Bayesian Estimation of Generalized Log-Linear Poisson Item Response Models for Fluency Scores Using brms and Stan

In this paper, published in Journal of Intelligence, we show how generalized log-linear Poisson item-response models for fluency scores (including the 2-Parameter Poisson Counts Model) can be estimated and interpreted in a Bayesian multilevel framework using brms and Stan.

Accounting for Variable Task Discrimination in Divergent Thinking Fluency Measurement – An Example of the Benefits of a 2-Parameter Poisson Counts Model and its Bifactor Extension Over the Rasch Poisson Counts Model

In this paper, published in Journal of Creative Behavior, we introduce new psychometric models for count/fluency tasks (tasks in which individuals have to provide many instances). Notably, we extent the Rasch Poisson Counts Model (RPCM) to account for variable discrimination (2-Parameter Poisson Counts Model) and to account for local dependencies/nuisance factors (Bifactor 2-Parameter Poisson Counts Model).