Should I reproduce figures from other publications in my thesis to make it more self-contained?

Question

I am working on a computer science master's thesis wherein I occasionally cite figures and tables from other works. Currently, the way I am doing this is something like

Foo et al. (2021) visualise that this and that is detectable (see their figure 2).

... where "Foo et al. (2021)" is hyperlinked to the bibliography at the end of the thesis with LaTeX.

Personally, I think this is sufficient information for the reader to know where to find the information I am discussing. On the other hand, I am biased because I have already read the research and hence have an imprint of the figure in my head (and/or have the figure pulled up in a separate PDF). I have a feeling that just pointing to "their figure" is somehow sloppy. I think my doubt comes down to these worries:

1. If the source is behind a paywall (this is never the case for me, though), the reader might not have access to the figure.

2. If the source is hard to find (which sometimes happens with really old sources) or even disappears (if they are published on university servers instead of official publishing servers), the reader might never have access to the figure.

3. Even if the source is two clicks away (jump to bibliography and open in browser), the reader still has to stop reading and do all of that, and also now has to scroll back up to where they jumped away from. Backreferences probably won't work since I cite most of my sources at least three times.

4. The reader needs an internet connection.

5. The reader needs to be reading on the computer.

6. Unlike internal references to my own figures, I have no hyperlink that specifically points to "their figure", only to the document in which it appears. This breaks stylistic consistency.

Should I reprint the discussed figures and tables in my thesis? I would of course put an appropriate citation alongside them to make it obvious that these aren't my own.

One immediate objection I have is that we have a page limit, and hence reprinting figures when they exist elsewhere seems like pointless inflation of valuable space. Putting them in appendix would have the same issue with jumping away from the text, but at least the reader doesn't need internet.

---

Note: this is not a question about whether I can reprint a figure/table. It is a question about whether I should.

Answer 1

Generally, a thesis needs to be more complete (self contained) than a normal paper, though the advisor can give more detailed guidance about what they consider acceptable.

But, if you want to reproduce the images and such from the work of others, you need to deal with copyright issues. Your citation comments seem appropriate and you are avoiding plagiarism, which is the other major issue - good job there.

But, especially, for images, care is required. An image can carry a lot of information so reproducing a published image (or even an unpublished image created by someone else) is different from quoting a few lines from another paper.

There are some exceptions to copyright law for scholarship, but they don't completely override the requirement to quote "sparingly". If you want to use such images, the best recourse is to ask the publisher for permission. It is likely to be given so long as your work doesn't diminish the economic value of the original. Things behind paywalls, will, of course, have their own issues and calculations by the publisher.

Even reproducing figures has issues, just as over-paraphrasing does.

But, for an academic work, pointing to originals is normally sufficient, as the likely readers also probably have, or can obtain, access to the things you point to.

The dance is a bit subtle, of course. Talk to your advisor.