Having collected data, it can be tempting to publish several articles on the same dataset. However, there is a difference between publishing articles with clear overlap, and distinct studies. In short, when the overlap is sufficient so that the studies could have been combined for a more impact article, it is not good. It distorts the literature and wastes the time of the reader, reviewers and editors.
Data overlap is acceptable in the case where: There are different and distinct questions with multiple and unrelated endpoints; and /or the data collection is large or drawn out that it is impractical to publish all simultaneously and other researchers would benefit from earlier publications.
To read more on this, see:
Kirkman, B. L. and Chen, G. (2011), Maximizing Your Data or Data Slicing? Recommendations for Managing Multiple Submissions from the Same Dataset. Management and Organization Review, 7: 433–446. doi:10.1111/j.1740-8784.2011.00228.x
Academy of Management ethics video series:
(Thank you to Elizabeth (Liz) Solberg for the presentation the content of this post is based on)