HI- there really is no 'silver bullet' answer to this. I will try to give you a pragmatic answer rather than a text book answer and note, I am answering from primarily the perspective of collecting ‘interview data’- Starting at the beginning I believe that the interview questions should be aligned to the research questions/objectives/purpose of the study. This will ensure (or try to ensure) that the data derived from the interviews are relevant in the first instance. The actual conducting of the interviews per se must also be well balanced ensuring that respondents do not sway far away from the interview questions when answering. Hence, researcher interview skills are necessary. You may also ask questions that are double validation based (asking the same question but in a different way) to ensure validation of responses. Thereafter coding of the data is very important and in this case, it depends on how you choose to code and/or the software platform you use. Eg. Nvivo makes coding and analysis easier utilising various features within the program, but there are many more out there. You can also adopt a framework for coding and analysis. I generally adopt Braun and Clarke (which you can google and see the steps involved). Whilst analysis of qualitative data is inductive whereby the data predicts the themes, I do tend to align themes to the research question/objectives of the study. One should try their best to eliminate subjectivity and bias from the analysis. However, qualitative analysis is generally known to be subjective due to the nature of responses received and interpretation thereof.