Denison Scientific Association: Practical Visualization and Interpretation of Generalized Linear Mixed-Effects Models in Linguistic Research

Generalized linear mixed-effects model (GLMM) is a flexible statistical tool that helps us understand patterns when data points are not fully independent. In collaboration with Dr. Mary Beaton (Modern Languages), this project applies a GLMM to explore how social and linguistic factors interact in the pronunciation of word-final “r” in Lorain Puerto Rican Spanish.

Rather than focusing on the specific results, this talk highlights how to visualize and interpret interaction effects in a GLMM. By plotting predicted probabilities across generations, genders, and linguistic contexts, we can see how multiple influences overlap and shift, making the model’s structure and meaning accessible even without a statistical background.