A problem that we here at Bioedit often come across is duplication of data both in tables and manuscript text. The problem is particularly acute when authors have to meet strict word counts.
Many authors seem to be unsure about the best way to present their data, so they tend to adopt a “belt and braces” approach and present in two different formats. However, this simply introduces redundancy, makes the text difficult/awkward to read, and more often than not involves a lot of editing to reduce the word count of the manuscript to ensure that it meets target journal specifications.
Many (although not all) journals will include a statement in the “Instructions” page to the effect that “Data presented in tables should not be duplicated in the text”, but this seems to be widely ignored.
As a general rule, we suggest that were possible data should be presented in table format. This usually means that it is visually more appealing to the reader and complex relationships/correlations and easily be seen and appreciated. Trying to explain a complex data set in the text requires a considerable amount of grammatical/scholarly gymnastics. The result is often rather disjointed, difficult to read/understand and confusing.
It is much better to include the data in a table and then simply provide the reader with an overview, referring the reader to the specific Table where necessary. For example, the authors may simply state that x is larger than y, that x fell by more than y, or that X in group A was larger than X in group Y, without actually having to mention specific numerical values.
Avoiding duplication will mean that authors have more room to discuss the implications of their results and avoid drastic editing/cutting to meet a specified word count. As more and more articles are submitted for publication year on year, the space available in journals is becoming more restricted (this is still true for journals that publish online). Concise, well written and easy to read articles will have a greater chance of acceptance; not cluttering your work with data that can easily be presented in table format will go a long way to achieving these goals.