How to set up an article writing machine (Re: Writing Scientific Articles)

[Guest guest]

These questions that Vijay has put up are great pointers in writing scholarly articles. Indeed, scholarly research is a highly writing intensive task. For writing primary research articles and systematic reviews, a practice that we have found very useful over the years is to do the data analyses first.(assuming that the data have already been collected), set up the tables, summarize the key points from the results tables, and then write the summary of article or the paper that we are going to present first. Everything else follows that summarization process.The tables as derived from the data analyses can undergo several iterative changes as we continuously debate our findings to test that the findings are indeed robust based on the data we have in our hands and the findings that we have found. In general, we bias against ourselves at every step in the data analysis and argue that the findings may not be true. Also, we try to check our findings through at least more than one means to see if the different approaches produce similar results. If they do, we know we are in the right track, if not, more analyses are needed or we refine our data analytical strategy.Once the summary (it does not matter whether the specific journal requires us to send a summary or not, summary writing is for our quality control purpose, and the summary can serve as the starting point for writing abstracts whenever required) is finalized based on the tables, we write the results section and do some research and try to explain the findings, the natural extension of the findings etc for the discussion section. After that is done to our satisfaction, we go for a thorough description and writing of our methods section, keeping in mind particularly about the steps we had taken about sample selection and what did we do for removal of bias (one way or other). The introduction and the background section is done last, tying into the main message of the paper. Every research paper should tell a story, some are exciting like a short story with a tail hanging in the end prompting the researcher and the reader to embark on a new journey. For example, causal research typically present findings that encourage others to replicate the original work. Other researches tell other types of stories. Building a theme based message around a central storyline is a very important task before the student or practitioner who has to write a research report. The more attractive/powerful the story, the better the report reads.Best,ArinSent from my iPadOn 28/03/2011, at 3:30 AM, "Vijay" <drvijaythawani@...> wrote:

Since we mostly write these, we will discuss this in greater details.

The most popular style of writing these is IMRAD which stands for Introduction, Material & method, Results And Discussion.

In these sections we answer the questions:

Why did you start? (Introduction)

What did you do? (Methods & material)

What did you find? (Results)

What does this mean? (Discussion)

Vijay