This Chapter is about methods and techniques in data-collection during a qualitative research. We mentioned earlier that qualitative research is eclectic. That is, the choice of techniques is dependent on the needs of the research. Although this should be true for almost all social research, it is particularly so with qualitative research in that the appropriate method or techniques is often identified and adopted during the research. Qualitative research is also multi-modal. The researcher may adopt a variety of research techniques, or a combination of such, as long as they are justified by the needs. The discussion below is therefore not to identify a set of techniques unique to qualitative research, but rather, to introduce the methods and techniques most commonly used in qualitative research, and the issues related to such use.
We shall introduce the methods and techniques in three broad categories: observations, interviews and study of documents. These are also the basic methods used in cultural anthropology (Bernard, 1988:62). Indeed, the discussions about qualitative research in education can be viewed as a particular case in cultural anthropology.
Observation usually means the researcher’s act to find out what people do (Bernard, 1988:62). It is different from other methods in that data occur not necessarily in response to the researcher’s stimulus.
Observation may be obtrusive or unobtrusive. A researcher may simply sit in the corner of a school playground and observe how students behave during breaks. He may also stand by the school gate and observe how students behave at the school gate. Such cases of observation may be seen as unobtrusive. In other cases, the researchers may not apply any stimuli, but their presence per se may have some influence on the scene. The most common example in this category is classroom observation. Although the researcher may just sit quietly at the corner of a classroom, the presence of the researcher may change the classroom climate. It is, nonetheless, still observation.
Observation is a basic technique used in almost all qualitative research. Even if other methods or techniques are used, the researcher remains the most essential “sensor” or “instrument” and hence observation always counts (McCracken, 1988:18-20). For example, when interviewing is used, a qualitative researcher also takes into account the tonic or facial expressions of the informant, because they help interpret the verbal responses. Such expressions are only sensed by observation.
If the interview is done in the field, then the surroundings of the interview site also provide meaningful data for the research. The surroundings can only be depicted through observation. Hence observation is indispensable in almost all occasions of qualitative research. However, the term observation may sometimes go beyond what is seen. It also pertains to what is heard, and even sometimes what is smelled. Case 4.1 provides one of such examples.
Case 4.1: Classroom Observation Scheme
In the IIEP project on basic education, Leung designed for the Chinese research a scheme for classroom observation. Classroom was taken as one of the environmental factors affecting students’ learning. The scheme was designed after Leung stayed in local schools for two days. The scheme did not confine itself to the performance of the teacher, although that was a part. The figure on the next page shows one of the six sections of the scheme.
Different writers have different ways of classifying observations. Without running into juggling of definitions, we shall briefly introduce observations as participant observations and non-participant observations. More detailed classification of observations can be found in Bernard (1988), Goetz and LeCompte (1984) and Patton (1990).
Participant observation is perhaps the most typical of qualitative research. Some authors even use participant observation as a synonym for ethnographic research. Different writers may have slightly different definitions of participant observation. The following description by Fetterman is perhaps the most agreeable to most researchers.
Participant observation is immersion in a culture. Ideally, the ethnographer lives and works in the community for six months to a year or more, learning the language and seeing patterns of behaviour over time. Long-term residence helps the researcher internalize the basic beliefs, fears, hopes and expectations of the people under study. (1989:45)
Immersion of the participant can either be continuous or noncontinuous. The three classical cases we quoted in Chapter 1 all include participation in the continuous mode. Li’s study of classroom sociology (Cases 3.8 and 3.9) involved one year’s continuous residence. In the second and third year she went to the school three days a week. She combined continuous with noncontinuous participant observations. Fetterman used noncontinuous participation when he was doing qualitative evaluation of educational programmes.
Case 4.2: Noncontinuous Visits
In two ethnographic studies, of dropouts and of gifted children, Fetterman visited the programmes for only a few weeks every couple of months over a three-year period. The visits were intensive. They included classroom observation, informal interviews, occasional substitute teaching,interaction with community members, and the use of various other research techniques, including long-distance phone-calls, dinner with students’ families, and time spent hanging out in the hallways and parking lot with students cutting classes. (Fetterman, 1989:46-7) II. Environment of the classroom
1. The classroom is on the _____ floor of the school building.
2. The classroom is near
( ) residential area ( ) factories
( ) road(s) ( ) field
( ) marketplace
( ) others _______________________________________
3. The number of windows which provide lighting and ventilation to the classroom: ( ) satisfies the required standard
( ) is below the required standard
4. The main artificial lighting facility in the classroom is: ( ) florescent tubes total no.__________________
( ) light bulbs total no.__________________
5. Condition of lighting during the lesson :
( ) bright ( ) dim ( ) dark
6. Ventilation in the classroom:
( ) well ventilated ( ) stuffy ( ) suffocating
7. Quality of air in the classroom:
( ) refreshing ( ) a bit smelly ( ) stingy
8. Environments for listening:
( ) very quiet ( ) occasional noise ( ) noisy
9. Classroom’s floor structure:
( ) concrete ( ) log ( ) mud ( ) carpet
10. Classroom’s floor condition:
( ) clean ( ) some litter ( ) full of rubbish
11. Classroom’s wall conditions:
( ) smooth & clean ( ) some stains ( ) dirty & damaged
12. Classroom’s area: _____________m2; area/person: _____ m2.
13. Space use in classroom:
( ) looks spatial ( ) fairly crowded ( ) very crowded
14. Furniture and other article arrangements in the classroom: ( ) orderly and tidy ( ) messy
1Figure 1 Classroom Observation Scheme (Designed by Leung Yat-ming) Whyte’s experience in the Italian slum (Case 2) is perhaps the nearest to ideal in participant observation. He stayed in the community for two years. He experienced the life of a member of the Italian slum. In Whyte’s case, native membership allows the researcher the highest level of participant observation.
Most researchers are denied such an opportunity, often because of constraints in time and resources, as we have discussed at length in Chapter 3. Under all sorts of constraints, at best the researcher “lives as much as possible with and in the same manner as the individuals under investigation” (Goetz and LeCompte, 1984: 109). In these circumstances, the researchers may not claim that they was doing ethnography, but it is legitimate to apply ethnographic approach and techniques to the study (Fetterman, 1989:47). Participant observation in its broad sense therefore tolerates different lengths of time and different degrees of depth. There is a full range of possible modes of participant observation, what Wolcott calls “ethnographer sans ethnography” (Wolcott, 1984: 177).
The most frequent case in education is that a researcher may stay in a school and become a teacher in that school. The researcher identity may or may not be disguised. The researcher may then, as a participant, observe teachers’ behaviours in teaching, in meetings, in conversations, and so forth.
Sometimes, the researcher is readily a member of the community (say, a school) and may still carry out research as a participant observer. However, in this case, the researcher should be aware of his/her knowledge of the community and should be cautious that such knowledge would not lead to preoccupations about the school under research. In cases where the researchers have successfully gained membership (as Whyte did in the Italian slum), the distinction between a native member and the researcher-as-participant begins to blur. This insider-outsider dialectics will be further discussed later.
Strictly speaking, nonparticipant observation involves merely watching what is happening and recording events on the spot. In the qualitative orientation, because of the non-intervention principle, strict nonparticipant observation should involve no interaction between the observer and the observed. Goetz and LeCompte assert that in the strict sense “nonparticipant observation exists only where interactions are viewed through hidden camera and recorder or through one-way mirror” (1984: 143).
Dabbs (1982:41), for example, used hidden camera in Atlanta at a plaza in Georgia State University, and studied an informal group that frequently gathered during the morning break. There are examples of using hidden video-cameras in school toilets to study drug problem among students, or to use unnoticed audio recording device to study student interactions. The use of audio or video recording device often invites concern in ethnical considerations. Such problems are similar to those arising in using one-way mirrors in interviews or psychological experiments. Such cases are rare in policy-related research.
Another case of nonparticipant observation with ethical problem is disguised observation, or covert observation. A typical example is Humphrey’s (1975) study on homosexual activities. He did not participate in such activities, but offered to act as “watch queen”, warning his informants when someone approached the toilet. Another famous example is Van Maanen’s covert study of police. He became practically a police recruit. Over more than a decade, he “slipped in and out” of the police in various research roles (Van Maanen, 1982). Covert observations are again rare in research which is related to educational decision-making.
Hidden camera or recorder and covert observation occur only exceptionally. Most author would accept the watching of audience behaviour during a basketball game (Fetterman, 1989:47) or the watching of pedestrian behaviour over a street as acceptable examples of nonparticipant observations. Interaction between the researcher and the social community under study is often unavoidable. We have again discussed this at length in Chapter 3 under the notion of researcher intervention. If we perceive the problem of intervention as a matter of degrees, then the distinction between participant observation and nonparticipant observation begins to blur. The general principle across the board is that the researchers should minimize their interactions with the informants and focus attention unobtrusively on the stream of events (Goetz and LeCompte, 1984:143).
Wolcott’s study of school principal (Case 3) was perhaps the most intensive type of nonparticipant observation that one could find in the realm of education. (He also used other supplementary methods as mentioned in Case 3). He did live with the school for two years, but he did not participate as a school principal which was his subject of study. He saw his role as one of “participant-as-observer” (Wolcott, 1984:7). So was Li’s study (Case 3.8) of classroom sociology in her first year.
She did stay with the school as a teacher but she never became a student which was her subject of study. The following two years of her study, however, was not nonparticipant observation because she applied experimental measures. During the UNICEF research in Liaoning, the basic method I used was interviewing and not nonparticipant observation, but I did have, at times, nonparticipant observation when debates occurred between the local planners and the provincial planners (Case 3.7), or when planners chat among themselves about their past experience in the field.
The most frequently employed nonparticipant observation which is relevant to educational decision-making is perhaps observation at meetings. Typically, the researcher attends a meeting as an observer. The researcher tries to be as unobtrusive as possible and records everything that happens during the meeting. When Wolcott did his study on the school principal, he was present at all meetings unless he was told otherwise (Wolcott, 1984:4). The following was my experience of a non-participant observation in China.
Case 4.3: A Validation Seminar
I realized during the UNICEF research in Liaoning (Case 4) that one essential step in the planning for basic education in China was validation. When drafting of an education plan was complete, the draft plan had to undergo scrutiny in what is known as a validation seminar. In essence, all those related to the plan, including leaders at all levels, representatives of all relevant government departments, experts from all areas – are invited to discuss. Relevant documents are sent to the participants well in advance. They are then asked to comment on the plan during the validation exercise. Only “validated” plans are submitted to relevant machinery for legislation. The validation seminar for Liaoning was unfortunately held before the UNICEF research. I got an opportunity, however, a year after in 1988, when the Shanghai educational plan was to undergo validation.
The host of the meeting agreed to send me an invitation. I attended the meeting in the name of an “external expert”, although I made clear to the host that my major task was not to contribute. They agreed. During the meeting, I was able to observe the roles of the various “actors” during the meeting. I was also able to talk to individual participants during tea breaks and meals to understand their background and their general views about educational planning. I was able to do a number of things over the two-day meeting: (a) to classify the over 40 participants into technocrats, bureaucrats, policy-makers and academics; (b) to understand the different extents in which the participants contributed to the modification of the plan; (c) the disparity in capacity among participants in terms of information and expertise; (d) the inter-relations between the different categories of actors and (e) the function of the validation exercise. In the end, I concluded that validation was a way of legitimation, which employed both technical (expert judgement) and political (participation) means to increase the acceptability of the plan before it went for legal endorsement. The political aspect came to me as a surprise. It indicated a change in the notion of rationality among Chinese planners and policy-makers.
Interviewing is widely used in qualitative research. Compared with observation, it is more economical in time, but may achieve less in understanding the culture. The economy in time, however, makes ethnographic interviewing almost the most widely used technique in policy-related research.
Interviewing is trying to understand what people think through their speech. There are different types of interviews, often classified by the degrees of control over the interview. Along this line, we shall briefly introduce three types of interviewing: informal interviewing, unstructured interviewing, semi-structured interviewing, and formally structured interviewing. We shall also briefly introduce key-informant interviewing and focus groups which are specific types of ethnographic interviewing.
Qualitative research of course has no monopoly over interviewing. Interviewing is also frequently used in research of other traditions. The difference between ethnographic interviewing and interviewing in other traditions lies mainly in two areas: the interviewer-interviewee relationship and the aims of interviews. Ethnographic interviewees, or informants, are teachers rather than subjects to the researcher, they are leaders rather than followers in the interview. The major aim of the interview should not be seeking responses to specific questions, but initiating the informant to unfold data.
Readers may find more detailed discussions about ethnographic interviewing in Spradley (1979) who provides perhaps the most insightful account of the subject. In-depth discussions about ethnographic interviewing can also be found in Bernard (1988), Patton (1990), Fetterman (1989) and Powney and Watts (1987).
Informal interviewing entails no control. It is usually conversations that the researcher recall after staying in the field. It is different from “observation” in that it is interactive. That is, the informant speaks to the researcher. By its own nature, informal interviewing is the most “ethnographic” in the sense that it is not responding to any formal question. It is part of the self-unfolding process.