Questioning techniques can be subdivided into individual interview methods, focus groups and questionnaires, often administered as organisational surveys. In all cases, the job-holders (usually experts) are asked a number of questions about their work. This may be to solicit their knowledge of the job, key tasks and their relative importance, and the cognitive, social and other skills needed to complement their technical expertise. In addition, interviews and surveys can be used to ask about attitudes, safety concerns, typical workplace behaviours of self and others and aspects of organisational life that influence work behaviour.
Interviews are frequently used to examine system usability, reactions, attitudes and job analysis (Stanton et al., 2005). interviews can be structured, semi-structured or unstructured (Gillham, 2005; Kvale, 2004), as described below:
• Structured interviews use pre-determined questions, either open or closed, dependent on the data requirements to elicit information from the interviewee. Benefits of structured interviews include ease of administration and collection of focused data. A further benefit of a structured interview is that it facilitates comparison of responses from different respondents to specific questions. However, this type of interview demands that the interviewer has relevant background knowledge of the topic under examination, and requires the interviewer to access static knowledge that the interviewee must verbalise in some form.
• Unstructured interviews comprise open-ended questions, encouraging substantial discussion and interaction between the interviewee and interviewer, and does not pre-suppose any previous knowledge of the topic by the interviewer (heiman, 1995). unstructured interviews provide a useful introduction to the domain for the interviewer, as well as increasing familiarity with the appropriate terminology.
• Semi-structured interviews, on the other hand, generally use pre-determined, but open-ended, questions, which can be adapted to individual requirements, and allow a certain degree of flexibility in exploring a wide range of issues that may arise during the interview session (Breakwell et al., 2006). the questions are designed to obtain information about how an individual performs the subtasks of the main task under investigation, thus is very useful for collecting task knowledge and information about certain types of skills, such as procedural skills. Benefits of a semi-structured interviews include the efficient collection of declarative knowledge, further exploration of issues raised by the interviewee, and a structure for a set of consistent questions such that the interviewer can react to the information given and seek clarification if required. Furthermore, unanticipated topics can be discussed (Redding and Seamster, 1994), and the interviewer requires relatively little prior knowledge of the domain beyond necessary terminology. one of the main disadvantages of semi-structured interviews is that, similar to structured interviews, the interviewer generally requires some understanding of the topic area to know when to follow up particular points.
Each interview type has its advantages and disadvantages (see Table 9.2), and the selection of interview type depends on the purpose of the data being gathered.
Table 9.2 Advantages and disadvantages of types of interview
Type of interview
• ease of administration
• focused data
• ability to compare responses
• requirement for relevant background knowledge by interviewer
• lack of flexibility to follow up topics raised
• flexibility to follow up topics raised
• some previous
relevant background knowledge required
• no previous relevant background knowledge required
• minimal preparation
• digress from purpose of interview
Semi-structured interviewing can also include techniques under the heading of diagramming. Diagramming techniques, e.g. laddered grid (Burton and Shadbolt, 1987), refer to the construction of a diagram representing relationships between particular tasks or job concepts, within a domain involving the relationship between elements, or subsystems. For example, experts can be provided with cards and asked to arrange the cards to represent the relationships between them (Hoffman et al., 1995). In addition, acquired data can be used to validate results from scaling and rating tasks, and can also be used to direct a structured interview. Disadvantages are that it relies on key points from the task or job being known beforehand, and that it is more suited to rule-based or procedural domains.