online QDA logo - Home Page



Bookmark and Share


Semi-structured interviews

Qualitative interviews can range from fully structured to unstructured. it all depends how far the topics and questions are pre-planned by the researcher. A fully structured interview is one where the researcher has decided on the topics to be addressed and the questions (and subsidiary questions and prompts) to be used. It is very like the use of a questionnaire in survey except that all the questions are open-ended. In contrast, an unstructured interview leave it open to a large degree what the respondent will say. Researchers usually have a short list of topics they want to cover, but no questions and the topics may be covered in any order that seems appropriate at the time. The narrative interview is a form of unstructured interview and in this case the respondent is often simply asked to 'tell their story' on the topic without any further instructions from the researcher. A semi-structured interview lies between these two extremes. It will contain some key topics that need to be talked about and maybe also some questions and suggested prompts and/or supplementary questions. More structured interviews are used where the researcher knows what kinds of answers to expect. Less structured interviews are used in a more exploratory way, where the researcher is unsure of the kinds of things that the respondent will say.

For more information related to the stucturing of interviews see:

ESDS Qualidata teaching resource: exploring diverse interview types

Bias is where the outcomes of research and its results do not properly represent what actually happened, what people really felt, what they said and meant or even what data you collected.

For more information related to bias see:

Book iconKvale, S. (2007) Doing Interviews. London: Sage. See Chapter 7, Interview Quality.

And generally on the quality of qualitative research see,

Book iconFlick, U. (2007) Managing Quality in Qualitative Research. London: Sage.

Book iconSeale, C.F. (1999) The Quality of Qualitative Research. London: Sage.

An interviewThe interviews


Authors of this page:Dawn Clarke and Graham R. Gibbs

Affiliation: University of Huddersfield

Date written: 19th August 2008




Learning outcomes

  1. Learn


Frances chose to use interviews as her main data collection method. She had a choice of several types of interview approaches which may have been: structured, semi-structured or unstructured.


Loudspeaker iconFrances explained her choices:

I didn’t want to do an unstructured one [interview] because I felt that I could be too close that if I was just to go in and talk with a participant, part of my experience might get into that and it would be too close and would be one way of again putting bias into it [the research]… and corrupting the data. but what I did was, I used my experience to think right well what are the areas that it would want to look at and how would that translate in into questions.


For many researchers the reason they undertake a project is that they have had personal experience in the field that has raised questions that the research might address. Whilst this can be useful because it gives an insight into the research area it is also necessary to take deal with any biases that experience can give rise to.  Frances was aware that her own experiences could have a negative effect on her research outcomes by introducing bias into her work inasmuch as her own experiences could override, influence, or dismiss the experiences of the participants. One implication of this for her research design was that she had to choose an approach for data collection that managed potential bias.

Nevertheless, Frances did use her professional and personal experiences to develop her design.


Loudspeaker iconHear Frances explain her design.

I knew that I wanted to watch this [the effects of whiplash] over time because if the research suggested that six, twelve months and beyond they [the participants] were having problems, so were there any indicators that we might be aware of, right at the outset, that might suggest that actually this person may have problems, we [medical practitioners] may need to do other things, they [the patients] may come back into the system. So with that in mind… I wanted to see them over a year so I  got this year timescale, I would like to have done it longer. And that was based on my experience again - my problems got worse as time went on and so that’s where this [whiplash injury] doesn’t make sense, there’s some thing not quite right here so I used that [personal experience] as a basis.


Here we can see how both Frances's personal and professional experiences had influenced her approach. She already knew that her own injury had got worse long after her medical diagnosis found her to be recovered. This was the main impetus for her research. That is to say, there was a discrepancy between her personal and professional experiences; they did not match.

Furthermore, Frances was already beginning to consider where she would like the final findings of the research to be disseminated and used 'we medical practitioners may need to do other things'. This was an important consideration that informed her research design because she needed the results to be able to stand up to medical and academic scrutiny. Dissemination and the potential audience is an important element for all researchers to consider in their research design, through their choice of data collection method(s), data analysis and particularly when writing up in order for the findings to be accessible to the intended audience(s).


Creative Commons License
The resources on this site by Graham R Gibbs, Dawn Clarke, Celia Taylor, Christina Silver and Ann Lewins are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.


Case ContentsNext page   1   2   3   4   Back to overview