After presenting a paper at a conference last year, I was invited to submit an expanded version of the paper as a journal paper. The review of the paper has recently been concluded, and the editors recommended acceptance subject to minor revision.
While the paper was under review, however, we developed a more accurate approach which reveals that the finding reported in the paper was too optimistic and what was proposed in that paper did not seem as promising as previously thought.
I have been contemplating which action to take:
- Withdraw the paper.
- Revise the paper significantly to include the new methodology and the new results.
- Keep the paper nearly as it is, only addressing the reviewers' comments (none of the reviewers questioned the previous approach), but adding a note that our latest approach produces something quite different.
I am personally not comfortable with option 3, although it may be the easiest and fastest thing to do. I am deliberating between options 1 and 2. Option 1 because the result doesn't look as promising as before and may not be interesting enough to warrant publication. However, if I withdraw the paper, there is no chance to warn readers that the results presented in the conference paper were too optimistic. Hence option 2, which is the option I am inclined to take. It however needs more time and I am not sure what to say to the editors.
A good review panel would not select your paper purely due to the impact of your result but also with respect to your approach to solving the problem. This may include how existing methods were modified and how novel methods are applied. This would also concern your analysis of the methodology along with related results.
Withdrawing the paper (option 1) should be the last thing you ought to be thinking about. That is unless you plan to publish it in a different journal after much improvement, but this also has its own implications.
There is nothing wrong with going for option 2. In fact it is to be most welcome in the ethics of research. Hopefully the reviewers would also appreciate such honesty.
Most may prefer option 3 as this would mean that your paper is published without any further ado. But since this makes you -- the author -- uncomfortable, you rather don't do it. It is not worth the feeling of regret once it gets published.
Convey your points precisely and see if you could try to describe the fallacy of the overly optimistic result in your communications. You may also include this as part of the updated manuscript if possible. Note that all communications to the reviewers must pass through the editor as part of the procedure.