Prompt Engineering for Psychology
LLMs are already being used to code interview transcripts, synthesize literature, screen participants, and deliver interventions. They also raise questions about bias, consent, and behavior that psychology is uniquely positioned to answer.
Where this is showing up in Psychology
- Tools like Elicit, Consensus, and ATLAS.ti's AI coding are being used for literature review and qualitative data analysis in labs worldwide.
- Woebot, Wysa, and Dartmouth's Therabot trial represent an early wave of LLM-based mental-health tools — with mixed clinical evidence and active ethical debate.
- The APA, JAMA Psychiatry, and Nature Human Behaviour have published guidance and critiques on AI in clinical and research practice.
- Research from groups like Binz & Schulz (PNAS, 2023) shows LLMs replicating — and sometimes amplifying — human cognitive biases, opening a new line of psychology-native AI research.
Projects you could build in this course
- A tool that codes open-ended survey responses against a theoretical framework
- A RAG assistant for literature review across a specific area of psychology
- A journaling or reflection tool that demonstrates the design tradeoffs of AI in mental-health contexts