I think that discussion section is the most difficult part of a paper because the researcher must read many other studies and choose the most common to his work comparing all the results.

I too agree with Maria. Whenever writing a discussion to any paper, it needs a lot of literature search and then conclude the studies with relevance to the previous literature adding what new the study is going to add to the existing scientific knowledge. SO discussion needs to be very precise, informative as well as conclusive with sufficient evidences.

Other than these, I feel writing an abstract is also a very important skill. Here the word limit sometimes makes it very difficult to include each and every part of the study. Sometimes the situation becomes like 'What to include and what to leave', though everything seems to be important as far as study is concerned.

For my part the abstract and the title are always the toughest to write. They deserve extra attention because they might be the only part of the article people actually read. So they should not only be good enough to trigger the interest of the reader but also contain your the key message you want to convey.

I totally agree with Elsa. The abstract is the teaser. After reading the abstract the reader decides whether to go on or not. Therefore the abstract should not only highlight the findings but should clearly state what is special about the findings. Most importantly the abstract should provide the significance of the findings to a broader community.

For me: taking care of the journals style requirements, for both, text (# words) and figures (size, layout limitations). Preparing good figures takes the most time, and it is the first thing I do, so I get a good overview over the results and the story. Writing M&M and results then is simple, since it is just a report of what was done and observed. Very important and more difficult is the abstract, for which you have to have the story plot in mind and get it to the point. The last thing I write is the discussion, and this is quite difficult for the reasons others mentioned above.

Probably the hardest is deciding on the story-line - what data goes first and how does it all fit together. The bit I hate the most is proof-reading.

I agree with Jochen, one of the hardest things is getting the figures just right and adapting to a journal's style requirements. But of course, the hardest thing of all is waiting for the journal to get back to you with feedback on the paper once you have submitted it :)

I agree with Maria and Alok, discussion is the most toughest part as you need to row through literature and find the relevance to your search and to show that you added something new to the field and not only repeating others work. For other parts of the article its easy to finish it as you need to follow the instructions to author to the journal where you want to submit your article to. For the abstract although its very important but sometimes you are restricted to the journal instruction as some Journal put limitations on the number of words in the abstract and this would not give you enough space to add all the important findings in the research

Thank you all for the responses. I too agree with most of you, Discussion is a difficult task. Framing a good abstract is also a tough job. Most of the people read abstract only and they read the paper only if abstract is appealing enough.

I agree with Rana. However, once I have good solid results, I don't worry much about writing!

All the best