Critical thinking with AI?
Don’t trust. VERIFY!
(👆 there’s your AI mantra for the week)
I just came up for air after a lengthy project helping someone study for the California Psychology Laws and Ethics Exam (CPLEE), the licensing exam for psychologists who practice in CA. They’d already tried one of the 3rd party exam prep companies and found it terribly insufficient.
So my part was twofold: help them master the content, and make the studying as efficient as possible.
The CPLEE is a beast. You learn all the applicable CA laws and regulations (950+ pages), the APA ethical principles, and the HIPAA regulations. Then you learn how it all fits together – perhaps the hardest part given the numerous overlapping and conflicting guidelines between the three!
CPLEE questions are mostly scenarios, not just factual recitation, and they’re nuanced and difficult. Mastering the exam requires a combination of test-taking strategy, close reading, content knowledge, and deep focus throughout the 2.5 hour exam.
So we needed to ingest huge amounts of information, analyze it to find overlaps, evaluate the conflicts, choose the best answers, and apply them to real-world scenarios. (Bloom’s Taxonomy, anyone?) We needed to build practice test questions, correct misconceptions, and address conflicts between the sources. We needed to build a knowledge base, testing strategies, and frameworks for how to reason through the questions. All in the 6 weeks before the exam! 😅
What a great way to combine learning science and technology!
On the learning science side, we leaned hard into spaced-repetition as a learning aid with Anki, an open-source flashcard software that uses spaced-repetition algorithms to enhance learning and retention. Quizzing, even (surprisingly) in advance of learning the content, is a really powerful way to learn.
And of course we incorporated AI! Multiple models of paid ChatGPT and Gemini, plus NotebookLM. (Some of the custom AIs for critical thinking / legal evaluation are pretty expensive, so we went with two lower-cost options.) I wanted to harness the potential of AI to ingest all that information and “think” critically about the content. Basically: to make a super-powered study buddy. 🦾
Plus we worked on strategy, mindset, and mental resilience in the face of a grueling learning marathon.
And finally, we also relied heavily on (drumroll please) our own brains and critical thinking! Because it turns out the big AIs just can’t be trusted to tell you facts! Wow do they love to hallucinate. 🤥
I’ll follow with more on this process over the next weeks. Lots of learnings and advice to pass along.
For now, I’ll just finish with the results: they passed! 🎉