WHERE WE'VE BEEN

We've described how to operationally define and measure variables in such a way that our measurement processes are as reliable and valid as possible. We've also discussed ways to present and interpret the scores that result from our measurement processes and the strategies we use for designing instruments that assist us in the data collection process.

WHERE WE'RE GOING NOW

In this chapter we'll describe how to collect reliable and valid data in naturalistic settings-such as schools, classrooms, and communities. This kind of information is often informal, interpretive, and of direct value to educators. In addition, it often supplements the kind of data described in preceding chapters and provides input and a context for cause-and-effect research, which will be covered in later chapters.

CHAPTER PREVIEW

As we stated in chapter 1, the purpose of Level I research is to ascertain what outcomes are occurring in an educational setting. When teachers or researchers already know what variables to observe and can easily operationally define them, it is often a straightforward (though sometimes difficult) process to apply the strategies described in chapters 4-5 to assess the degree to which these variables are occurring.

In other cases, however, while it is obvious that there are problems to solve, it is necessary to conduct research to clarify what the problems really are and to identify variables that may need to be measured by more structured methods. In addition, sometimes the variables are unknown or genuinely difficult to operationally define, so it may he necessary to conduct interpretive research to explore and develop variables and to give deeper, fuller meaning to them. The labels qualitative, interpretive, and naturalistic have been applied to these kinds of broad-based research efforts to describe or interpret educational settings.

After reading this chapter, you should he able to

1. Define and give examples of qualitative research.
   
   
2. Describe the relationship between qualitative research in education and research methodologies in ethnography, human ecology, and sociology.
   
   
3. Describe the role of the participant observer in qualitative research.
   
   
4. Describe effective qualitative research strategies, including interviews, observations, and content analysis.
   
   
5. Describe how samples are selected for qualitative research.
   
   
6. Describe the four types of field data.
   
   
7. Describe the important issues in establishing the reliability and validity of qualitative research.
   
   
8. Describe four major phases of qualitative research and the activities typically conducted during each phase.
   
   
9. Describe the characteristics of a good qualitative research report.
   
   
10. Describe the major problems that occur in qualitative research and provide strategies for overcoming these problems.
    
    

The strategies described in the preceding chapters are quantitative in the sense that researchers using them will count or rate behaviors or in some other way assign scores as a result of a data collection process. On the other hand, researchers using the strategies in the present chapter will collect data not to assign scores but rather to develop well-founded, deep, general interpretations of a situation, Quantitative research strives to be objective, but the human nature and the imprecision of measurement processes in education guarantee that research in education can never be as objective as that in physics and biology. Qualitative research, on the other hand, admits that the analysis of human activities is by necessity largely subjective and attempts to make the subjective analysis of human behavior as unbiased as possible. Specific strategies within this category of research often include the following:

• The use of field studies
  
  
• The use of participant observers
  
  
• A major concern with the personal feelings and thoughts of an individual or the members of a group with respect to events in their lives
  
  
• The use of broad descriptions of the context of social communities
  
  
• An awareness of the invisibility of everyday social and personal events and structures
  
  
• An awareness of personal, local, and different meanings of seemingly the same structures events, and social rules
  
  
• The questioning of assumptions and existing social systems
  
  
• The holistic, broad, deep perception of existing social systems and the people in them

Qualitative research is not new. In some areas of the social sciences, such as anthropology, qualitative research has long been the primary strategy for developing and testing hypotheses. Qualitative research has likewise had a long history in education, dating back at least to Barker's (1964) Large Schools, Small Schools, which presented overwhelming evidence of the benefits of small schools for the personal and social growth of adolescents. The results of that study were all but ignored by educational administrators, who were on a campaign to consolidate schools, partially on the promise that larger schools would be more efficient to run - a promise that was little supported by research and was seldom fulfilled. Coleman (1961) studied the culture of adolescents in high schools in the Chicago area and concluded, among other things, that adolescents do form a society and that a psychology of learning based on students as isolated entities would he markedly incomplete. In such ways, qualitative research has played a prominent role by contributing to the understanding of the educational process and its context.

In another sense, however, qualitative research in education is a new field. There has been a tendency to consider as "true research" only the quantitative and experimental research described in the other chapters of this book. Recent theorists have balked at this perception and have pointed out that since human behavior is complex and often highly subjective interpretive research strategies are necessary to supplement traditional, quantitative educational research (Erickson, 1986). Quantitative and qualitative researchers have simply responded in different ways to the challenge posed by the fact that experimental control, in the tradition of physics and biology is clearly impossible in education and the other social sciences. Quantitative researchers have responded by developing or applying numerical data collection processes, research designs, and statistical procedures that enable research and measurement in social sciences to parallel closely the work of natural science researchers. Qualitative researchers, on the other hand, have responded by using strategies focusing on the "objective analysis of subjective meaning" (Erickson, 1986). There is considerable validity in both of these approaches.

In this book, we take the position that there is an objectivity and subjectivity continuum and that it is improper to force all research studies into one category or another. In fact, the two types of research are perhaps of greatest value when combined. An appropriate strategy is to understand the basic principles of both qualitative and quantitative research, to know the strengths and limitations of each, and to be able to interpret the results of both methods.

The newness of some aspects of the emphasis on qualitative research presents certain problems in discussing it. The terms, labels, and methodologies are sometimes not well defined. One of our goals is to clarify some of these ambiguities in later sections of this chapter.

THE ORIGINS AND IMPACT OF QUALITATIVE RESEARCH

When they conduct qualitative studies, educational researchers often borrow from such scientific fields as social anthropology, ethnography, ecology, clinical psychology, sociology, social psychology, child development, and family studies. When you read qualitative research reports, you will sometimes see them referred to as ethnographic or ecological studies. What this means is that the researchers borrowed heavily from the field of ethnography in conducting their research or relied heavily on the methodologies employed in the field of human ecology.

Students and researchers specializing in the qualitative sciences like anthropology and ethnology take one course or more on the research methodology unique to that field. It is safe to assume, therefore, that we are not likely to convey to you within a few pages "how to become an excellent researcher" in any of these qualitative fields. Rather, the following pages will offer examples and guidelines from a few of these fields so that you can understand what the researchers do and can interpret their findings more fruitfully. In our brief examination of these fields, we shall focus immediately on how they can be applied to educational settings.

Ethnography

Ethnography is a part of cultural anthropology. Ethnographers view human actions in broad social contexts and look for multiple meanings about these actions. They study such topics as languages, loss of cultural elements, the influence of exceptional individuals, taboos, religion, child-rearing patterns, marriage, eating habits, housing patterns, gender roles, and games and recreation. By vividly bringing to our attention the extraordinary variability of cultural behavior, ethnography helps make us far more conscious of our own behavioral patterns. It is a science aimed at understanding the diversity of human cultures.

 

BOX 9.1 Good and Bad Reasons for Doing Qualitative Research As university faculty members, we often encounter students and colleagues who tell us they "want to do qualitative research." What they mean is that they become nervous when they hear the words measurement and statistics and they think they have found a way around these distasteful topics. (Sometimes it's even worse: they are simply inept thinkers who want to avoid hard work, and they think qualitative research is easy.) They think they have found a way to do research without doing statistics and without the careful thought and time-consuming planning that must go into effective data collection processes or good research designs. This line of reasoning is seriously wrong! Quantitative research is not really all that difficult to conduct. It is not necessarily highly mathematical. It requires logic and planning - but so does qualitative research. Good qualitative research is no easier to conduct than good quantitative research. Teachers do not conduct qualitative research simply by visiting an innovative school, conversing with some students and teachers, and refraining from using structured measurement processes. They do not conduct qualitative research by simply taking a sabbatical, visiting the school for a whole year, reading many books related to the strategies employed in the school, theorizing carefully about the philosophy behind the school, and writing a 400-page book about their experiences. To conduct qualitative research, they would have to apply strategies like those described in this chapter. The right reason to do qualitative research is because the situation calls for qualitative research.

The ethnographer's task is to describe, interpret, and understand in considerable depth people and the cultural scenes in which they exist. Ethnographers do not make value judgments or evaluate people's actions; they would be concerned if researchers tried to "improve" the people or the cultural situations with which these participants are involved. The ethnographer's orientation is to discover how things are and how they got that way, rather than how they ought to be. The interest is long term - they examine cultural scenes for years or decades rather than for days or weeks. In educational settings, ethnographers believe it is critical to focus on environments such as classrooms, school buildings, school systems, and central offices as cultural scenes.

Ethnographies are analytic descriptions of intact cultural scenes. These descriptions present the shared beliefs, practices, artifacts, folk knowledge, and behaviors of peoples. Ethnographies are both empirical and naturalistic. They are first-hand, holistic descriptions of a total phenomenon in its context. From these in-depth descriptions, ethnographers generate the major variables and phenomena affecting the participants' beliefs and behavior. The findings that are generated often give rise to or support emerging theories.

Human Ecology

Human ecology is the study of the total system of influences surrounding people. As teachers well know, the classroom is a complex environment of cultures and overt and covert message exchanges expressed in a number of ways. Variables in an ecological study of a classroom could include the written and spoken language; the body language of movement, gestures, and posture; changes of tone, speed, tempo, pitch, and intensity of voice, including pauses and any intruding sounds; the social processes of negotiation and turn-taking; humor in communications; and patterns of thought, feelings, behavior, and inner responses. All of these variables influence what is communicated, how it is communicated, and from and to whom the communications are sent in the classroom.

In a classroom ecology, the communicators include the students and teachers, any announcements over the intercom, and any messengers, parents, visitors, administrators, or special teachers who come into the classroom. An examination of these communicators and their activities can help develop a rich, descriptive understanding of the social and cultural foundations of a classroom. Thus, they are important sources of information in educational research. Further, classroom teachers are continual, active participant observers of many classroom variables and thus can be full partners in ecological qualitative research. Classrooms (and school buildings and school systems) are culturally complex entities in which students, teachers, administrators, secretaries, janitors, parents, and the community play roles. All of these groups and individuals interact with one another. They all have their own contexts, cultures, patterns, and methods of relating. In addition, schools exhibit both formal and informal curricula. Finally, the classroom ecology includes various patterns of formal and informal social control.

Sociology

A part of education's task is to aid in the socialization and acculturation of children and adolescents. How this is done, whose culture is passed on, and how it is communicated in schools should be a major part of educational research. Sociology looks at the functioning of groups of people and the influences of social structure. It is concerned with such factors as social change, social status and stratification, social power, social problems (such as crime, poverty, ethnic tension, and overpopulation), bureaucracies, families, and social institutions. In addition, sociologists are concerned with interpersonal relationships, conflicts, and management styles. They also study evasion tactics such as ambiguity, ignorance, avoidance, and the use of intermediaries, rituals, secrecy, lying, and joking in social relationships. Sociologists study education and view it as a major cultural learning experience for everyone in our society.

BOX 9.2 Margaret Mead: Qualitative Researcher Anthropologists have pioneered the social research methodology of the active participant observer. When anthropologist Margaret Mead lived with the Samoans in the South Pacific, she was not too much older than the adolescent Samoan girls; she was thus able to gain the girls' confidence and that of their parents and other members of the community. She described in detail the cultural aspects and events in all of their lives and the physical, biological, and social environments in which they lived. Her book Coming of Age in Samoa (Mead, 1928) not only described the lives of adolescent Samoan girls but gave Americans considerable new insight about their own culture and about how adolescent girls are treated in the United States. Young women have been reading the book ever since, to gain deeper insights and broader perspectives about their own lives, feelings, and behaviors.

One of the problems of educators is to deal effectively with all the social classes. Perhaps some educators "tolerate" lower-class habits and values and reinforce "good" middle-class values, while trying to convince the "elites" of society to financially support the schools. Family poverty is another problem for which schools make accommodations, by providing free lunches, special tutoring, and often public health services. An additional problem is that teachers are predominantly from the middle classes and tend to emphasize the middle-class values of attendance at school, punctuality, respect for property, and personal responsibility, while school boards may be drawn from business and professional groups or, in rural areas, from landowning farmers.

The preceding paragraphs have shown the degree of overlap between sociology and education. In fact, there is a well-developed field of the sociology of education, which itself conducts educational research. Educational sociologists point out that schools are a part of society and not apart from it; thus, schools cannot be understood without some understanding of society, which schools reflect. They tell us that schools seek out students and indoctrinate them with specific knowledge and social values. In fact, one of the major goals of teaching is to change the beliefs of the students. It is clear that the content and methodologies of sociology have much to offer education.

Summary: The Impact of Qualitative Sciences

We have summarized only a few of the qualitative sciences whose research methodologies are relevant to education. As you read these descriptions, it probably became obvious to you that there is considerable overlap among the activities that researchers in these various fields perform. For example, ethnographers and human ecologists often focus on the same variables as do sociologists, but from a slightly different perspective. The recommendation of qualitative theorists in education is that educators similarly borrow from these fields and apply these methodologies in proper perspective to education.

To exhibit a broad understanding of students, teachers, administrators, and parents in educational research it is important to be knowledgeable of the theories and research methodologies of such fields as anthropology, ethnography, human ecology, and sociology, as well as their data collection, analysis, and interpretation methods. Most readers of this book will realize that they have at least briefly encountered the methodologies of these fields as part of their liberal arts training to become teachers. As qualitative research reports continue to be increasingly available in the professional literature, you will have opportunities to review these methodologies, understand their applicability to education, interpret their findings more appropriately, and perhaps begin to employ them yourself in further understanding educational problems.

QUALITATIVE DATA COLLECTION METHODS

Once we have realized the value of qualitative research, it is useful to examine specific methods for recording a continuing sequence of events and a wide range of variables. Effective qualitative methods enable the researcher to probe situations in considerable depth and breadth, learn the participants' personal feelings and views of activities, determine the social structure and context of the classroom or other educational setting, and put all the observations into a holistic, all-encompassing, phenomenological picture.

The Participant Observer

The key data collector in many forms of qualitative research is the participant observer. As the name indicates, this is a person who participates in the setting or process being studied and also makes careful observations of what is happening. Participant observers (who in classrooms can certainly be teachers) are not strangers to the situation; therefore they know the language, phrases, and particular vocabulary common in educational situations. This knowledge allows the participant observer an understanding of the events in the classroom and educational culture from the beginning phases of the data collection. The participant observer can ask sensible questions about educational events and can develop a strategy for data collection of all kinds: observations, interviews, questionnaires, visual and auditory recordings, content analysis, and formal and informal testing.

It takes practice to be a good participant observer. Explicit awareness must be built. A good way to develop this expertise is to work with a colleague - each of you should write detailed, explicit descriptions of rather ordinary educational processes which you observe, or conduct separate interviews of the same people. Then compare notes with your colleague to see how they agree and to see which variables were reported by one but not the other. Discuss the reports of your observations and interviews to determine what you may have misinterpreted, overinterpreted, or omitted. Then repeat these activities several times with different observers and interviewers with progressively more complex social events and interviews. Continue this process until you see and hear variables in much the same way.

The process described in the preceding paragraph will produce observers and interviewers who are in general agreement with one another (in psychometric terms, these researchers develop reliability). Through this process the observers and interviewers can demonstrate that there are real social and psychological (but subjective) variables in the educational world that can be observed by others. Being reliable data collectors doesn't necessarily indicate that the participant observers are valid reporters and interpreters of that reality, but without replicable data there is little hope of having valid data.

Bernard (1988, p. 159) suggests that even with prior training and practice and good rapport, "it takes at least three months to achieve reasonable intellectual competence in another culture and be accepted as a participant observer." He suggests that researchers pick sites that are easy to enter, have their research project well documented, use contacts with the participants who can help them gain access to the culture, and think through in advance their responses to questions about what they are doing, what they want to learn, what good their research is, and whom it will benefit.

Classroom teachers can be heavily involved and often full partners in qualitative research, certainly at Level I research. They can be collaborators and participant observers with others at Levels II, III, and IV. Teachers are deeply and personally involved in the classrooms and with their students. They are only too aware of discrepancies between learned theories of human behavior and actions and what they see everyday. Further, they see unusual students and social events for which there are few descriptions in the published literature. They know through personal experience that family, neighborhood, and community contexts are important. Every classroom has a distinctly local, unique, and compelling reality.

A teacher serving as a participant observer can markedly reduce the problem of reactivity of the members of the class or school being observed. Teachers understand educational systems and how they operate. They fit in well, and the class and school participants soon go on with the business of teaching and learning and let the participant observer conduct interviews, make observations, and generally collect data with a minimum of change in behavior. Teachers acting as participant observers can facilitate collecting field data on-site in the classroom or other educational setting in which they participate.

One of the problems of being a teacher participant observer is that teachers and other educators are not naive about the culture they are observing, and familiarity with the classroom can serve as a biasing factor that will interfere with objective data collection. Much of what teachers do every day is done so automatically that perhaps they are intellectually almost unaware of it, or they read their own motivations and thoughts into what others are doing as they go about their tasks. In either event, they can form an inaccurate interpretation of observed events or, perhaps worse, not recognize that an event or a response is taking place.

In spite of the problems, however, the participation of teachers in qualitative research, collaboratively or individually, holds the promise of contributing markedly to the profession of teaching, to education, and to one's personal growth and insight.

TYPES OF QUALITATIVE DATA

Generally, qualitative research data fall into four classifications: observations, interviews, documents, and research instruments of various kinds (such as questionnaires, surveys, and personality, attitude, and cognitive tests). Many of these qualitative data collection methods are available to teachers, counselors, and administrators, who should use them in understanding their educational milieu. However, in some cases, qualitative researchers remain deliberately uninformed about variables and relationships until after their data collection; this enables them to approach the situation without preconceptions that may bias their research. In other cases, the participant observers approach the research setting with a more specific plan and listed variables, behaviors, events, and settings of probable interest to them. This preplanning is based on prior experiences and reviews of the literature and gives them a tentative organization for their data collection in the field.

Seidman (1991) points out that effective qualitative interviewing and observation must take into consideration the power of the social and organizational context of people's experience. He urges researchers to become aware of their own experience with issues of race, class, gender, and age, as well as the way these factors may be influencing the participants.

The information obtained by these less formal means of data collection needs to be organized and interpreted (the characteristics of a good report are summarized later in this chapter). The results of qualitative studies often provide a useful basis for developing other data collection strategies, including further qualitative research, structured interviews, questionnaires, and surveys.

Observations

Some observational methods can be (and in some instances should be) conducted under rigorously controlled conditions - highly trained researchers can observe specific, operationally defined units of behavior that are easily agreed upon and counted, as was discussed in chapter 6. However, in this chapter we shall consider open-ended observations, which are broadly defined as observations of behaviors under ongoing, natural conditions. These observations involve seeing and hearing events that people take part in, not asking after the fact what people have done or why. For example, researchers can observe types of teacher-pupil exchange in a naturalistic situation to look for patterns based on ethnic, social status, or gender differences or based on the type of program in which students are participating. In addition, it is possible to observe administrative as well as instructional activities. For example, continuous observations of the principal can help determine the kinds of problems encountered during a typical day and how these tend to be handled - essentially, a day, a week, a month, or a year in the life of a principal.

The obtrusiveness or unobtrusiveness of these observations, as discussed in chapter 6, must always be a consideration. Here, the emphasis of qualitative research leans heavily toward observations and measures that are as unobtrusive as possible. Reactivity of the participants being observed or interviewed is a major problem for qualitative researchers. They want to see behaviors essentially unmodified by their own presence. Mixing in (or "hanging out") with those observed, if this is possible, is more friendly, builds rapport, and is certainly less reactive than more systematic time sampling, behavior counting, or videotaping of social groups. However, the more informal approach requires much more reliance on the observer's memory and increases the possibility of bias. If such biases can be minimized, the quality of the natural data is likely to be better and more meaningful.

Interviews

Interviews are another major form of data collection in qualitative research. The richness of the responses in both breadth and depth can add markedly to the understanding of the classroom or school. In qualitative data collection the informal interview essentially has no structure. These interviews resemble ordinary conversations, except that the participant observer often makes a point of ensuring that the conversations take place and continue. Notes are afterwards made of the informal interview's contents based on the memory of the conversation. The informal interview is used to explore interesting phenomena in the cultural scene and to establish rapport. At a slightly higher level of structure, there can still be a minimum of control, but there is a clear plan in the interviewer's head about the areas and people of which information is wanted. The purpose of the informal interview is to allow a structure so that the persons being interviewed tell the interviewer their information in their own terms. Tape recordings of interviews are often helpful to preserve and cross-check information.

In qualitative research, the interviewer is often shopping for information but is not totally certain regarding what that information may be; he or she doesn't know the aspects and topics about the social group that the informant knows. It is wise to guarantee the informants' anonymity and to explain to them why they have been chosen (e.g., because they have observations and feelings about particular situations, events, or social relationships; or they were picked at random to obtain a sample that represents a group of people).

The informant must have the needed information and must be available to give the interviewer the time to talk. The interview tone should be cordial, supportive, and nonthreatening. The general principle of many interview formats is to tell the informant, in rather general terms, what it is the interviewer wants to know, and then let the informant talk.

Counselors and teachers with counseling courses should be especially good interviewers. They will recognize the general approach of the qualitative interview as nondirective and Rogerian in style. Teachers and counselors typically have the genuine interest in people that is so necessary to be a good interviewer. They are often able to develop an open, reflective communication pattern with the informant and build a feeling of trust and rapport. Being supportive of responses with smiles, nods, and a nondirective "uh-huh" is important. Showing attentive silence and learning to wait are both valuable interviewing skills. Another effective skill is to be able to reflect the gist of the last statements in a questioning manner, indicating that you want either affirmation that your notes are right or further clarification of the responses.

The purpose of naturalistic interviewing strategies is to keep the informants talking and to express your interest in what they have to say. Interviewers should not indicate an approving or disapproving valuation by their words, tone of voice, or gestures. It is important to probe to obtain further, more detailed, and relevant information, but if it is necessary to interrupt, this should be done gracefully and not accidentally. The interviewers must reduce their reactions and interventions to a minimum. Reinterviewing the informant a second and third time after a lapse will sometimes allow even more information to be obtained.

Often, it is useful to take on-the-spot notes or use a tape recorder and have a list of topics available that should be covered. If there are sensitive issues to be discussed, it is better to ask the informant about them in the middle of the interview and begin and end with questions about less sensitive topics. As part of the interview, it is useful to take notes on the surroundings in which the interview was held and to record the date and time. It is also important to record the informant's attitude (e.g., was the informant open? evasive? cordial? nervous? afraid?).

Documents and Other Artifacts

Formal and standardized instruments (such as questionnaires, surveys, and tests, which are discussed in chapters 4-8) and documents can also be part of qualitative research data. Documents in this sense are often referred to as artifacts. They can include either current information (e.g., instructional materials and communications within the school building or school system) or archival information (e.g., newspapers, school board minutes, teachers' union records of contract negotiations, court records, architects' plans for school buildings, tapes of televised programs, diaries, tax roll data, transcripts, and textbooks). Archival documents are often valuable even though they may have errors and could well be incomplete. The data from both types of documents should be collected unobtrusively, although the records could have been made under obtrusive conditions (e.g., perhaps newspaper reporters asked pointed questions about school board meeting procedures).

One of the main ways to analyze documents and other qualitative data is through a process called content analysis, whereby the researchers look for themes or concepts in the natural language. For example, they could count or analyze the number of references to administrators, teachers, students, and parents to ascertain the relative importance of each of these groups in a school system's communications. In children's themes, the relative emphasis on the concepts of cooperation and competition could be assessed. The several concepts in teachers' evaluation statements about pupils could be coded and counted to determine the curricular and social goals of the teachers. Newspaper articles reporting on the activities in schools could be counted, their length noted, and the topics analyzed; this would indicate the public's interest in education or the facets of the system that the administrators would like the public to know about. Variables can be both conceptually and operationally defined, as well as illustrated with examples from the documents themselves. Current documents can be generated by having students write themes on various topics for content analysis.

SELECTING SUBJECTS AND INTERPRETING QUALITATIVE DATA

The sampling of situations, sites, and informants for qualitative research is usually not done in the same way as for quantitative research. As chapter 8 showed, quantitative researchers often sample people and sites at random. Whatever findings are true in the random samples can be generalized, within limits, to the populations from which they are drawn. In qualitative research, informants and events are selected for their unique ability to explain, understand, and yield information about the meaning of expressive behavior or the way the social system works. This is called theoretical sampling. Purposive samples can also be chosen to yield maximum information related to specific issues.

The samples may be unique or unusual in order to promote insight about a behavior that is not usual. For example, a researcher may be interested in understanding the feelings of a child with a specific disability, the dynamics of a chaotic classroom, the effect of unusual school building architecture (such as an open classroom) on class instruction, or the educational progress of a newly immigrated Vietnamese community. All of these situations are unusual, and subjects to be observed or interviewed may be selected specifically to learn about social and personal variables operating in the educational situation or to gain insight about ways a class or school functions. Informants are chosen because of their key involvement in a social group and their competence to tell the participant observer what they observe, feel, and think; they are chosen also because they are believed to reflect the views of others in the social group.

At other times, convenience samples are selected simply because they are available. Precautions need to be taken to determine whether such samples are reasonably representative of a more general population. This can sometimes be done by checking the sample's demographic variables (e.g., sex, age, social class) against the population of interest. The logic of sampling (discussed in chapter 8) indicates that nonrandom samples are often biased; these biases present a problem that must be considered when interpreting qualitative research.

THE FOUR TYPES OF FIELD DATA

A major problem with qualitative data collection is remembering the data long enough to record it. Bernard (1988) recommends collecting four basic types of field data: field jottings, field notes, a field diary, and a field log. Field jottings (Figure 9.1) are taken whenever the researcher observes or hears something important. They are written on the spot to avoid the problems of forgetfulness and selective memories.

Field jottings serve as a source for field notes (Figure 9.2), which are summaries of field data collected during the day or over another designated period of time. The notes are supplemented by all the other information collected, including recordings, documents, and notes about the overview and understanding of the social scene as the participant observer sees it at that time. These notes are compiled immediately, at least on a day-to-day basis, so that the data are fresh and other activities do not interfere with recollections. Bernard suggests spending an hour or two per day collating and integrating a day's field jottings and writing the field diary. The field notes are a precursor to the reports the participant observers will ultimately compile.

The field diary (Figure 9.3) is a personal chronicle of how the participant observer feels about the social situation that he or she is in; it should also chronicle the relationship of the observer to those being observed. Later this diary will help the participant observer to more fully interpret the field notes and be alert for personal biases. The field log (Figure 9.4), much like a ship's log, relates the chronicle of daily events: how the participant observer planned to spend time, how time was actually spent, who was seen, what their names are, what they talked about, and who else needed to be seen and what needed to be asked. The plan of the day and the actual activities of the participant observer may actually be quite different, but the log helps keep the data collection organized.

Eventually, the sheer bulk of field notes becomes unwieldy, so organizing and summarizing becomes important. An essential step is coding the information to help structure and report the field data. The coding process (Figure 9.5) consists of looking for patterns and organizing information around these patterns. At a minimum, coding should note the informant's name, the names of those observed, the place of the observation, the date and time of the data recording, and the initial indications of the variables involved. It is during this coding process that concepts and insights arising from the data collection begin to emerge. This development of insights about categories, concepts, social structures, and meanings is one of the principal aims of qualitative research. The codings should also summarize the methodological techniques used and describe the events that occurred during the period of data collection.

Analytic notes (Figure 9.6) are essentially the result of the qualitative researcher's conceptualizations of how he or she thinks the culture and social groupings are structured and organized. They result from a careful examination of the field notes, jottings, diaries, and logs. They are inductive. They look for broad, speculative explanations, organizing principles, and concepts that explain and give structure to the many observations and data that are contained in the field notes. Analytic notes will not be many, but they are the key to good results from qualitative research. They are inductive generalizations arising from specifics and operational definitions. They result from "eyeball" analyses and sometimes from discussions with objective, insightful colleagues who have expertise as qualitative researchers but have not been so close to the field setting. This inductive procedure leads to the generation or refinement of theory. This is often referred to as grounded theory because it is generated from the data rather than developed first and then tested through the collection of data.

Qualitative researchers must be open to negative evidence as well as positive evidence about a theory, and they must deal with unusual cases in the proposed explanations of behavior. As they seem to understand and can explain the behaviors, qualitative researchers develop alternative explanations (hypotheses) and check them against the data as well. A valuable aspect of qualitative research is that it allows you to develop variables and an understanding of patterns of behavior in social groups over a large number of naturally occurring events. The field data descriptions are used to help operationally define the variables and describe in rich, natural language how these variables change in relationship to other variables and circumstances. Good field data depend on systematic fieldwork conducted over a long period of time.

The human mind is a fast and effective organizer and recognizes patterns in data, but it is not too effective in storing large amounts of finite hits of data. Thus, qualitative researchers should make copious notes on a timely basis, add to these field notes daily in a reflective and structured mode, and record their feelings about the situation as well. On the basis of these voluminous amounts of data, researchers should then make interpretations of it, impose a meaningful structure, develop variables, and finally present their readers with their understanding of the social entity with which they have en involved. This process of repeatedly looking for patterns of data and variables in field notes and developing constructs that account for these patterns, comparing them with each other, and redefining and reconceptualizing them into more coherent variables and patterns has been called the "constant comparative method" (Goetz & LeCompte, 1984).

Students seemed to start by working in isolation. Even within the small groups, members did not interact with one another except by showing partners what they had done. By the second day, interaction was accelerating within groups. When the teacher suggested looking at the work of other groups, interaction increased greatly. Groups began sending delegates to other groups to find good ideas that they might use. When groups first started sharing ideas, at first the group with the good idea simply did the work lot the group that requested help. The teacher discouraged this. She showed them how to help by giving ideas without doing all the work for the group receiving the assistance. Figure 9.6 Example of Analytic Notes from Field Data

RELIABILITY AND VALIDITY OF QUALITATIVE DATA

The concepts of reliability and validity (discussed in chapter 5) must be reexamined and expanded for qualitative data. While some qualitative theorists (e.g., Wolcott) would maintain that such quantitative terms do nor apply at all in qualitative research, others would maintain that the essential definitions still apply in this chapter. Reliability means that the data collection process is not self-contradictory - that the data collection is both consistent and stable. Validity means that the observations, interviews, or content analysis really contain the information that the researcher thinks they contain. At their most general level of definition, these concepts are obviously important in any form of research; otherwise, why should anyone pay attention to such research? However, because of its more subjective nature, qualitative research introduces some nuances into the use of these terms.

The concept of reliability is sometimes ignored in qualitative research. If an event is rare, how can two independent observers describe it and then later check their level of agreement? Qualitative research is enormously labor intensive. Many groups and individuals are observed and interviewed over long periods of time, up to a year or more. Many sources of data are tapped, and extensive written records are made. Then more writing is done to further elaborate on the descriptions. To have more than one person do all of this is rare. Clearly, to attempt even to duplicate these observations would be enormously time-consuming and expensive. However, for educational research to maintain credibility as a science, the concern with reliability must be addressed.

Kirk and Miller (1986), two anthropologists, refer to the types of reliability relevant to qualitative research as synchronic and diachronic. Synchronic reliability is the similarity of observations made within the same time period (agreement among observers or measures). Diachronic reliability is the stability of an observation over time. The general rule is that diachronic reliability is distinctly limited by the reliability of observations of phenomena at a given point in time. (In addition, the validity of one's interpretations are limited by both types of reliability.)

Aside from the problems of interpretation arising from low reliability, biases of various types are a major threat to the validity of qualitative research. One type of bias concerns the time span or settings sampled, which may not be typical of those to which the results of the data collection process will be generalized. This problem is treated by collecting data over a lengthy time span and by selecting settings judiciously.

Another major bias concerns the subjectivity of observers. They tend to be biased about what they see, and since qualitative research often involves interpretation, further problems occur when they misinterpret what is really happening. Much of what teachers and educators do every day is done so automatically that they are almost unaware of it, or they read their own motivations and thoughts into what others are doing. In either event, they can form wrong interpretations of observed events or not recognize that an event or a response is taking place. In phenomenological terms, they are reporting their own reality rather than the reality they should be observing. The way to deal with this source of bias is to train researchers to be aware of and control their tendencies to be biased. An effective strategy to increase validity is to do a member check - that is, ask participants if they think your interpretation is valid (but do this in a way that avoids obtrusive interference with the natural setting).

Although there are problems with validity, compared with more traditional quantitative research, qualitative methods are more valid in that they really do get at the underlying concepts being observed rather than measure an artificial entity created by a data collection process. To the extent that interpretation is necessary and the interpreter has interpreted correctly, then the subjective qualitative measurement will be superior to a more objective, quantitative assessment of the same outcome or situation. The problem, of course, is that there is no good, objective way to assess the quality of subjective interpretations and insights. We can muster empirical evidence to demonstrate the validity of quantitative instruments, but this is much more difficult to accomplish for qualitative strategies. This does not mean that quantitative strategies are superior. As we have stressed from the beginning of this book, multiple methods of data collection are always preferable to a single method. Teachers and researchers should derive the benefits from both approaches to research problems.

Maxwell (1992) has identified five general types of validity in qualitative research:

• Descriptive validity refers to the overall accuracy of the descriptions making up a study.

   

• Interpretive validity refers to the degree to which the researcher correctly interprets the activities and feelings of the people in the study (i.e., as the participants themselves would interpret them).

   

• Theoretical validity refers to the degree to which the researcher's explanations represent a legitimate application of the concepts or theories that the researcher thinks they represent.

   

• Generalizability refers to the degree to which an account can be extended to situations or populations not directly studied (both within the study and beyond the study).

   

• Evaluative validity refers to the degree to which judgments based on the study are legitimate.

The validity of any type of research is established by logical methods. Some research studies rely heavily on the mathematical, statistical, or research design strategies described in chapters 5, 10, and 11 of this book to support their validity; this is why they are often referred to as quantitative studies. On the other hand, researchers who use the qualitative methods described in this chapter rely on experience, insight, and reasoning to demonstrate the validity of their studies. The preceding presentation of Maxwell's discussion of validity meshes very well with the phases of qualitative research described by Kirk and Miller (1986), which are described next.

THE PHASES OF QUALITATIVE RESEARCH

Kirk and Miller (1986) describe four phases of qualitative research methodology, which serve as a useful synthesis of what has been discussed in this chapter These phases are as follows:

1. Invention or preparation

2. Discovery

3. Interpretation

4. Explanation

 

Invention or Preparation

The first phase is invention or preparation. This phase involves reviewing the literature about a problem; developing the aspects of the context and social group that might be of particular interest; gaining access to and becoming accepted by the social group and developing social networks and contacts within it; having your preconceptions challenged or changed; dealing with practical constraints to data collection as they emerge; and identifying your own biases. Developing questions about interesting aspects of the field situation, locating key informants, and developing research instruments are also part of the preparation for data collection.

Kirk and Miller state that taking notes during the first views of a cultural scene and its social context are crucial. They point out that first observations of social scenes and cultural settings in the field seem to evaporate as soon as the ethnographers get close to them. The same may be true of too much prior knowledge of a social scene from personal experience or a study of the literature. This knowledge can bias researchers and lead them to think that they understand (only too well) certain personal reactions of the individuals involved or to recognize only certain variables. Thus, some purists in qualitative research refuse to bias themselves by reviewing the prior literature and theory about classrooms or school systems. Nevertheless, it is important for researchers to document their first impressions about a social context and scene because such impressions will fade rapidly, and much will be lost.

A feature common to all qualitative research is the need to become acquainted with the participants and to build rapport. At an early stage in the research process, it is useful to learn where the participants usually work and gather informally; to map the physical characteristics of the classroom, building, and neighborhood; to record the demographics of the participants by role; and, in general, to create a preliminary description of the social and physical scenes to be studied. The demographics should include the participant's age, gender, education, socioeconomic status, ethnic identity, and position in the various social organizations. In addition, use-of-time maps and a record of daily, weekly, and monthly episodes are useful. It is also helpful to determine the concerns of the students, teachers, community, and parents - perhaps by collecting anecdotes, stories, and even gossip about how the participants view one another. These data will suggest what is socially and personally important to the participants and how they evaluate their role in the group. At this point, too, the observer develops trust and rapport with the participants. He or she also takes care to note unusual features of the physical and social functioning of the group because all too soon these features will fade from consciousness, seem familiar, and no longer be questioned or perhaps even recorded.

At an early stage it is also important to note the location of documents - records, minutes, curriculum guides, memos, letters, public relations handouts, newspapers, policy statements, etc. - that will form the collection to further inform and substantiate the observer's description of the group. Observations, interviews, document collection and analysis, depiction of physical structures, content analysis, open-ended questions, focus groups, role playing, formal and informal tests, and product analysis can each form a part of the all-encompassing description of the group.

Discovery

The second phase, discovery, occurs when the researchers identify specific places and times to make their observations, conduct interviews, and collect other types of data. An initial plan for the research needs to be developed. In even simple social groups, an almost unlimited number of observations could be made. A research plan lists the most interesting situations, events, people, and data that can be used.

This phase is conducted initially to allocate the researcher's time to accumulate most efficiently data that will shed light on the questions, individuals, and social events of interest. Research in general is a continual, circular process of asking questions and collecting data so that you can ask better questions, collect more data, and so on. One of qualitative research's great strengths is in the broad observational abilities of the researchers to see far more than they expected to see and hear far more than they expected to hear.

The scholarly preparation, planning, and first phases of data collection are essential; but an alert, open mind eager to follow interesting clues, identify and record surprises in the first observations, develop more interesting questions, and follow the more intriguing social events is a great strength of qualitative research. It is practically a certainty that these initial research plans will be reworked, in some cases radically. (In one example, an initial qualitative study of school dropouts and underachievers showed that one aspect of the problem was the lure of major money to be made by adolescent boys becoming homosexual prostitutes [Reiss & Rhodes, 1959]. This was totally unanticipated and was a major diversion of the research effort as first planned.) All new questions, variables, and events must be incorporated into the total scope of the data collection plans on-site in the field. In addition, the data collection plans continue to change throughout the fieldwork as more new opportunities arise. Kirk and Miller (1986, pp. 66-67) state that "qualitative research is defined by the location of hypothesis-testing activity in the discovery, rather than the interpretation, phase." The discovery phase ends with all the field data in hand for the attainment of the research objectives, as modified during the fieldwork. The discovery phase fully ends when the researchers exit from the field site.

Interpretation

Interpretation is the third phase of qualitative research and constitutes an ongoing analysis of the field data and its overall meaning. (Some would argue that this should be done, in part, in the field also, where further inquiries are possible.) Interpretation occurs partly through a reconsideration of the reliability and validity of all the data, considering whether the conceptual interpretations of the variables are reasonably correct and the overall understanding of the social scene, its context, and those involved in it are also reasonably correct.

This phase involves a repeated, careful examination and consideration of all the field data combined with the researcher's insights gained from his or her personal involvement on the scene in the field. Data related to the same event from several sources are brought together, further hypotheses are developed, and old and new hypotheses are further tested against other data and alternative hypotheses until a theory evolves from the field data that appears to be a general explanation. The field data are converted into categories and relationships through a process of multiple readings and sorting multiple sources of records into piles relating to such aspects as themes, concepts, individuals, groups, and scenes. This can be done by making multiple copies of the raw data and cutting them up before sorting. File cards can be used, but computers are commonly used these days. Various software programs are available, and the computer's natural language storage capacity and ability to classify, subclassify, and organize data are nearly ideal for the functions that go into developing grounded theory. Assertions are made, tested, and then retested against many examples in the data. During interpretation, the researcher also looks for discrepant cases to see how they fit the theory or whether other theories and explanations are needed.

Data are pieced together to help determine patterns. The search is made for key links that pull together data related to the same event, person, or scene. The whole process of interpretation and explanation is one of pattern analysis, and the attempt is to bring together in a meaningful fashion as many items of data as possible into a meaningful whole. This can bedone in part by maps, figures, diagrams, and tables. The aim is to persuade the readers of the research that sufficient evidence exists to support the assertions made and that the patterns of evidence do exist and the theory (i.e., the conceptual explanations) is indeed valid. It is not the thick, rich descriptions - quotations, vignettes, frequency counts, etc. - that demonstrate the descriptive validity of the report, but all of these combined with the interpretive perspective make the overall presentation valid.

Explanation

The final of Kirk and Miller's four phases of qualitative research is explanation. In this phase a message is produced to communicate the data, findings, and broad, deep understanding of the social and personal relationships of the field site. The report must present the organized data and the relationships among the variables that the data represent; this is the evidence that will support the ultimate assertions. The links between the concrete data and the abstract concepts must be made clear. Researchers should make explicit their own expertise, possible biases, and interpretive stance. Next, they should consider carefully the audiences who will read the report; these could include fellow researchers, administrators, policymakers, teachers, and members of the context community. For the research audience the adequacy of the research methods and data will be of importance, and the interpretations and particularly the hypotheses generated need to betaken into consideration in future research. For policymakers and administrators it is appropriate to present a wider range of alternative policy options that have been generated by the research and some data about the consequences of each. For this group it is particularly helpful to present the practical constraints, inhibitions, and frustrations of those in the classrooms and school buildings.

For the practitioners the concern is whether the field situation is similar to their own classroom or building and whether there are suggestions, prescriptions, or recipes in the report for what works. Also, researchers should question whether the theories and explanations generated help explain the actions, behaviors, and responses of others and themselves. In part, the value of these explanations lies in the specificity of the qualitative descriptions of the contextual situation and the classroom or building so that the teachers and principals can determine whether there is enough similarity between the research and their own situations to allow them to generalize about their own rooms and buildings.

The last audience includes members of the community, parents, other school staff, and students. These people may want to be more generally informed about their school - what is news and not news, what might be regarded as a positive, negative, or neutral evaluation. It is crucial that for this group (as well as to some extent the professional group) reputations and privacy be protected. Certain material may have to be excluded from the public reports. In any event, for all of the audiences the researchers must be judicious and consider that everything they write may ultimately become public information.

CHARACTERISTICS OF A GOOD QUALITATIVE RESEARCH REPORT

The amount of writing included in field notes, logs, diaries, and reflections - especially for the long periods of time involved - is considerable. By necessity, the reporting of the research is lengthy - a good qualitative thesis may run 500 pages or more, and many qualitative research projects are reported at book length. Further, the writing itself must almost be of novelistic quality to gain and keep the attention of the readers. Some of the greatest insights about human actions, thoughts, and feelings are presented in novels, and at least one psychologist has said that to be a good clinician requires taking a few psychology courses and reading Dostoevsky. With this advice in mind, we have found that English and history teachers are among the best qualitative researchers. Erickson (1986) says that the narrator must convey to the readers the particulars and meaning of the everyday life contained in the social scene; ground the abstract analysis concept results in concrete particulars; provide evidence for the analytical results; and present vignettes and quotations in the rhetoric to convince the readers that what is presented is typical of the social unit and that the results can be generalized to other similar events and social units.

Having stated that qualitative research is often time-consuming and that qualitative reports are often extensive, we should hasten to indicate that it is possible - and often desirable - to conduct and report abbreviated forms of qualitative research. For example, teachers doing any informal observation can benefit from applying aspects of what we have described in this chapter. Following guidelines of qualitative research will help make these observations a better "objective analysis of subjective meaning" and will lead to more insightful interpretations of what has been observed. In addition, when we discuss external validity in chapter 15, it will become evident that the results of higher-level research can be generalized more accurately if researchers employ qualitative methods in collecting, reporting, and interpreting data. As later chapters will show, good educational research is often an integration of effective qualitative and quantitative strategies.

PROBLEMS AND LIMITATIONS OF QUALITATIVE RESEARCH

Erickson (1986, p. 140) lists four types of problems that can occur because of poor procedures in qualitative research:

1. Inadequate amounts of evidence. Assertions in research conclusions may be unwarranted by the data or may not include events and scenes that would have confirmed the statements.

    

2. Inadequate varieties of kinds of evidence. The researcher may have failed to employ triangulation of data by using several sources or field sites, or may have failed to employ the multimethod, multittait, and alternative hypothesis approaches to data collection and analysis.

    

3. Inadequate attention to disconfirming evidence. Erickson strongly recommends deliberate searches for potential disconfirming data.

    

4. Lack of attention to discrepant cases. The researcher may fail to examine on a comparative basis individuals and instances that do not conform to the initial theories and explanations that seem to hold for most individuals and cases. Discrepant case analysis also needs to be done as data collection is in process to help change and sharpen theoretical presuppositions. Discrepant cases and individuals' data analyses help suggest alternative hypotheses that need examination. (Ethnography places in the foreground discrepant cases, which may contain more information than the normative cases.)

Sadler (1981) has indicated areas of pitfalls and cautions for those attempting to observe and make generalizations and inferences in qualitative research (as well as providing commonsense approaches to drawing conclusions from evidence). Sadler lists 10 such problem areas:

1. Data overload. It is well known that humans have severe limitations on the amount of data they can process. While the human brain is unexcelled in its capacity to extract patterns and relationships from noisy environments, it has distinct limits on receiving, recalling, and processing data. Naturalistic observations can severely tax these cognitive processing limits.
   
   
2. First impression. It is also well known that the order in which observations are made is important. First impressions tend to dominate observers' later inferences and assertions. Closely related is the problem of confidence in judgment. Researchers (and also administrators) tend to stick to a decision once it is made, even in light of considerable contrary evidence.
   
   
3. Availability of information. The ease with which an observer can obtain data influences conclusions. Personal relationships can ease or hamper data collection. Data that are readily obtained will be relied on more than data that are difficult to obtain. When initial hypotheses are developed, instances of positive observations that confirm the hypotheses will tend to be unconsciously selected, noticed, depended upon, and later stated as fact. Even when there are later negative or conflicting observations, they tend not to be noticed.
   
   
4. Positive and negative instances. Even balances between negative and positive observations do not cancel out. Sadler notes that to be aware of ambiguity is not the same as to be ignorant.
   
   
5. Redundancy versus novelty of information. Novelties and extreme observations tend to be overweighted and given more importance than redundant or consistent data.
   
   
6. Uneven reliability of information. Lack of notice of uneven reliability of data causes all data to be given equal consideration.
   
   
7. Missing information. Missing data tend to be filled in by the observers later when making interpretations, often in unpredictable ways.
   
   
8. Base-rate proportion. Base-rate proportions against which later comparisons are made are often unstable as the result of observations on one or a few cases. Faulty or unstable base-rate proportions produce inaccurate comparisons and interpretations later.
   
   
9. Sampling considerations. From the study of sampling theory (see chapter 8) it is well known that small samples are unstable and that humans tend to be insensitive to sampling considerations. In qualitative research, where samples are often small (e.g., a class, a student, or a school building), the local sample results may not hold true for the larger population.
   
   
10. Co-occurrences and correlations. There is a problem with interpretation of co-occurrences and correlations. A documented correlation is not necessarily a cause-and-effect relationship; rather, it is merely a co-occurrence of events, perhaps caused by some prior condition. (See chapter 13.) Exacerbating the problem, many observers tend to rely on positive co-occurrence instances as evidence of a strong correlation.

WHAT IS THE ROLE OF QUALITATIVE RESEARCH IN EDUCATION?

At the beginning of this book we argued that teachers should do research every day. This is true of qualitative research as well as quantitative research. While a careful, full-fledged qualitative study consumes a large amount of time and energy, the basic strategies of qualitative research are applicable to the daily life of teachers. For example, teachers can benefit from becoming more skillful at observing their students in their natural environments, from being aware of the wide variety of factors that interact with one another to influence their students, from taking and using field notes, and from generally being more aware of the impact in their classrooms of variables they cannot yet quantify.

Scientific research methodologies have been developed over a span of 400 years to reduce the human errors of observation and logical inference. Research methods in the human sciences are much newer than that, but numerous problems of biased observation and inaccurate inferences and explanation have been noted, and methods have been designed to reduce these errors. These methods and concerns need to be heeded by all in the human sciences. Qualitative, sociological, cultural and social anthropological, and human ecological research methods are valuable in developing the basic understandings of students, teachers, administrators, parents, and the social contexts, scenes, and events in which these people live, study, and work.

Qualitative research reports are valuable as instructional tools. For example, they have the "you are there" quality, which allows education students to obtain the "feel" of classrooms, school buildings, and systems. Teaching itself requires an ethnographic perspective toward the students and a knowledge of the social learning environment and context. The reports of the human sciences of social and cultural anthropology are about as close and involved with the real life of education as one can get out of books. They serve as excellent entrees to the educational world prior to the students actually becoming a part of it. Reading books is not as good as the real thing, but it gives a much wider perspective than can be obtained in a few classrooms; and, of course, it is less disruptive and cheaper.

BOX 9.3 Examples of Qualitative Research Studies Health, S. B. (1982). What no bedtime story means: Narrative skills at home and school. Language and Society, ii, 49-76. This study provides a good qualitative description of how students in homes with various socioeconomic backgrounds differ in the language skills that relate to educational success. Jankowski, M. G. (1991). Islands in the Street: Gangs and American Urban Society. Berkeley: University of California Press. Without personally committing criminal behavior, the author became a participant observer of gangs in Los Angeles, New York, and Boston over a period of ten years. His report gives useful insights into why young people join gangs, what gangs do, and how people react to gangs. McAuliffe, S. (1993). Toward understanding one another: Second graders' use of gendered language and story styles. Reading Teacher, 47, 302-309. The author uses qualitative strategies to determine that boys and girls have distinct styles of communicating during story time and to explore the nature of these different styles. Wong, E. D. (1993). Self-generated analogies as a tool for constructing and evaluating explanation of research in science teaching. Journal of Research in Science Teaching, 30, 367-378. The author uses a systematic anecdotal approach to describe students coming to new insights because of their use of analogies to understand unfamiliar phenomena.

Good educational research is time-consuming and labor intensive. There are many important educational problems that need better solutions than now exist. Therefore, the best-quality, most informative research needs to be done, research that will give us the broadest, deepest understandings that can be obtained. Inevitably, qualitative and interpretive data and methodology will be part of all educational research that provides these roads to understanding.

The present chapter has focused on the contribution of qualitative research toward determining what is going on in educational settings and presenting causal explanations for what is happening. In addition, qualitative methods can be integrated with quantitative methods to contribute to higher levels of research as well. These contributions will be discussed in later chapters.

PUTTING IT ALL TOGETHER

Let us return now to the days when Mr. Anderson was first planning to do research on children's attitudes toward animals. At about that same time, a colleague, Helen Witherspoon, expressed a similar interest in doing research, but she took a different approach.

Mrs. Witherspoon was working on a graduate degree and needed a research project for her thesis. She identified two schools in her area. In one, the principal, who was a friend of hers, was concerned that the children seemed to display a flagrant disregard for animal life. The other school was recommended by the local humane society as a place where students had a reputation for being especially kind to animals.

Mrs. Witherspoon conferred with some teachers and the principals at both schools. She arranged to visit the classrooms on a regular basis and to make observations and conduct interviews with the students. She agreed to do this without disrupting the normal classroom routines. Two third-grade teachers were very receptive in each school, and these four teachers agreed to keep detailed notes. Mrs. Witherspoon prepared guidelines that listed general behaviors she wanted these teachers to look for.

For the entire school year, Mrs. Witherspoon spent two hours each week in each of the four classrooms. The teachers in each class considered it to be a useful language arts activity for the children to converse about interesting topics, and so Mrs. Witherspoon was able to hold individual and group interviews with the children. The teachers introduced her as a teacher who was taking a college course and needed to talk to children. The students seemed to like her, and they spoke freely. The conversations ranged among many topics - sometimes focusing directly on animals but also covering matters only indirectly related to animal life. During these sessions, Mrs. Witherspoon let the children talk freely. She tape recorded the conversations, so it was not necessary to take notes while she talked with the children. Shortly after each session she listened to the recordings and took notes based on these interviews and her other observations. In addition, the teachers had the students write occasional essays on topics of interest to Mrs. Witherspoon, and they shared these essays with her. The teachers were pleased that as time went on the children seemed to be des eloping both written and oral thoughts more effectively.

In addition, Mrs. Witherspoon spent a half hour each week interviewing the four teachers.

At the end of the year, Mrs. Witherspoon spent a week summarizing her data. She agreed that the students in the second school really were more humane in their attitudes than those in the other school. She drew many conclusions, which are too numerous to summarize here. One of the most interesting was that perhaps materialism was an important factor in attitudes toward animal life. In the past, she had thought that programs on humane values should focus on bringing cute animals into the classroom, but her insight about materialism suggested to her that it may be more important to find ways to reduce personal greed and materialism among the children. She needed to collect more data to confirm and expand her ideas.

When Mrs. Witherspoon shared her findings with Mr. Anderson, he found her ideas to be helpful. He arranged for Mrs. Witherspoon to administer his Fireman Tests in her classrooms. With Mrs. Witherspoon's detailed knowledge of the children, Mr. Anderson felt that she could provide valuable information to help validate these tests. In addition, he was particularly struck by her ideas on materialism. He found that her insights helped him to better understand the results he had found in his own research.

SUMMARY

Qualitative researchers examine naturalistic settings to look for factors that might give meaning to those situations. Qualitative research can be either exploratory or fully interpretive in nature. It is useful in its own right and frequently provides a basis for subsequent quantitative and experimental research. It offers insights into the reasons behind the events that occur in experimental research settings. Qualitative research is often conducted by participant observers, who have the advantage of being very close to the situation and can observe it in as natural a manner as possible. Participant observers have to take deliberate efforts to avoid biases and collect information as objectively as possible.

This chapter has described the basic methods of qualitative researchers and the types of data they collect. Good qualitative research depends upon systematic fieldwork over a lengthy period of time. This chapter has also discussed problems of reliability and validity in the collection of qualitative data and has examined factors that are likely to interfere with the qualitative research process. In addition, this chapter has described the characteristics of a good qualitative research report and has discussed the role that qualitative research can play in education.

What Comes Next

Starting in chapter 10, we'll begin to focus on conducting and interpreting other types of research - quantitative research that deals with cause-and-effect relationships and with generalizing results in education.

DOING YOUR OWN RESEARCH

When conducting qualitative research, your goal is to collect and report objective data that are free of the artificial constraints often imposed by quantitative methods. Consider the following guidelines, which are based on principles discussed in this chapter:

1. Use the naturalistic strategies described in this chapter to examine the whole picture. Avoid the accidental exclusion of important elements.
   
   
2. Avoid personal biases. The best way to do this is to be aware of the sources of bias described in this chapter and to take steps to prevent their occurrence.
   
   
3. Employ a reasonable strategy to compile data completely but unobtrusively. Bernard's (1988) suggestion to use field jottings, field notes, a field diary, and a field log is one good strategy.
   
   
4. Be aware of the problems that can occur in qualitative research, and avoid them. An awareness of the lists compiled by Erickson (1986) and Sadler (1981) can help you prevent costly mistakes in planning and implementation.
   
   
5. Remember that there's no law against integrating quantitative methods with your qualitative methods, if this can be done without disrupting the collection of data in a naturalistic setting.

In addition, even if your research plan emphasizes quantitative methods, consider supplementing them with the qualitative strategies described in this chapter.

FOR FURTHER THOUGHT

1. Is it reasonable to expect teachers and other educators to both do a good job educating students and to engage in qualitative research?
2. What types of problems in education are most amenable to qualitative research?
3. Complete this sentence by answering the designated questions: "Qualitative research ... (does what?) (To what or whom?) (When?) (Where?) (How?) (Why?)"

REFERENCES

Barker, R. (1964). Large schools, small schools: High school size and student behavior. Stanford, CA: Stanford University Press.

Bernard, H. R. (1988). Research methods in cultural anthropology. Beverly Hills, CA: Sage.

Coleman, J. S. (1961). The adolescent society: The social life of the teenager and its impact on education. New York: Free Press of Glencoe.

Erickson, F. (1986). Qualitative methods in research on teaching. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed.). New York: Macmillan.

Goetz, J. P., & LeCompte, M. D. (1984). Ethnography and qualitative design in educational research. Orlando, FL: Academic Press.

Kirk, J., & Miller, M. L. (1986). Reliability and validity in qualitative research. Beverly Hills, CA: Sage.

Maxwell, J. A. (1992). Understanding and validity in qualitative research. Harvard Educational Review, 62, 279-300.

Mead, M. (1928). Coming of age in Samoa. New York Morrow.

Reiss, A. J., & Rhodes, A. L. (1959). A socio psychological study of adolescent uniformity and deviation. Oct. 31, 1959, CRP 507, ERIC ED 002 890.

Sadler, D. R. (1981). Intuitive data processing as a potential source of bias in naturalistic evaluations. Educational Evaluation and Policy Analysis, 3(4), 25-31.

Wolcott, H. F. (1990). On seeking-and rejecting- Validity in qualitative research. In E. W. Eisner & A. Peshkin (Eds.), Qualitative inquiry in education: The continuing debate (pp. 121-152). New York: Teachers College Press.

ANNOTATED BIBLIOGRAPHY

Bernard, El. R. (1988). Research methods in cultural anthropology. Beverly Hills, CA: Sage. This is a good textbook from the field of cultural anthropology, a field that relies heavily on qualitative research and in which methodologists have developed and perfected numerous effective strategies.

Erickson, F. (1986). Qualitative methods in research on teaching. In M. C. Wittrock (Ed.), Handbook of research on teaching (3rd ed.). New York: Macmillan. This chapter presents a good summary of the status and methods of qualitative research in education.

Finders, M. (1992). Looking at lives through ethnography. Educational Leadership, 50(1), 60-65. This article includes excerpts from several ethnographic studies. These can give teachers insight into how qualitative research can contribute to their daily practice.

Patton, M. Q. (1987). How to use qualitative methods in evaluation. Newbury Park, CA: Sage. This book covers issues to consider when deciding whether to use qualitative research methods as well as strategies for conducting qualitative studies.

Seidman, I. F. (1991). Interviewing as qualitative research: A guide for researchers in education and the social sciences. New York: Teachers College Press. This is a good introductory text for researchers with no background in qualitative research. It connects interviewing techniques with broader issues in qualitative research.

Research Report Analysis

The research report in Appendix C is primarily an example of quantitative rather than qualitative research. However, the conclusions section incorporates an important qualitative component that makes the report more meaningful. Find that qualitative component on page 464 and answer the following

1. In what way does this qualitative information enhance the research report?
   
   
2. In what other ways could qualitative research strategies have been incorporated into this study and report?
   
   
3. What would a primarily qualitative study of the same topic look like? (Give a brief description of what the report of a qualitative study could say.)

ANSWERS:

1. This qualitative information gives us an anecdotal example of what the numbers in the report tell us. The reader of the report can vividly picture the conversation described in this anecdote, and this puts the quantitative descriptions into perspective.
   
   
2. Additional anecdotal descriptions could further elucidate the thinking processes the students went through. For example, the report says that a major advantage of the computer simulations is that students can run programs repeatedly and learn from their mistakes. The report could have described how some students did this and how they acted differently on subsequent runs of the programs. Likewise, an anecdotal description of how the teachers implemented the programs and what the students did while running them would help support external validity, because teachers reading the report could more easily decide how closely the activities on which the report is based resemble what goes on in their own classrooms.
   
   
3. A fully qualitative study would devote attention exclusively to issues such as those described in the second paragraph of the "Discussion and Conclusions" section and in the answer to the preceding question. For example, a participant observer could watch students running the set of programs and report on the sorts of things that the students and teachers did when using them as a tool for studying the designated topics. The report could focus on the thinking and feelings expressed by the students and teacher and on specific behaviors the students engaged in that would seem to demonstrate known psychological principles such as cooperative learning, attribution theory, and strategic thinking. Additional descriptions could describe the social context (e.g., parental, administrative, and community pressures, forms of support, and interests).

Note that similar qualitative components could improve the quality of most quantitative research.