The interview is a method for discovering facts and opinions held by potential users of the system being designed. It is usually done by one interviewer speaking to one informant at a time. Reports of interviews have to be carefully analysed and targeted to ensure they make their impact. Otherwise the effort is wasted.
Because of the one-to-one nature of the interview, what is talked about can address directly the informant's individual concerns. Mistakes and misunderstandings can be quickly identified and cleared up.
Consider the information you require, and prepare an 'interview schedule'. This is a set of topics that you need to obtain that information. Decide on the order in which you will cover the topics. For each topic, ensure that you have an 'askable prompt' (how you will ask for the information you need) and an explanation of each topic (in case the informant does not understand the 'askable prompt'). If you want to do a highly structured interview, each topic will be broken down into a series of sub-topics, each with their own 'askable prompt'.
The ideal interview situation is composed of an interviewer and a respondent. Several respondents may be interviewed simultaneously. If there is more than one interviewer, there should be one principal interviewer or chairperson. There should never be more interviewers than there are respondents.
Decide how you will record the informant's responses. In order of preference, these are: your memory, concurrent written notes by yourself, tape recorder, video.
Decide how you will present the interview results (make an indicative table of contents), and check with the intended audience that this is useful for them.
Do at least one trial run of the interview. Make sure you know the interview schedule extremely well (by preference, commit to memory.)
There are typically four phases in the interview:
- The "nurturing" phase. This is the initial warm-up to the interview when the parties to the interview introduce themselves and talk briefly about neutral topics to establish themselves.
- The "energising" phase. Here the area of discourse, and any existing problems are identified.
- The "body" of the interview. This is the peak phase of activity, where the interviewer is continually probing, bringing out the 'askable prompts' in the predetermined order to understand the range of responses the respondents produce. It is important at this stage for the interviewer to remain analytical and neutral. If the interview is fairly free in structure, the respondent may direct the order of topics, and the interviewer should follow them. Otherwise the order of topics is at the interviewer's discretion. Before this phase ends, the interviewer should check whether all the topics have indeed been covered.
- The "closing" phase. The interviewer should summarise what he has learnt from the interview, and ask the informant whether this is correct. The informant should be asked whether they thought the interview covered all the areas of concern, and whether there were issues which had not been touched upon. It is a good idea to spend a little time on how the informant felt about doing the interview, and whether there was anything that could be improved.
The biggest danger in using interviews as methods of data gathering is the unstructured nature of the resulting data, which is extremely easily mis-interpreted or censored.
The primary method of analysis that helps guard against censoring information that is difficult to handle or unexpected is to break up the text or notes from each informant into a set of simple propositions, using the informant's own words as much as possible. These propositions can then become the input to a content analysis activity.
The analysis should fit into the indicative table of contents agreed beforehand with your target audience. Notes and transcripts etc. may be contained in an appendix or appended CDs/ diskettes/ tapes.
Preece, J., Rogers, Y., Sharp, H., Benyon, D., Holland, S. & Carey, T. (1994). Human-Computer Interaction. Reading MA: Addison-Wesley.
Macaulay, L.A. (1996). Requirements Engineering. Berlin: Springer Verlag Series on Applied Computing.
When the interviews are over and the report has been written, ensure that the report makes its way to those people who will be most affected by it, and that it has been read. Follow up the initial report distribution within a week or so to ask if there are any questions or if any explanation is needed.
Interviews are sometimes used as methods for generating information for running a focus group or setting up a survey questionnaire that will be distributed to many people. However, they may sometimes be used as a direct input to design, so that the next activity to be carried out is either a card sort / affinity diagramming, requirements meeting or a paper prototyping session.
Lindgaard, G. 1994. Usability Testing and system evaluation. Chapman and Hall.
Blomberg, J., Giacomi, J., Mosher, A. & Swenton-Hall, P. (1993) Ethnographic field methods and their relation to design. In: Schuler, D. & Namioka, A. (eds.) Participatory Design: Principles & Practices. New Jersey: Lawrence Erlbaum.
Fowler, FJ (Jr), Mangione, TW (1990) Standardised Survey Interviewing. Sage Publications, Newbury Park.©UsabilityNet 2006. Reproduction permitted provided the source is acknowledged.