Saturday, February 17 | |
PS3 Poster Session 3 and Continental Breakfast |
Sat, Feb 17, 8:00 AM - 9:15 AM
Salons F-I |
Understanding Graduate School Speed-Dating with Generalized Linear Mixed Models (303588)*Christina Phan Knudson, University of St. ThomasHenry Z. Zuo, Macalester College Keywords: correlated data, mixed models, dating, generalized linear mixed models More Americans are pursuing advanced degrees, but do they fare well in their dating life? We explore the relationship between a graduate student’s success in speed-dating sessions and the type of advanced degree they seek. In speed-dating experiments, the binary decisions (“Yes, I want to see you again.” or “No.”) given or received by the same individual are associated. For example, a particularly attractive participant is likely to receive many "yes" decisions and a particularly selective participant is likely to give out many "no" decisions. Models for speed-dating data must take into account this correlation structure. To do so, we leverage the generalized linear mixed model and incorporate participant-specific random effects. We explain random effects in the easy-to-understand context of speed-dating, demonstrate the model-fitting procedure using R package glmm, and explain the resulting model interpretation. Please stop by to learn more about modeling correlated data!
|