Have you set your mind at conducting a survey? Good choice. But, in case you have chosen them because surveys seem to be the easiest option, then think again. Surveys are a popular choice among new researchers but they require far more expertise than that meets the eyes. There are plenty of ways in which your survey can go wrong. It is not an easy task to pull off a brilliant and successful survey and overcome the many flaws its methodological design is prone to. Here are a few steps (right from the beginning to the end) to help you execute an almost perfect survey for your thesis or your research paper. Ask yourself “why” you wish to conduct a survey and note the answers you find yourself coming up with. The noted answers miraculously become your research question/s. Conduct a literature search on the already existing published knowledge regarding your area of interest. Take a note of the pitfalls of previously conducted surveys, so you can avoid making the same mistakes. Decide which type (face-to-face interviews, telephone, online, or mail) of survey answers your question the best. Decide if a structured or an unstructured survey will best answer your question. Write down a list of questions (open-ended or close-ended) that helps you obtain a detailed answer your research question. Keep your questions short, simple, and precise. People are more likely to answer yes/no questions than those requiring descriptive answers. Choose a demographic that is best suited to answer your research questions, and ensure that your chosen demographic will allow generalization of your results. Undertake a power calculation to ensure you have the exact number of people who need to participate in your survey for the validity of your results. Test the validity and reliability of your questionnaire before you send it out to the participants to avoid faulty data collection. If you are conducting a telephone or personal interview, then undertake thorough training of the staff administering the questionnaire. The manner in which the questions are put forth impact the answers to them. Reach out to more participants than suggested by your power calculation. A large sample size ensures a more rigorous statistical inference. Compensate the participants for the time invested by them, if that is allowed. Pick the right or appropriate time to ask your questions. Participants who are pre-occupied or rushing will probably not provide honest or meaningful data. Devise a method of storing your data so that you can refer or obtain additional data from the data collected without any confusion or complications. Allow the possibility of missing data when coding your data. Data coding and entry is better done twice separately, because it not only helps you tally the accuracy of the data but also validate your codes. Begin statistical analysis on when your data set is positively “clean.” Use a mixed method approach to triangulate the results of your study. Surveys usually are well-known for insufficient response rates and inaccurate data. One must develop strategies that will enhance better and more meaningful participation of your targeted demographics. Approaching your participants in a manner that will encourage them to take a genuine interest in your survey is of utmost importance, and all researchers who wish to administer surveys must bear this in mind.
Topics
