From Prompt to Scale? Examining the Effects of Prompt Instructions on LLM-Generated Scales | Synapse