What are you analyzing?

In qualitative methodology relating to interviewing people, we do not use a "sample" because we are not trying to generalize information based on the people we interview. We use "subjects" or "participants."

Creswell (an internationally-recognized authority) recommended 5 to 25 subjects for phenomenological qualitative research. In addition, Giorgi (2009) felt that three subjects was sufficient for phenomenological inquiry in human science, if the participants clearly had the experiences the research wishes to explore.

If you have a data source that has a large number of data points, such as if you want to do content analysis of hundreds of news stories, it might be appropriate to analyze a subset of articles that is selected randomly. 

Thanks.

Aim is to analyze customer reviews about services on third party websites.

The number of reviews increase on regular basis.

Hi, Given that you are looking at reviews, and hence in a way "interviews" as Michael Marek suggested, in our experience it is best to stop at what is commonly called "saturation point", that is when an additional review or interview doesn't add any longer any new, not yet mooted or expressed information to what you know already from the reviews/interviews analysed up till then. Depending on the issue, that could mean going up to some 40 interviews, but after that, most of the time, it is simply a repeat of what has been said previously. Hope this helps

Hi, I agree with both researchers. The best way to get to "saturation point" is to start analysing your data, and than you will be able to see if any new data (=interviews) that you do, realy add new themes that you haven't heard before, or are they only repeating what you already heard before. Good luck! 

Thanks Michael W. Marek, Johan De Rycker  & Sigal Barak-Brandes. Your inputs were really helpful. Can proceed with better clarity & confidence now. Regards.

Allow yourself to be guided by the concept of empirical adequacy which is a principle that encourages researchers to not worry for expanded sample size so long as the size they choose could provide a pool of data necessary to enable them attain not only their research objectives but also enable them answer their research questions

Thanks Fred Amadi, your inputs added further dimensions in solution. Regards

I agree with the idea that to arrive at satuation point is enough.

Thanks Bashir.