Comparing a Fine-Tuned Model to an Engineered Prompt in the Context of Causal Connections in a Passage of Text