Learning Data Analysis With ChatGPT, as It Looked in 2023

Hong Chang-su’s book came out in May 2023, when many readers had only recently tried a conversational AI. I came from the humanities and had let programming slip away after changing fields. The book made me think that a chat window might give me a way to start asking questions about it again.

Begin With a Question You Can Check

The examples range from writing a small Python program to explaining a statistical concept, drafting a query in SQL or approaching a spreadsheet with VBA. I found them easiest to use when I could say what I wanted to learn and test the answer myself. A short example with data I already understand gives me somewhere to begin.

The book also imagines asking for career advice. A conversation can suggest paths or questions you had not considered. What it says about a particular employer or role still needs to be checked elsewhere. The same applies to code. Run it with inputs you understand and see whether the result matches what you would expect.

What the Book Practices

The appendix collects prompts about Python, introductory statistics, machine learning, R, SQL, VBA and financial data. Together they make the book feel closer to a set of practice sessions than to a promise that one program can make you an analyst.

Keep private records out of examples unless you know you are permitted to share them in the environment you use. Interfaces and controls have changed since the book was written, so I read its screenshots and product instructions as a record of that early period.

I did not come away feeling I could skip the foundations. If I understand what a column means, what an average leaves out and what I should expect a query to return, I can ask a better question and recognize an answer that deserves another look.