Why do you do a pilot study? You do it in order to make sure that your methodology will generate the desired results. It is a reliability and validity check for your research method. You do not want to conduct your research method on your entire sample, and then discover different results than what you wanted.
For example, you have developed a survey that no one else has tested before to achieve a specific objective. You are planning to run the survey over tens or hundreds of participants. You may want to run this survey on a small number of participants in a pilot study.
In the pilot study, you almost do a qualitative mini research. You will be giving your survey to a few individuals who will answer the survey. Then you may sit with these participants, and make sure that they understood the survey questions correctly. Then, you will make sure that the way they answer the survey questions is exactly how the questions are intended to be answered.
During the pilot study, there is a good chance that you will better understand your topic. You may realize that you need to remove, modify, and add a question. You may want to reword the questions so that it implies the intended purpose of the question. Alternatively, you may realize you want to go one step backwards to achieve your purpose.
A pilot study is needed every time you do not have enough evidence that your questions have been tested. If you have translated questions from another language, then you need to test the new language wording. If you have modified existing survey, you want to make sure it holds the same level of validity. If you have developed your own questions, you want to make sure they are valid and can achieve their desired results.
During the pilot study, you will get a mini result of your research. These mini results provide you with sample data of your final results. It is a good choice for you to try the analysis you have planned for to make sure you know how it will turn out to be.
Doing the analysis based on the pilot study will pave the way for the final results. You may build your data sheet in your analysis software such as SPSS, Nvivo, or any other software. This same preparation will be used later on. This process may give you hints to modify the final methodology to better suit the following analysis stage.
For example, you may find out that the data you collected has open question that can be a multiple-choice question. This will help you a lot in the analysis later on. Another example, you may find out that you missed an important question that can enhance the validity of the survey sample. Let us say you forgot to ask if the respondent is over 18 years. How can you make sure to remove those who are not 18 years old from the final responses? One way would be having such question in the survey.
Study validity and reliability are major foundations for the entire study. If your methodology has been proven invalid or unreliable, then your entire study is meaningless. Pilot study helps in achieving reliability and validity.