theoraizer: AI-assisted theory construction

The Causal Loop Diagram (CLD) method is a structured approach to theory development in which domain experts collaboratively identify causal relationships between variables. However, CLD construction is labor-intensive, time-consuming, and cognitively demanding. Large Language Models (LLMs), with their advanced text processing capabilities and extensive knowledge base, offer the potential to support theory construction and reduce workload. This paper presents 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 theoraizer to define a list of putative variables, then query an LLM to evaluate candidate causal links between these variables and provide supporting literature. Rather than replacing expert judgment, theoraizer functions as a thinking partner, helping researchers assess causal relationships and construct proto-theories. We describe how LLMs can be used to automate key steps in CLD construction and demonstrate empirically that agreement between theoraizer and human experts is comparable to the agreement observed between experts themselves. This suggests that LLMs can reliably assist in identifying putative causal structures, making the theory-building process more efficient. In addition to improving efficiency by quickly generating an initial candidate CLD, theoraizer supports creativity in the modeling process by suggesting alternative perspectives and highlighting variables or relationships that may have been overlooked. By combining the CLD method with the generative capabilities of LLMs, theoraizer drastically reduces the work required to arrive at a candidate CLD and provides users with a standardized, multi-stage framework for constructing CLDs that support early-stage theory development.