research, yet researchers frequently apply statistical methods meant for non-categorical data in their own analyses and in published work. Recent modeling evidence shows that analyzing categorical data using non-categorical models substantially inflates Type I error rates, as well skewing effect sizes and introducing interpretation errors (Liddell & Krushke, 2017). This course is designed to provide practical experience and theoretical background for statistical methods for categorical data. We will particularly focus on logistic and ordinal regression models, including mixed effects (“hierarchical”) regression. In addition, we will explore issues that arise in the interpretation of categorical response data in the presence of potential response biases (Rotello, Heit & Dubé, 2015). We will explore analytical techniques, such as Signal Detection Theory, to address these issues. Along the way, we will also introduce the R statistical programming environment, the tidyverse tools for organizing your R workflow, and project management using the Open Science Framework.         

Instructors: Caroline Andrews & Brian Dillon

Caroline Andrews is a PhD candidate in the Department of Linguistics at the University of Massachusetts Amherst. She studies the ways that humans access and implement complex linguistic constraints during real time sentence comprehension, drawing on insights from theoretical syntax and memory research. Her research is particularly focused on how past linguistic experience influences processing of an incoming syntactic structure, and how lexical information mediates access to abstract syntactic information. In the course of her work she employs both traditional NHST and Bayesian statistical approaches to continuous response time data and categorical judgment studies (binary and Likert scales).

Brian Dillon is an Associate Professor in the Department of Linguistics at the University of Massachusetts Amherst. He studies how readers encode and manage linguistic information in memory during the course of routine language understanding, in particular during reading. To do this he uses a range of computational and experimental tools and investigates data from a range of different languages such as French, Spanish, Hindi, and Mandarin Chinese. 

REGISTRATION FOR THIS WORKSHOP IS CLOSED.

Space is Limited!! Please register early to reserve your seat, and to assist in our planning.

• Five College Undergraduate and Graduate Students…………………………..$150/person
• Five College Faculty………………………………………………………………..$250/person
• Non-Five College Undergraduate and Graduate Students…………………………..$275/person
• Non-Five College Faculty………………………………………………………………..$400/person

Registration note: The Five Colleges include: UMass Amherst, Amherst College, Hampshire College, Mount Holyoke College, and Smith College. Registration closes for each workshop 2 full business days prior to the start date. Payment of registration does not guarantee a spot in these space-limited workshops. ISSR will offer you the choice of placement on a wait-list or a refund for any workshops that are over-subscribed. If paying with departmental funds or personal checks, contact Karen Mason (mason@issr.umass.edu).

Cancellation note: In cases where enrollment is 5 or less, we reserve the right to cancel the workshop.  In cases where the registrant cancels prior to the workshop, a full refund will be given with two weeks notice, and 50% refund will be given with one week notice.  We will not be able to refund in cases where registrant does not notify us of cancellation at least one week prior to the beginning date of the workshop.