How many interview need to conduct in a qualitative research?

Home Forums Default Forum How many interview need to conduct in a qualitative research?

Viewing 15 posts - 1 through 15 (of 23 total)
  • Author
  • #3890
    Kien Nguyen

    Hi everybody,

    My name is Kien. I am a researcher. Can you answer some questions for me? Thank you firstly.

    Here they are:

    – How many interview need to conduct in a qualitative research?

    – What type of interview (semi-structured or unstructured) will you use?

    – How long do you usually conduct a interview?

    Thanks again.


    Hello Nguyen,
    Your all questions’ answer is that it will depend on the question that you are intended to answer. Furthermore it will depend on the theoretical framework and nature of inquiry.


    Jenny Hall

    Hello Kien
    I think Dave has answered all your questions. One of my friends did her PhD studying just one person!. Also don’t just assume you have to interview- there are other qualitative methods for collecting information.
    best wishes

    Kien Nguyen

    Thanks Dave for your detailed and useful answers.



    Kien Nguyen

    Thank Jenny and Muhammad.



    Sam Stott

    hi kien,

    you’ve specified which paradigm (qualitative) but not which methodology you are using. this will determine if and what kind of interviews are suitable methods for collecting/generating data.

    in qualitative research, you’re likely to need enough data (from whichever methods chosen) to reach theoretical saturation.




    How many interview need to conduct in a qualitative research?
    To my understanding you need to conduct interviews until you reach the point of saturation and redundancy starts in the interviews. However, you have to consider the approach you are using and the time-frame you’ve got to conduct the research.

    What type of interview (semi-structured or unstructured) will you use?
    Depends on your research question, methodology, the context of research and the type of informants. To my experience, it’s better to combine approaches. Unstructured (informal conversational interviews) are rich source of information, and rapport building is easier here.

    How long do you usually conduct a interview?
    There is no hard and fast rule about this. Basically it depends on the type of interview (informant or key-informant), the amount of information you want to get from a interviewee, and the amount of time s/he can give you. Dave made wonderful explanation about this.

    Best wishes.

    Betty Akumatey

    I suggest you read the ff 2006 article: “How Many Interviews Are Enough? An Experiment with Data Saturation and Variability”
    Greg Guest et al.

    Kien Nguyen

    Many thanks Betty, Gorkey and Sam for your useful suggestion and guide.



    Hi Nguyen,

    To answer all these questions, my suggestion is the book: “The long interview” of Grant McCracken.

    Until next time.



    Hello Betty,

    I liked your hint.

    Thank you.


    Kien Nguyen

    Hi Fernanda,

    Thank for your help.



    Roger Gomm


    Hi Kien

    How many interviews ? That really depends on the scope of the claims you are going to make for your findings. In simple terms there are two kinds of research: one kind collects information about how people behave, think,feel and so on. Here if you are happy just to claim that your findings apply to ten people, then ten interviews is sufficient. Ten might be sufficient to make claims about a larger number of people if you can find a convincing argument that these ten can stand as representatives of larger numbers who would have said much the same in interview as did those you actually interviewed.Much will rest on how diverse is the wider group towards whom you are generalising your findings – the more diverse it is, the larger the number of interviews you will need to conduct in order that your small group of interviewees covers the range of diversity in the larger group (diversity here as relevant to your research topic). Alternatively you can restrict your generalisations to a rather homogeneous wider group.


    The other kind of research collects information from people about specific events or contexts, for example about a road traffic accident. The number of interviews you need here is the number of people who have important information to tell you about the event – one might be enough, but enough to cross-check stories is better.


    If you take the grounded theory route and follow the advice of those who say you need as many interviews as are required to reach theoretical saturation, the how many question can’t even be given a speculative answer in advance. The number will depend on your theory and your imagination in plotting routes of constant comparison, as well as contigent factors such as the number of red herrings which are thrown up by earlier interviews. However, the answer is almost bound to be more interviews than you have time for.



    Hi Betty,


    The Guest et al article is interesting, but I wondered whether it counts as grounded theory, particularly in terms of the idea of a standardised codebook, and reviewing and revising it every six interviews.  Do you have any thoughts?



    Daniel (“applied” researcher in HIV in Melbourne, Australia)

    Kien Nguyen

    Hi Roger Gomm,


    Many thanks for your answer. It’s so detailed.


    Best Wishes,



Viewing 15 posts - 1 through 15 (of 23 total)
  • The forum ‘Default Forum’ is closed to new topics and replies.