Rewriting for the data you collect

Kerry mentioned in class that the first writing up for qualitative research paper will always be horrible. Some unexpected or discouraged things may happen: the results from the data you collected do not match the theory; the results of from the data may be totally different from the theory the paper is arguing; the results from the data may not be the one you expect. From Kerry’s experience, a good qualitative research paper has to be wrote, improved and garnished many times (at least three times).

I totally agree with Kerry. From my experience, even for a quantitive research paper, it may need to write many times before it is ready. For example, the results from regression may not be significant, which requires the reshaping of the model; the results imply extra findings, which needs us to add up additional points that is not expected at the beginning; the results seems weird, which requires us to delete some extreme data (e.g. outliers). Furthermore, researchers have to present their draft research paper in front of colleagues or conferences where a lot of suggestions and recommendations come in. This requires a lot of amendments for the research proposal. Lastly, even when the research paper is sent to journals for publish, there may be still a long way to go before it can be published. I heard from supervisor that it is very likely or even impossible for a paper to be published when it is first time submitted. The reviewers may make many comments on research draft. I know one friend that it took him eight years to publish one of this paper. That sounds scary, but I know that is the value of published papers.

But there is two things I am wondering. First, we have to rewrite and reshape our research draft if the results from the data are not consistent with our expectation or are weird. Will this distort our original idea (e.g. the evidence we collect may not be the one testing our original research question)? Does this mean the data we collected does not match our research questions and theory? Is it a good idea to recollect data that is more pertinent to our research question and theory? Second, sometimes, we can not directly ask questions from participants. Instead, we have to rely on interviewees’ answers and guess/infer the implications for our research questions. I am wondering is this ok because our thinking may be different from what  the interviewees think? Also, how about interviewees intentionally misguide researchers in their answers?