Can I use "public domain" data in MIT-licensed software?

Question

I want to use a small dataset that a university professor posted on his web page (a list of English words). The data is part of a larger corpus of data that accompanies a textbook (written by said professor). On the web page, it say that "Public-domain sources for all programs and data of [software described in the book] are freely available." I want to use this data in a piece of open-source software that I am creating, which I intend to license under the MIT license. I plan to include a copy of the data with the software source code (which I will ultimately post to Github).

My questions are:

1. Am I legally allowed to do what I have described above?
2. If so, how should I cite the data? I was thinking of writing a few sentences about (1) the link to the data itself (2) the citation of the textbook (3) the quote from the professor's web page that says "Public-domain sources for all programs and data of [software described in the book] are freely available."
3. What if I want to make my own modifications to the data before publishing it as part of my software (e.g. removing certain undesirable words from the list)?

Answer 1

You should check if the dataset you found on the website is actually the version which the professor published in PD. The language "Public-domain sources for all programs and data of [software described in the book] are freely available." indicates to me that this might not be the case and you might have to ask for the public-domain-version of the dataset.

If it is in public domain, then you don't even need to mention from where you took it. But being polite and giving proper credits to the professor is always a good idea.

If the dataset is actually in public domain, then you can obviously do whatever you want, even add or remove parts of it. That might be necessary to accommodate legal requirements in different countries.