I think there are two situations where your supervisor might ask you to continue a project someone else started:
- Someone produced some data, but never wrote a paper about it
- Someone wrote a paper and your supervisor thinks there is more to investigate
My question is:
How can I decide whether it is worthwhile to continue a project or not?
And I ask this question independent of whether the field of research is worthwhile.
You might also ask my question this way:
What are clear indications that my predecessor did sloppy work that might look good in a paper or presentation, but might become very annoying if you trie to understand in detail how he/she did it.
I mean first of all there are technical things:
- Did your predecessor leave a structured folder system of his work?
- Did he (I'll stick to he) leave all raw data?
- Are the sampling methods well documented?
- Did he document more about his analysis than you can find in his paper?
(papers are sometimes not very informative about how it was done, I mean how it was actually, actually done)
- What analysis tools did he use? Will it be available to you? How good will it be replicable? (Think of software compatibility and licensing issues)
- Is there a "source code" of the analysis that you will be able to reexecute and modify or will you get only vague descriptions of how the analysis was done?
- If it involves code: how well is it documented (commented)?
The social component:
- Is your predecessor still approachable and how (meet in person, phone, e-mail)?
- How willing is your predecessor to still contribute to the project?
- How much time does he actually have to support you?
- Is he still interested in further insights about his past project?
- How much does your supervisor really know about the project? Will he be able to help you with details?
There might be other things to consider:
- Will I be able to produce substantial additions to my predecessor's work? Or will my work be the "and someone else tried this with it". You either might benefit from the fame of your predecessor's work or might stunt in the shadow of it.
Of course my supervisor should know better than me, whether its worthwhile. He has a better overview in the field of knowledge and he probably supervised the project, so he was somehow involved and should know what was done and how it was done. But I think you can't always trust the judgement of your supervisor. Especially if he is a visionary, he might see more in the data/project than there actually is. He might see more what he wants to see in the data than what there actually is. And my experience is, the higher up someone is, the less he really knows about the projects he is "involved" with.
It would be good to have sort of a checklist. When most of it is checked you could say "Ok, let's start the project" and if not it is a clear indication of "Keep your hands off this project".
Maybe in a continuing question I could ask "How to leave a research project for someone else to continue" or "How to work in a way that someone else can join in or continue on his own at any time".