theoraizer: AI-assisted Theory Construction

The Causal Loop Diagram (CLD) method is a technique for theory construction in which domain experts collaborate to identify causal relationships between variables. However, CLD construction is labor-intensive, and the input required from experts grows quadratically with the number of variables involved. This limits the method to the construction of small graphs. Large Language Models (LLMs) can generate large amounts of content through a standardized pipeline and, therefore, offer the potential to lift these limitations. This paper presents \texttt{theoraizer}, an R package and Shiny app that enhances CLD construction by integrating LLMs as a digital extension of the expert group. Researchers can use \texttt{theoraizer} to define a list of putative variables, after which it queries the LLM for putative causal links between these variables. This method drastically reduces the amount of work required to arrive at a candidate CLD and provides scientists with a standardized, multi-stage framework for constructing candidate theories involving large numbers of variables.