Hope you all almost done with the prof mania (except the big boys of final year).
I am continuing from where I left the part 1, in which I tried to tell you about research basics of Team formation and idea collection.
The third step after you have a genuine idea is forming a closed questionnaire. By that, I mean go through the article related similar papers and find what were the basic findings by those papers. e.g. Almost all the papers start with demographic data of age, gender and other similar but relevant parameters. Go through the similar papers published, make a list of questions. The basic question is the idea of the study, i.e. suppose you are doing a study on post-op survival after a particular operation. The basic question is "Status of the patient" and options are A) Alive and B) Dead.
This question defines the dependent variable in your questionnaire on which the other independent variables will influence. e.g. effect of a particular complication on survival.
By forming a closed questionnaire, I don't mean to do stupid surveys (although they are good to start with) but surveys are public opinions that form the lowest type of research a person does, consequently, there is not much credit you gonna get out of that. However, questionnaire here means you devise a form on which you can note your findings, closed means that you choose the answer options for questions, like in the above example is either alive or dead, there is no third option, however, you might wanna add more options in age like <10>
After devising a questionnaire, you should show it to some reliable but experienced person who will tell you what changes you need to make it better. But thats an option only for the beginners, not for people who already have a couple of papers at their hands.
The next step is "getting the questionnaire filled". Now this is the real hard work, you need to do it sincerely and consistently. Auto-fill options simply don't work, SPSS will give you the proof that if you fill the data yourself, there is no means you'll get reliable results in terms of statistics, thats why there is the concept of randomization.
There are two ways to get the questionnaires filled, prospective and retrospective. Both are acceptable methods, however, the retrospective ones saves a lot of time and efforts. Prospective means you go to the patients' beds and ask them questions as well as note data from their files, and it is also better in sense that you don't miss anything. Retrospective on the other hand means you go to the files section of the wards, get the records out of tons on dust, and sit there in the back rooms full of rats and mites, but rapidly take out the files of the disease of interest and note the data on the questionnaires formed by you. But despite quicker way of data collection, you'll have to compromise at several options that won't be mentioned by some doctor who was doing a special sunday duty and was too tired to note all the findings. But thats also acceptable, as long as you mention at the end in your manuscript that how many patients data was insignificant in your findings.
Note the data as it is, don't try to manipulate it or make it significant yourself if it is not, you are not there to prove anything right or wrong that nature controls, just keep it the way it is.
Hope I've made myself clear on the above two points, well thats the half way mark you'll get to, but as I said earlier, it needs persistence and lots of efforts.
Dr. AK Malik