formulas for writing publishable papers

Question

I have started a PhD on a subject I have quite some industry research experience, just to get the title.

My approach to the PhD is to come up with a novel solution or method and tell all about it on a paper (the sexy, the good, and the bad) as an independent observer. However my tutors seem to be uneasy with my approach and they claim that for a paper to be publishable I should omit/minimize the bad things, although they know that doing so is bad science.

I have the impression that people in the academia might have formulas for writing publishable papers. After all, to publish seems to be the only valid metric in Academia.

I would appreciate if someone could point out some of the non-written rules. That would help me to avoid the frustration it causes me the writing of academic articles.

Thanks.

PS: I ask this because I want to avoid ending up like this folk.

Answer 1

First, let me state that in my experience, there is no such thing as a formula for writing papers. There are too many different types of results to be presented in too many different types of papers.

That said, however, there are most certainly patterns of presentation and argument that emerge as being particularly effective. In my observation, many successful researchers have some form of such patterns that many of their papers adhere to. Ultimately, for a researcher with integrity, this is not about chasing a metric, but about effective communication and about having an impact on their research community and, hopefully, the larger world.

That said, I think that the approach you propose of relating your work as though you were an independent observer, is not a good way of presenting, because you are not an independent observer. An approach of "minimize the bad" is even worse, because then you are attempting to conceal the truth. Instead, I recommend thinking of yourself as an "honest advocate" for your work.

You are a human being who believes the work you're doing is good and producing interesting results: in the introduction and related work section of the paper, you explain why you think the problem you're working on is interesting and how the results you've produced fit into the larger world. You must, however, be honest with yourself and others about the strength of the results that you have found. That means:

• Be scrupulous about distinguishing between what your evidence shows and its relationship to your hypotheses: e.g., "the data is consistent with our model, though not strong enough to rule out the alternate hypothesis"
• Don't value judge "good" and "bad," but instead report the circumstances under which statement is known to hold or fail. It's OK if something only holds in a narrow case if that case is interesting and possibly important.
• Publish early and often, so that you don't feel like your papers are so high stakes. You might not get "glamour journal" papers, but you won't need to worry about being scooped, and you can get credit for the knowledge byproducts discovered during your search even if your ultimate goal is not achieved.

More important and more difficult than anything else, you need to be prepared for the fact that you may find your work has been going down an unproductive path and that you need to change your direction.