Interview Schedule
You have been asked to prepare interview questions to help you understand how a user will perform different tasks in a software application that you will design.
The package will follow the following outline:
Select Relevant Types of Interview Question
By the end of the package you will be expected to
The interview schedule will then be ready to PILOT,
and any changes made, where necessary, in a REVIEW. When a final version
is ready it will be possible to IMPLEMENT the questionnaire before it
is ANALYSED to answer the original RESEARCH QUESTIONS.
Defining the Research Questions
You now need to "convert" these research questions into questionnaire items that will provide you with useful information. There are a number of options available.
1. Open Questions
With open questions you do not guide the respondent in any way – you do not put ideas into their heads. In this way you are providing the respondent to answer the question in the way that they think is most appropriate. You obtain data that is both highly valid and often very detailed. The data normally comes in a great variety of forms and so analysis may take a long time. Further, in analysis, as the aim is generally to make the data quantifiable, there will need to be a large degree of interpretation.
2. Closed Questions
A variety of closed question types have been developed to make attitude surveying quantifiable. These can be categorised as
A. Bi-Polar
With these questions there are just two answers, normally "Yes" and "No" (but could also include Male / Female, Working / Unemployed etc.). While these questions are easy for the respondent to answer, and the simplest to analyse, they do not provide a great deal of useful information. Many interviewers follow up these questions with an open or multiple choice item.
B. Multiple Choice
As in testing, multiple choice items require the respondent to choose one (or more) from a list of options. In comparison to bi-polar questions, they are only slightly more difficult to answer but provide the analyst with a lot more information for only a little more effort in analysis. Perhaps the only difference lies in the appropriateness of providing only two or more than two options as alternatives. The difficulty in designing multiple choice items is in providing all suitable answers to an item, without leaving gaps and without providing alternatives that nobody will choose. As an alternative many item designers like to include a final choice of "Other (please specify)" so that the respondent has the choice of providing an alternative that the questionnaire designer had not thought of. Multiple choice items naturally lend themselves to items that ask for factual, or objective, responses. In an interview situation, multiple choice items are difficult to administer because they normally require the interviewee to look at the choices. If an interviewer presents these choices orally then it is easy for the interviewee to forget one of the options.
C. Prioritisation
This type of item is very similar to the multiple choice format in terms of design and analysis issues, but requires the respondent to process the options in a different way, adding a further category and greater sophistication to the data for analysis. The respondent is provided with alternatives, as with multiple choice, but instead of being asked to choose just one is asked to prioritise, or order, the alternatives.
D. Importance / Satisfaction
The Importance / Satisfaction item type allows respondents to express their attitudes (as they do in a Liekert scale, below), but also allows them to show the importance they give to different aspects of an issue, as they do in the prioritisation type, above. Respondents are asked to provide a score (e.g. from 1 to 4) for a statement depending on how important they think that item is, They then provide another score (using the same scale) to express how satisfied they are with the current system. The numerical responses allow for easy analysis, and as the final analysis depends on a difference between the two scores there is normally a clear indication of which problems to tackle first.
E. Liekert Scale
This type of item has become common in questionnaires, especially those that ask the respondent for their opinion on something. It is best suited to attitudes and opinions that are ‘gradable’. The respondent is asked to gauge their response to a statement on a scale (often agree to disagree). The Liekert scale has quickly become a standard for attitude testing because it is both flexible enough to allow people to express their feelings, while at the same time it is quantifiable, allowing for easy analysis. Respondents often object to certain statements, and so one option for the designer is to allow respondents to identify which statements they could not respond to.
The best way to find out useful, accurate information effectively is to prepare an interview schedule. This is a list of questions that you will ask the interviewee, with follow-up questions prepared before you go into the interview. An interview schedule should start with more open questions, that allow the interviewee to respond in their preferred way. The interviewer should then be ready to ask for more accurate information. At this stage the interviewer can use more closed types of questions, based on the interviewee's answers to the open questions.
There are a number of factors that need to be taken into consideration. The first is accuracy. The aim of the interview is to gain accurate information. This information will be essential in ensuring that the final design will conform to customer needs and expectations. An interview is a very labour-intensive research method, and so you must consider the effectiveness of your questions. You must be able to obtain the maximum amount of information from a minimum number of questions. In order to make your questions more effective they must be as unambiguous as possible. That means, that the interviewee must understand clearly what you are asking. Above all, you need to be prepared. Your interview schedule must be able to obtain ALL of the information that you will need for the design. You need to prepare more questions than you will need, to make sure that you do not miss any information.