Theory And Method PDF

P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 THEORY AND METHOD 2...

P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 THEORY AND METHOD 2 INTRODUCTION 17 THEORY AND SOCIAL RESEARCH 18 What Theory Looks Like 18 Taxonomies, 18 r Models, 20 r Typologies, 21 r Paradigms, 22 Description and Explanation 23 The Verification and Discovery of Theory 23 Deduction, 23 r Induction, 24 THE DEVELOPMENT OF RESEARCH IDEAS 25 Curiosity 25 Pure and Applied Research 25 Public Sociology 26 Already Existing Theory 26 Training and Experience 28 INTELLECTUAL CRAF TSMANSHIP 29 Learning the Craft 29 Using Research Imagination 30 SUMMARY 30 KEY TERMS 31 EXERCISES 31 SUGGESTED READINGS 31 REFERENCES 31 a guide in the collection and analysis of data. INTRODUCTION Moreover, in many studies where the researcher In Chapter 1, the scientific method was des- does not begin with a thoroughly defined theory, cribed as a process that contains theory and the we can create theory by searching through the testing of hypotheses or hunches suggested by data for recurring patterns of behavior. the theory. Indeed, in all of the various social sci- The application of social theory to research is ence disciplines,1 and no matter what method one of the prime areas for the use of the research of data collection is employed, theory is used as imagination. This chapter explains the relation- ship between theory and methods in detail. It 1 Psychology, sociology, political science, anthropology, and describes the various forms that theory may economics. take. It compares and contrasts research that is 17 https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 18 Theory and Method designed to verify existing theory with studies represent conceptual frameworks that guide the that are designed to create new theory. Then, analysis of data as they are being collected. To the source of research ideas or how researchers understand how these schemes work, consider decide what to study is discussed. Finally, the what happens when the morning mail arrives idea of research as a “craft,” or set of related skills at a bustling business office. It has to be sorted; that can be practiced and learned, is introduced. perhaps it is distributed in boxes or pigeonholes, each one with a person’s name on it or with the name of a department. In this analogy, the morn- THEORY AND SOCIAL RESEARCH ing mail is the data. At first it is all jumbled and There are numerous options open to the inves- undifferentiated, but we can make sense of it by tigator in choosing a theoretical perspective. placing each piece in the proper box. Many theo- Moreover, the many different kinds of social the- ries approach the explanation of social reality in ory vary in terms of the specific concepts they a similar way, by atomizing it, or breaking down contain, how complete or well articulated they observable phenomena systematically. Data are are and in the way they are expressed on paper, organized according to a diagram or to a list of with words or diagrams. An exhaustive review is categories, corresponding to a row of mailboxes. beyond the scope of this book; however, in this Thus, each bit of information gathered by the chapter, some examples of theories that have researcher, whether it is what someone says, how been particularly influential in social research someone looks, or what someone has written, are offered to illustrate three key points about fits in somewhere in a well-articulated scheme. the relationship between theory and the re- If a taxonomy, model, or typology merely sup- search process: plies labels for each of its conceptual “boxes,” it is basically descriptive, rather than explana- r The scientific method is compatible with a tory. It may be suggestive of theory, but it can- variety of theoretical approaches to data. not stand as a complete presentation of theory. r Creating and expanding theory itself are We can make only a few generalizations about prime motives for research. the business office by looking at each pigeonhole r The choice of theory is crucial because it may separately. What makes these schemes more influence one’s topic for investigation, as well valuable theoretically is the explanation of how as one’s conclusions. their various components relate to one another. Thus, each datum not only belongs in a certain category; it also carries implications for the rest What Theory Looks Like of the data. It is useful to know how many letters “Hunches,” as described in the diagram of the go in each mailbox; it is perhaps more important scientific method (see Figure 1.1 in Chapter 1), to know why some mailboxes are always much are really theories in embryo form; they are spec- fuller than the rest. ulations that have a relatively short life span. The foregoing generalizations are illustrated They may never actually be written down. By in the following examples. contrast, social theories are more elabo- rate, general explanations of human behavior, TAXONOMIES. Perhaps the easiest of the classifi- and they usually take a concrete form. Some catory schemes to understand is a taxonomy, of the formats used to express theory in social or list of categories. An example depicted in Ta- research are taxonomies, models, typologies, ble 2.1 is the Taxonomy of Educational Objectives and paradigms. To show what theory actually (Anderson, Krathwohl, and Bloom, 2000), used looks like in practice, these expressions of the- in research on teaching and teaming in schools. ory will be examined in this chapter. The authors created definitions of general con- Taxonomies, models, and typologies are sche- cepts such as knowledge and intellectual ability mes for classifying data. Some of them are and separated each into its component parts. extremely lengthy and intricate. In practice, they They designed this to be an exhaustive list, in https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 Theory and Social Research 19 that there are many occasions where students Table 2.1. The Taxonomy of Educational showed their “comprehension” of the lessons Objectives by making outlines of chapters, reciting what Knowledge they had memorized, or putting the teacher’s Knowledge of specifics lectures into their own words. There might be Terminology fewer entries in other categories, that is, compar- Specific facts atively less opportunity for students to apply, to Knowledge of ways and means of dealing analyze, or to synthesize what they had learned. with specifies These data might help the teacher understand Conventions why some students appear uninterested in class Trends and sequences or are discipline problems or why test scores Classifications and categories are lower than expected. Because the taxonomy Criteria shows that there are many types of learning Methodology that the data indicated had been ignored, the Knowledge of the universals and researcher may suggest some specific changes abstractions in a field in assignments and lectures to involve students Principles and generalizations more actively in learning. Theories and structures Thus, taxonomies may point the way toward Intellectual Abilities and Skills the explanation of human behavior and to- Comprehension ward policy recommendations. Aside from the Translation exhaustive description of knowledge and abili- Interpretation ties, there is no real theory of learning expressed Extrapolation in the list of categories in Table 2.1. Nowhere in Application the scheme is it stated that the curriculum in Analysis each school subject must contain all types of Analysis of elements knowledge or that tests must measure all kinds Analysis of relationships of intellectual ability. The taxonomy does not Analysis of organizational principles state that it is more difficult, or more desirable, Synthesis to acquire one type of knowledge than another. Production of a unique communication It does not specify the degree of difficulty for Production of a plan or proposed set of each academic skill, nor does it specify that operations each must be acquired in a logical sequence Derivation of a set of abstract relations or order. In fact, there is no particular relation- Evaluation Judgments in terms of external criteria ship proposed between any of the elements in the taxonomy; they are merely labels. As the From Lorin W. Anderson, David R. Krathwohl. A Tax- authors themselves note, researchers may use onomy For Learning, Teaching, And Assessing: A Re- the categories “in very arbitrary fashion,” out of vision of Bloom’s Taxonomy Of Educational Objec- sequence in the scheme, so long as their defini- tives, 1/e. Published by Allyn & Bacon, Boston, tions remain intact (Anderson, Krathwohl, and MA. Copyright
C 2001 by Pearson Education. By Bloom, 2000:10–11). permission of the publisher. Thus, the primary utility of taxonomies is for description. A complete list of objectives in any other words, to contain every type of knowledge field, whether it be education, business, com- and every kind of intellectual ability and skill. munity organization, or medical treatment, is The taxonomy is an aid in sorting data. The invaluable as a baseline against which to mea- researcher may, for instance, observe the work sure performance. Having commonly accepted of a particular teacher for several weeks, placing definitions of concepts such as intellectual skill each class exercise and homework assignment makes it easier to compare behavior in a variety somewhere in the scheme. It may be discovered of settings. This same list of categories could https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 20 Theory and Method The Metropolitan Mental Federal & State Health Center Agencies Executive Director Medical Director Director of Director of Director of Clinical Services Administrative Consultation and Services Education Adult Emergency Day Geriatric Management Counseling Services Treatment Services Information Services Services Evaluation Payroll Court Clinic Youth Aftercare Billing, etc. Services Services Program Development Case Management Mental Retardation Residential Direct Responsibility & Mental Health Services Contract or Affiliation Agreement Figure 2.1. An organizational model. be used to evaluate the curriculum in English, atric services and adult counseling, communi- chemistry, sociology, history, or any other school cate with each other directly, although they all subject. Finally, the taxonomy is useful for devis- are the responsibility of the director of clinical ing strategies of data collection. It tells us what services. the possible types of behavior are in a given Like the taxonomy we have already examined, setting; thus, we are encouraged to look for this model is suggestive of theory. We may, for them, and, if some behaviors are not present, instance, speculate that there could be friction we can wonder why not. between administrators and clinicians because they rarely, if ever, interact on a daily basis. MODELS. A model is a visual depiction of how In analyzing the role of the executive director, something works. It is a prototype to which the we may look for evidence of tension between real world is compared as data are gathered. administrative and clinical duties. Finally, the Some models are almost entirely descriptive, in chart may lead us to investigate the difficulties that they are really labels arranged spatially in a of coordinating the several, diverse departments diagram. Figure 2.1 is an organizational chart of on the left of the diagram. a mental health clinic. It tells us who the prin- These hunches and insights may help us cipal actors in the bureaucracy are, and it spec- determine how the organization works, but the ifies the lines of communication and authority chart does not contain a real theory of group between them. It also details the responsibilities functioning or structure. It could not explain of the various parts of the clinic. The adminis- why two clinics, set up with substantially the trative wing, on the right, relates to the depart- same official positions and lines of authority, ments that provide direct service, on the left, may differ greatly in their effectiveness and effi- through the office of the executive director. The ciency. The diagram, informative as it is, does various clinical departments, for example, geri- not indicate how many employees work for the https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 Theory and Social Research 21 Father's Respondent's Education Education Figure 2.2. A model of the process of Respondent's socioeconomic achievement. Present Occupation Father's Respondent's Occupation First Job clinic, the caseload, what sort of neighborhood we often must sacrifice rich, descriptive detail. it serves, or other details that may be theoreti- Both descriptive and explanatory mod- cally relevant to the study of mental health care els are thus useful in the social sciences, but delivery. the descriptive models are only aids to theory con- By contrast, Figure 2.2 presents a model that struction, whereas explanatory models are theory is primarily explanatory. The components of itself. Figure 2.2 are not parts of any single organi- zation. Instead they are variables (character- TYPOLOGIES. A typology is a device for analyz- istics, attitudes, or behaviors that can be mea- ing all the logical combinations of at least two sured and that take on differing values). These variables. Figure 2.3 shows a simple descriptive variables are used to explain part of the process typology for examining a population of college of social class formation; namely, what leads a students according to the variables grade-point respondent (person who is the subject of inves- average and intelligence measured by an IQ tigation) into a particular occupation. (intelligent quotient) test. Nine student “types” According to the model in Figure 2.2, a respon- are conceptualized in the cells, or boxes, in this dent’s choice of employment is influenced by the table. A person classified as a “straight arrow” nature of the first position that person obtained (type 1) is someone with both high grades and a on entering the job market. Equally important, high IQ. A “striver” (type 8) has fair grades but a in theory, is the amount of education that the low IQ. Two other types are “marginals,” whose respondent has received. Two other variables poor grades are perhaps more consistent with are thought to be significant – the education their low intelligence test scores, and “apathet- and the occupation of the respondent’s father. ics,” who, despite superior IQs, have only fair The model thus reflects the idea that a person’s grades. own educational history is greatly influenced by This scheme lets us compare all students on his or her parents. A father’s occupation may a given campus; everyone may be placed some- influence his children’s career patterns directly where in the typology. However, it has additional (“following in his footsteps”) as well as indirectly, theoretical utility. We expect to find some people through the amount of education that they are with average-tested intelligence in each of the given and the initial occupational goals that they three grade-point average categories (types 4, 5, are encouraged to set for themselves. There are and 6). The individuals whom we have labeled many statistical procedures that may be used to “underachievers” and “overachievers” (types 3 verify the relationships among the variables in a and 7) are probably more unusual, in that there model of this type.2 is a great disparity between their tested intel- Figure 2.2 has much more explanatory power ligence and their grades in college. These are than Figure 2.1, the model of the mental health precisely the sorts of people we should examine clinic. To create general explanations, however, closely if we want to understand fully the the- oretical connection between natural talent and 2 More on this point in Chapter 19. actual performance. Yet, without the typology, https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 22 Theory and Method Intelligence [IQ] High Average Low 1 4 7 High “Straight Arrows” “Pluggers” “Overachievers” Grade-Point Average 2 5 8 Average “Apathetics” “Normals” “Strivers” 3 6 9 Low “Underachievers” “Slackers” “Marginals” Figure 2.3. Typology of college students according to grades and IQ scores. we might fail to isolate them, or we might con- that their performances are genuine. Thus, centrate exclusively on students whose IQ scores Goffman forces us, as observers, to confront were more closely related to their grade-point the manipulative, sometimes artificial quality of averages. people’s contacts with one another. If the “act” is successful, the audience gives people approval PARADIGMS. Another sort of theorizing is of and confirms them in their roles. Only “back- broader scope and not as easily diagrammed as stage” in areas hidden from the public are people those we have looked at so far. It is the applica- permitted a respite from their acting chores. tion of a paradigm, or coherent worldview, to If we accept this paradigm, we see the basis social life. In effect, when we follow a paradigm, of social reality continually shifting along the we put on a pair of glasses that colors all behavior dimensions of managing impressions, putting with a particular interpretation. our “best foot forward,” and hiding imperfec- A classic example from social psychology is tions. The categories Goffman develops to orga- the work of Erving Goffman (1959), who has nize an enormous quantity of data provide attempted to explain social interaction by strong insights into interpersonal relations. No uncovering its basic processes. The inspiration one act is seen as being any more real or true for Goffman’s paradigm is Shakespeare’s claim than another. Acting per se is part of the human in Hamlet that “all the world’s a stage and all condition. the men and women merely players.” He takes Another paradigm, one of the most influential that notion seriously and offers evidence for it. in twentieth-century American social science, is In fact, Goffman’s work is called dramaturgi- the image of society associated with Talcott Par- cal because of the close analogy between social sons (1902–1979). It is known as structural life and what occurs “on stage.” According to this functionalism in sociology and political sci- paradigm, human beings are all “actors” who, ence and as systems theory in social work depending on the situation, must play a variety and business management. According to this of roles for society, the “audience.” People are paradigm, every element of a society that exists constantly trying to convince their audiences over a period of time serves a distinctive function, https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 Theory and Social Research 23 helps to maintain the social system, and is sup- how something works, we must know its dimen- ported by public consensus or agreement (Par- sions and the identity of its components, as well sons, Shils, and Smelser, 1962/2001; Alexander, as the general principles of its operation. There- 1998). Society is much like any living organism, fore, in making use of theory, a balance must claims Parsons, so that a change in any one social be struck between description and explanation. institution will result in corresponding changes We must have both, but here is a dilemma: As in other elements of the system. Any alteration in we become wrapped up in the vivid descrip- the economic structure of a society, for example, tion of a single event or person, it becomes diffi- will cause complementary changes in the polit- cult to generalize about other similar or related ical, religious, and educational spheres. Func- phenomena. Conversely, if we place less empha- tionalists argue that any social system is always sis on description than on general explanation, moving toward a state of equilibrium. In this theory tends to become a series of disembod- paradigm, society is much like a rubber ball that ied, self-evident propositions. Perhaps there is may on occasion be squeezed out of shape but no foolproof solution to this dilemma, but the is always striving to return to its original form. attempt to solve it is a constant challenge in The questions for functional analysis are there- social science. fore, What function is performed by each social element, and what would be the consequences The Verification and Discovery of Theory for the social system as a whole if it were absent? The explanations provided by functionalism Thus far some of the many ways that social the- are not very concrete. They seem abstract and ory may be expressed have been shown. How- general because they are meant to apply, at ever, the form that theory takes does not deter- several levels of analysis, to families, business mine its place in the research cycle. We may enterprises, ethnic groups, nations, and even begin with observation and gradually discover the world system. Nonetheless, functionalism or create a model, a typology, or a general expla- has been an attractive paradigm because it con- nation for behavior. Or we may initiate research firms the scientific notion of an orderly uni- with the theory firmly in mind and through test- verse, in which there is a place and a reason ing attempt to verify it. This section discusses for every element of society. It makes a very the implications of these two approaches for the complicated world seem more intelligible by ongoing practice of research. proposing that the relatively small social groups to which we belong operate according to the DEDUCTION. In the physical sciences, the typical same principles as the larger society. Finally, research strategy is to begin with a theory and although the paradigm of functionalism is rel- then to subject it to observation. This mode of atively weak in providing specific explanations, inquiry is called deduction, and it has been it is extremely flexible. Like the dramaturgical most influential in the social sciences as well. We perspective, there is scarcely a human action or start with general principles and subsequently attitude that cannot be fitted into the function- deduce whether they are sound. Deductive the- alist conceptual scheme and vocabulary. ory does not emerge immediately from the data; A third key paradigm, which will be exam- it is conceived beforehand and applied to the ined in detail in Chapter 10, is inspired by fem- data. inism. It sees research about women, and per- To apply theory to data, the researcher for- formed primarily by women, as contributing to mulates hypotheses, specific predictions that the improvement of women’s position in society. follow from the general theory. Recall the model presented in Table 2.2 showing the process of socioeconomic achievement. This model Description and Explanation explains a person’s present occupation by con- Theory, as we have seen, takes many forms, but sidering his or her first job and level of education, all help us to understand the world. To find out as well as his or her father’s occupation and level https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 24 Theory and Method of education. How could we develop hypotheses induced. Measurement of variables is as impor- from this model? We would need to make some tant to inductive research as it is to deductive testable predictions based on its assumptions; research, but neither hypotheses nor measur- for example: ing tools are developed in advance. Instead, the r The less well educated a respondent’s father, research problem emerges as a result of direct confrontation with a set of behaviors. The pri- the less well educated the respondent r The higher the status of a respondent’s father’s mary focus for study, the development of theory, and the production of an analysis may emerge occupation, the higher the status of a respon- at any point in inductive research, even toward dent’s occupation the very end of the process. Induction is initially The researcher, in fact, might produce a series of exploratory and vividly descriptive. Investiga- additional hypotheses that could be tested using tors must take in a vast amount of information a variety of deductive research methods. because they have to develop categories for clas- Because the prediction of behavior is central sifying data, based not on already existing theory to the process of deduction, it is difficult to begin but on the actual range and variety of data in the that process with only a descriptive model or field. a very general paradigm. Each of the separate No list of examples could do justice to the vast variables in a deductive theory must be care- amount of excellent inductive research that has fully defined in advance, so that the researcher been conducted in the United States in recent will recognize them when they appear in the real years. Among the countless variety of topics world, and so that they may be measured. We that have been explored are inner-city life, the have to be as rigorous as possible because as we working poor, occupations, and leisure. Elijah create measures and variables, we are in effect Anderson (2003) uses induction to examine creating a yardstick against which everything we social class and street-corner life of the inner later observe will be compared. If our model sug- city. Mitchell Duneier (1999) describes the world gests that a large organization will experience of sidewalk booksellers. Jody Miller (2006) ana- difficulty in communication among its parts, we lyzes girls in gangs. Marvin Scott (2005) exam- had better be specific about what constitutes a ined “The Racing Game.” Cristina Rathbone large organization and what constitutes a bar- (2005) vividly recounts the experience of women rier to communication. We need such reliable in prison. As downturns in the economy have measures because we may have to examine as placed particular strain on the working poor, a many as a hundred different groups to confirm number of studies have focused on people who or disconfirm our theory. struggle to improve their position on the social The more general, abstract, and purely ladder. These include Sharon Hays’s (2004) anal- descriptive the model or paradigm, the greater ysis of the effect of welfare reform and Barbara the problems of measurement and explanation, Ehrenreich’s (2001) account of her own adven- and the more difficult it is to use in deductive tures in taking a variety of low-paid jobs around research. For this reason, models that specify the the country. relationship among several well-defined vari- There is a lively tradition of exploration of ables are more easily adapted to the process of various stages of the life cycle. Recent studies deduction. that have continued in this vein are Milner’s (2006) work on adolescent culture and identity INDUCTION. The major alternative mode of and Dorothy and David Counts’s (2001) exami- inquiry is induction. Its hallmark is the dis- nation of the on-the-road culture of senior citi- covery and building of new theory as research zens and their recreational vehicles. As you read progresses. The research cycle is begun with further in this book, you will learn some of the observation. From the data collected, a gener- methodological techniques that were employed alized understanding of behavior is gradually to obtain data in these diverse contexts. https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 The Development of Research Ideas 25 Although induction is usually identified only tive strategy is often not followed to the letter, with the research goals of exploration and des- even in the physical and natural sciences. In all cription, the explanation of social reality may disciplines, the development of theory is neces- also be created as research is being carried out sarily both inductive and deductive. Scientists (Glaser and Strauss, 1967; Strauss and Corbin, may begin with a theory, try to test it deductively 1998). By generating hypotheses on a day-to-day with actual data, and find that it does not predict basis, and discarding them when predictions are well. At that point, they may choose to modify not verified by observation, we may create a the- the theory to make it more consistent with the ory. In a classic study, Becker (1953/1993) began data. When they do this, they are beginning to to study marijuana use by looking at the history engage in an inductive process of inquiry. of people’s experience with the drug and eventu- ally formulated a theory of deviancy. Katherine THE DEVELOPMENT OF RESEARCH IDEAS Newman carefully observed and talked to young men and women who are part of the “working Even though the relationship between theory poor” in Harlem, primarily in the food industry. and research has been explained in this chap- She found that they “do not need their values ter, an important question remains: Where do reengineered. They do not need lessons about research ideas come from? This is one of the keys the dignity of work.” Newman (1999:297–298) to the research imagination! concludes that they need jobs that pay a living wage. Curiosity These, and countless other inductive expla- nations, tend to be highly valid, particularly Social researchers are generally intensely curi- when they are based on lengthy fieldwork. Over ous people. They want to know about those dif- many months the process of induction gradu- ferent from themselves, for example, an African ally eliminates the weaker alternative explana- society (Trefon, 2005); mental patients in a back tions for the behavior of the particular individu- ward (Knipfel, 2000); the very poor (Kozol, 1996); als observed. The chances become less and less or the very rich (Herman, 1999). Often they begin that important variables remain hidden from to investigate some setting or group for little rea- the investigator. However, inductive explanation son other than that they are intrigued by, or per- tends to be less reliable than it is valid. The mea- plexed by, a set of behaviors. What is it like to suring tools used in this mode of inquiry are be a woman in a motorcycle gang (Hopper and developed on an ad hoc basis, and they are influ- Moore, 1990)? Why can’t “Johnny” read at the enced to a considerable degree by the unique appropriate grade level (Spear-Swerling, 1997)? experience of individual researchers. Thus, if Why was the 2004 election so divisive (Sabato, we use inductive, as opposed to deductive, 2006)? These are questions that have inspired research strategies, the chances are increased social research. All scientists grapple with mys- that another investigator may come to equally tery. Social scientists, in particular, are attracted valid, yet different, conclusions. The theoretical to those whose actions and motives are, at least explanations created by induction also may be at first glance, unclear or puzzling. Once they are less generalizable from one setting to another.3 attracted to a subject, the logic of science fuels Despite these difficulties, inductive research their curiosity and their imagination. The scien- is no less scientific than the deductive approach. tific method disciplines the raw enthusiasm of The tension over reliability and validity is felt the researcher but does not dampen it. by all researchers, no matter what their field of inquiry. We may have to sacrifice a little of one Pure and Applied Research to improve the other. Indeed, the purely deduc- The investigator’s curiosity is not confined to 3 For more details on this point, refer to Chapter 9. bizarre settings and to social problems of an https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 26 Theory and Method immediate nature. Some science involves basic, dialogue with them. However, they also remain or pure research, that is, finding the solu- true to the theoretical insights of professional tion to questions that are intellectually challeng- social science. A leading proponent of public ing but that may not have practical applications sociology, Michael Burawoy (2005), makes the in the short run. Pure research in social science point that he is “unequivocally committed to the is primarily devoted to expanding theoretical values and practice of professional sociology – horizons; thus, the actual setting for data col- its rigour, its science, its research programmes, lection may be of secondary importance. If, for its care to get things right, its concern with the- example, we are keenly interested in how peo- oretical issues.” The primary purpose of pub- ple learn the grammatical structure of language, lic sociology, in his view, is “promoting public we might choose to gather data in an elemen- discourse.” However, he notes, “Without profes- tary school. There we would find out a great deal sional sociology there is no public sociology.” about how schools function, how teachers orga- An excellent example of these principles is nize their time, and how young children dress provided by Diane Vaughan (1996), whose aca- and play. Any and all of this information might demic investigation into the Challenger and Col- be of use to school administrators, teachers, and umbia space shuttle disasters has led to media parents, but our concentration in this setting exposure and numerous consulting opportuni- would be on data about language acquisition. ties. In the box on pages 27–28, Vaughan chroni- In fact, we could probably discover similar pro- cles her foray into the realm of public sociology. cesses of learning in another setting entirely, say, on a street corner, or in a home. Already Existing Theory applied research, however, is inspired by the needs of social action. Its findings and con- In addition to researchers’ basic curiosity and clusions are applied immediately to solve a prob- the problems they want to help solve, another lem or to improve the effectiveness of an existing major source of research ideas is the storehouse or proposed social program. Are the agencies set of theory that has been built and expanded by up to help poor people serving their intended social science. These ideas channel the devel- clientele? How many parking meters should be opment of research. It is easy to see why this put on Main Street? How can hospital emergency is so when we consider that deductive strate- rooms be made more efficient? These are exam- gies for the collection of data have dominated ples of questions that inspire applied research. the study of human behavior for the past sixty The answers to these questions may have theo- years. Even when we use the inductive approach retical significance, but the investigations were to data, however, it is extremely difficult to manifestly intended to help people. The results enter the field with a clean slate. Ideally, induc- of basic research may offer practical assistance tion begins with no theoretical preconceptions, to the public sooner than expected, but its pri- but although researchers may not write down mary purpose is the accumulation of knowledge hypotheses or sketch models in advance, they for its own sake. still cannot fully erase the memory of a lecture, a book they have read, or the example of concep- tual skill provided by another investigator whom Public Sociology they respect or admire. In practice, public sociology appears to be a In a book that is still widely read and dis- compromise between pure and applied research cussed, Alvin Gouldner (1970) argued that in that seeks to effect social change and public pol- the latter decades of the twentieth century there icy (Agger, 2000). Practitioners of public soci- was a reluctance to work at testing new ideas ology (e.g., Derber, 2000; Putnam, 2001) try to in the investigation of social life and a tendency communicate in language that resonates with the to deal with the same old assumptions about people, groups, and organizations that compose society, merely applying them to new settings. the nonacademic community, and to establish a Although we might agree with him concerning https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 The Development of Research Ideas 27 How Theory Travels: A Most Public Public Sociology The tragic disintegration of NASA Space Shuttle Columbia on February 1, 2003, sent me on an unexpected and remarkable eight-month journey in public sociology. Hours after the accident, I was deluged with press calls stemming from my study of the causes of the 1986 Challenger disaster. Recognizing the teaching opportunity and professional responsibility, I tried to respond to everyone. I was teaching the theoretical explanation and key concepts of the book, linking them to data about Challenger and Columbia as changing press questions dictated. Because the investigation went on for months, these conversations became an ongo- ing exchange where the press brought me new information, and I gave a sociological interpretation. I noticed that the concepts of the book – the normalization of deviance, institutional failure, organization culture, structure, missed signals – began appearing in print early in the investigation and continued, whether I was quoted or not. Two weeks after the accident, the publicity director at my publisher sent a copy of the Challenger book to retired Admiral Harold Gehman, who headed the Board investigating the Columbia accident. As the Admiral later told me, he read it mid-February, along with my jargon-free condensation published in a management journal. Persuaded of the relevance of the sociological analysis to Columbia, he sent copies of both to the Board. The Admiral and the Board members were experienced accident investigators, trained to look beyond technical causes to human factors, but the organizational focus and concepts of the book were new to them, helped make sense of their data and led them to other social science sources. The Admiral believed that history was a scene-setter, not a cause. Citing exam- ples from the Challenger case, I explained how historic decisions in NASA’s political and budgetary environment changed the organization structure and culture, ultimately affecting risk decisions, thus contributing to both accidents. He was dubious [but as we collaborated] information and ideas flew fast and freely. Extraordinary investigative effort, data, analysis, and insights were integrated into my writing; sociological con- nections and concepts became integrated across the chapters of the Board’s Report on the accident. It was based on their data but the outline of chapter topics paralleled my data and causal model. The Admiral, it turned out, was “delighted” with the result. The Board, too, accepted “History as Cause: Columbia and Challenger” as a chapter in the Report, along with its implications for the expanded causal model. The new centrality of sociological ideas and the connection with the Challenger acci- dent were not lost on the media. In press conferences, Admiral Gehman stressed the importance of the social causes. When he announced that I would testify before the Board in Houston, the field’s leading journal, Aviation Week and Space Technology, headlined “Columbia Board Probes the Shuttle Program’s Sociology.” My testimony covered the causes of the Challenger accident, comparison with Columbia, and iden- tification of systemic common institutional failures. The book’s theory and concepts traveled farther as my testimony – like that of other witnesses – aired live on NASA TV and video-streamed into TV, radio, and Internet outlets. The New York Times announced the equal weight the report would give to technical and social causes, identifying me as the source of the Board’s approach. The language of sociology became commonplace in the press. The theory of the book traveled one more place that August week. An AP wire story, “NASA Finally Looks to Sociologist,” revealed that NASA had invited me to headquarters to talk with top officials, who https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 28 Theory and Method shifted from denial to acknowledgement that the systemic institutional failures that led to Challenger also caused Columbia. Never did I foresee the extent of my involvement nor my impact. To give an idea of the extent of public and press interest in a sociological interpretation of the disaster’s causes, I had been quoted in print 50 times by the end of May. [I was] elucidating principles that bring sociology alive, out of textbooks, academic monographs, and classrooms and into the public consciousness and policy debate. Sociology was the instigator of it all. The theory and concepts that explained Challenger led to these con- nections because they were an analogical fit with the Columbia data and made sense of what happened for journalists and the Board. My book and university affiliation gave me the opportunity to engage in ongoing dialogic teaching – akin to daily grass-roots activism – but with these two tribunals of power with authoritative voice... translating the ideas of the book into grist for critical public dialogue. Excerpted with permission from Diane Vaughan. 2003. “How Theory Travels: A Most Public Sociology.” ASA Footnotes, 31 (8) November/December. the limitations of the theories bequeathed to us structural functionalism does not deal effect- by previous generations of researchers, we must ively with issues of social change and deviance also recognize that science develops through the (Harper, 2006). Their paradigm contrasts sharp- continual testing of already existing theoretical ly with Parsons’ because they see change and ideas (Kuhn, 1996). conflict as continuous and natural, not disrup- Often the setting or subject for research is tive of social order. Whereas functionalists main- chosen because it appears to be ideal for test- tain that every society rests on the consensus ing a theory or a part of a theory. An extraordi- of its members, their critics believe that every nary number of studies have stemmed from the society exhibits constraint of some of its members work of Parsons, Shils, and Smelser (1962/2001). (Ritzer and Goodman, 2003). Investigators, starting with functionalist theo- A change in theoretical position encourages ries, have analyzed large-scale organizations, changes in the kind of data that need to be schools, and the military, as well as the gen- acquired. Thus, as some researchers have moved eral phenomenon of social stratification. Davis away from the paradigm of functionalism, there (1945/1993) concluded that our system of dis- has been increased emphasis on the investiga- tinct social classes was of great value to society, in tion of groups that contribute to the disrup- part because we need to reward people of talent tion of the social system, for example, radi- and skill at a higher rate than those with less abil- cal students and gays (Burawoy et al., 1991; ity and ambition. The great disparity between Gamson, 1999). In addition, more attention has rich and poor in America is desirable, according been devoted to power relations of dominance to this view, to motivate people to perform the and subordination in the economy, law enforce- important tasks of leadership in business and ment, among racial and ethnic groups, and government. between the sexes. Studies based on Parsons’ paradigm have been challenged by those who see functional- Training and Experience ism as inherently conservative and overly sup- portive of the status quo and the interests of the Finally, in reviewing the sources of research “establishment.” Therefore, much research has ideas, we must consider the training and expe- been generated out of what social scientists see rience of the investigator. The nature of one’s as the limitations of systems analysis. Marxists employment is important because a certain type and other conflict theorists argue that of research may be expected in a particular https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 Intellectual Craftsmanship 29 department or agency. The priorities of the gov- tion or explanation in data analysis, nor does ernment bureaus that provide funding for the it tell us how much validity may be sacrificed social sciences also influence the goals of the re- to improve the reliability of a study. It does not search and the settings chosen. Often the selec- guide us automatically to the theory that fits our tion of a specific model or paradigm is influ- data best or to the most effective technique of enced by one’s colleagues and teachers at work data collection. Weighing such decisions skill- or in the university. fully is only partly a science; it is also a craft that These pressures can never be entirely elimi- calls for research imagination. The researcher is nated, nor should they be. Science is a cooper- both a scientist and craftsperson whose toolkit ative enterprise, and researchers often require includes a vibrant imagination. some assistance in formulating research ideas. Noted sociologist C. Wright Mills (1959/2000: The choice of theories, models, and paradigms is 215–223) explained that a true craftsman is crucial. It may, as we have seen, help determine someone who maintains the image of a com- the subject of an investigation. Even more signif- pleted product from start to finish, someone icant is the effect of theory selection on research who knows everything that goes into it, even if findings and conclusions. Whether a criminal is some of the tasks are performed by others. It is portrayed as a victim of society or as one who in this sense that the skilled researcher practices victimizes society may depend less on the crime a craft. Craftsmanship follows a plan devised by committed than on whether the investigator was the worker. This plan, whether it is the pattern for influenced by labeling theory or by functionalist carving a rocking chair or the design for exam- theory. ining a social problem, is of course shaped by To the layperson this may not sound very sci- the worker’s imagination and prior experience. entific, but the scientific method does not spec- A researcher may use a theory that has been ify which theory is most appropriate or the form used before, in the same manner as the car- in which it should be presented. Because these penter decides to make another chair like the choices are to be made by researchers, they need one in his parlor. However, the plans of both to be aware of the ways in which their prior train- are also highly individualistic and subject to ing, and the expectations of others, influence modification as the work progresses. The result, the decisions they make. Data do not “speak therefore, reveals the personality and interests for themselves.” It is up to the researcher, using of the worker. The scientific canon of objectivity social theory, to demonstrate the significance of does not mean that research has to be uniform data. It is an awesome responsibility. and colorless. The difference between ordinarily and finely crafted studies is comparable to the distinction between mass-produced and one- INTELLECTUAL CRAF TSMANSHIP of-a-kind items. Despite the freedom to create If by now you suspect that there are some dilem- which craftsmanship affords, there is, inevitably, mas in research methods that cannot be re- some mechanical drudgery in some phases of solved once and for all, you are right. To learn the work. However, says Mills, the craftsman is methodology and to do research itself requires “carried over these junctures by keen anticipa- a tolerance for ambiguity and living with some tion” of what the finished product will look like, uncertainty. It is always possible that the data we and by pride at its completion. collect may continually support our hypotheses but that our theory itself may be wrong. Or our Learning the Craft stated hypotheses may be supported by the data, but for reasons simply unknown to us. In generations past, young apprentices received The canons of science are a basic blueprint, instruction in the workshop of a master. They but to do good research, we must do more than first acquired some basic skills from books and follow their direction. The scientific method perfected the essential “hands on” techniques does not tell us whether to maximize descrip- by practicing continually. In teaching research https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 30 Theory and Method methods, the authors of this book can perform vate troubles, as workers, parents, consumers, some of the same functions. We will clarify the and voters, are in reality public issues. We must criteria that social scientists use in evaluating appreciate this point and try to get others to their work and the work of others. We will show understand it as well. In our journal we may keep you techniques for improving reliability and memos to ourselves, excerpts from books, half- validity, and we will offer you helpful suggest- baked theories, snatches of conversations heard ions as you go, step by step, through the research on the street, even our dreams. As the file begins process. When you finish, you will be acquain- to grow, it is useful to go through the entries, ted with most of the tools in the researcher’s trying to see which pieces of information seem “workshop.” to have things in common. This periodic rear- Your ability to use these tools is indispensable ranging of data itself constitutes an exercise of for the collection of data about the world, but the research imagination and frequently gener- it will not, in itself, make you an imaginative ates new ideas. We may find through such a con- craftsman or craftswoman. No rules or set of pro- tinual reorganization that certain key concepts cedures will ever replace a keen imagination in emerge and that many of our entries, previously producing a good study. Technique is important, seen as wholly independent and discrete, begin but showing off methodological expertise for its to fit into a larger mosaic. own sake is pointless. The purpose of research is One of the features of such a process is that to produce findings that add to our knowledge. we will eventually reach a point where we have generated more ideas than we could likely inves- tigate in a lifetime. We find ourselves necessar- Using Research Imagination ily setting priorities among our several ideas. In How, then, does one go about cultivating an Mills’s own words, “Any working social scientist “imagination”? Contrary, perhaps, to popular who is well on his way ought at times to have conception, creativity is not something that one so many plans, which is to say ideas, that the either possesses or does not. One can acquire the question is always, which of them am I, ought knack of being an imaginative researcher by con- I, to work on next?” (2000:198). A true crafts- tinually structuring one’s thoughts and ideas. It man or craftswoman, whether working in wood, takes work to develop this talent; how perceptive clay, paint, or ideas, is never without something an observer you are will depend more on your to do. own energy than on anything we can tell you. In his essay on intellectual craftsmanship, SUMMARY Mills (1959/2000) suggested ways of channeling mental energy. He said that ideas and problems The selection of an existing theory, or the con- for research gradually take shape as social sci- struction of new theory, is as important to social entists “play around” with concepts. It is not research as the perfection of techniques for the uncommon to begin toying with an idea and collection of data. Theory may be suggested by wait, sometimes several years, before actually taxonomies, models, typologies, and paradigms. beginning to work on it. During the interven- Most theoretical forms are compatible with the ing period, the problem remains in the back of scientific method. Creating and expanding the- the mind, and each time one reads something, ory are prime motives for research. The choice or has a relevant personal experience, a mental of theory is crucial because it may influence note is made. Part of intellectual craftsmanship not only the topic of an investigation but the is this continual reflection on ideas over time. research findings as well. Existing theoretical Mills, in fact, advocated that researchers keep models and paradigms exercise great influence a file, or journal, to serve as a reservoir for ideas. over the research enterprise, but neither these Entries in this file should continually reflect formulations, nor any guidelines for gathering one’s own life experiences. In this sense, to be social data, can substitute for a creative imagina- a scientist does not mean separating our per- tion. The individual investigator as practitioner sonal intellectual life from our career. Our pri- of a craft remains at the heart of the process. https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 References 31 KEY TERMS one week, reflect on the process of developing cate- gories for making generalizations about your activ- applied research ities. Why are you selecting these categories? What conflict theory have you left out, and why? deduction descriptive model dramaturgical paradigm SUGGESTED READINGS explanatory model Blau, Judith, and Keri E. Iyall Smith, eds. 2006. Pub- induction lic Sociologies Reader. Lanham, MD: Rowman & model Littlefield. paradigm A variety of public sociology studies are summa- public sociology rized, along with commentary on this emerging pure research method. respondent social theories Ritzer, George, and Douglas J. Goodman. 2003. So- structural functionalism ciological Theory. 6th ed. New York: McGraw-Hill. systems theory Among the most clearly written theory texts. taxonomy Spradley, James P. 1979. The Ethnographic Inter- typology view. New York: Holt, Rinehart and Winston. variable Numerous examples of different taxonomies and how they are constructed can be found on pages EXERCISES 132–154 of this book. 1. Choose some examples of research in the social Stinchcombe, Arthur L. 1987. Constructing Social sciences that attempt to offer theoretical expla- Theories. Chicago: University of Chicago Press. nation of some phenomena. For each, indicate This book compares and contrasts various types whether the theoretical explanation is arrived at of theory and shows how they are created. through an inductive or a deductive process. Then, Vaughan, Diane. 1986/1996. The Challenger say whether you feel the theories have been ade- Launch Decision: Risky Technology, Culture, quately verified by the data collected. and Deviance. Chicago: University of Chicago 2. Using the same piece(s) of research as for exer- Press. cise 1, try the following: A masterful sociological investigation of the r Indicate the assumptions that you believe accident that shocked the world and launched underlie the respective theories by creating a Vaughan’s career in public sociology. list of propositions that follow one another in sequence. r If the theory is stated only in verbal terms, try REFERENCES to create a model similar to Figure 2.2 to illus- Agger, Ben. 2000. Public Sociology. Lanham, MD: trate the proposed causal connections between Rowman & Littlefield. the variables in the theory. Alexander, Jeffrey C., ed. 1998. Neofunctionalism 3. From a textbook such as Shoemaker (2006), find and After. New York: Blackwell. two competing theories of delinquency. Indicate which theory you believe to be more reasonable Anderson, Elijah. 2003. A Place on the Corner. 2nd and why. What type of data would you need to test ed. Chicago: University of Chicago Press. each of your chosen theories? Anderson, Lorin W., and David R. Krathwohl, eds.; 4. On the basis of your personal observation, try Benjamin S. Bloom. 2000. Taxonomy for Learn- to construct a classificatory scheme for describing ing, Teaching, and Assessing: A Revision of Bloom’s the types of students on your campus. Taxonomy of Educational Objectives. New York: Longman. 5. Keep a journal in which you note aspects of your own everyday life that you believe have sociolog- Becker, Howard S. 1953/1993. Becoming a Mari- ical, psychological, or political significance. After huana User. New York: Irvington. https://doi.org/10.1017/CBO9780511819391.003 Published online by Cambridge University Press P1: JZP 0521879729c02 CUFX143/Gray 0 521 87972 8 May 30, 2007 10:21 32 Theory and Method Burawoy, Michael. 2005. “Public Sociology: Pop- Kuhn, Thomas S. 1996. The Structure of Scientific ulist Fad or Path to Renewal?” British Journal of Revolutions. 3rd ed. Chicago: University of Chicago Sociology 56 (3): 417–432. Press. Burawoy, Michael, et al. 1991. Ethnography Un- Miller, Jody. 2006. One of the Guys: Girls, Gangs, and bound: Power and Resistance in the Modern Metro- Gender. New York: Oxford University Press. polis. Berkeley: University of California Press. Mills, C. Wright. 1959/2000. The Sociological Imag- Counts, Dorothy Ayers, and David Reese Counts. ination, 40th Anniversary Ed. New York: Oxford 2001. Over the Next Hill: An Ethnography of RVing University Press. Seniors in North America. Peterborough, Ontario: Milner, Mu, Jr. 2006. Freaks, Geeks, and Cool Kids: Broadview Press. American Teenagers, Schools, and the Culture of Davis, Kingsley. 1945/1993. Some Principles of Consumption. New York: Routledge. Stratification. New York: Irvington. Newman, Katherine. 1999. No Shame in My Game. Derber, Charles. 2000. Corporation Nation: How New York: Alfred A. Knopf and the Russell Sage Corporations Are Taking Over Our Lives and What Foundation. We Can Do About It. New York: St. Martin’s-Griffin. Parsons, Talcott, Edward A. Shils, and Neil J. Duneier, Mitchell. 1999. Sidewalk. New York: Farrar Smelser. 1962/2001. Toward a General Theory of Straus & Giroux. Action: Theoretical Foundations for the Social Sci- Ehrenreich, Barbara. 2001. Nickel and Dimed: Or ences. New Brunswick, NJ: Transaction. (Not) Getting by in America. New York: Metropoli- Putnam, Robert D. 2001. Bowling Alone: The Col- tan Books. lapse and Revival of American Community. New Gamson, Joshua. 1999. Freaks Talk Back: Tabloid York: Simon & Schuster. Talk Shows and Sexual Nonconformity. Chicago: Rathbone, Cristina. 2005. A World Apart: Women, University of Chicago Press. Prison, and Life Behind Bars. New York: Random Glaser, Barney, and Anselm Strauss. 1967. The Dis- House. covery of Grounded Theory. Chicago: Aldine de Ritzer, George, and Douglas J. Goodman. 2003. So- Gruyter. ciological Theory. 6th ed. New York: McGraw-Hill. Goffman, Erving. 1959. The Presentation of Self in Sabato, Larry J., ed. 2006. Divided States of America: Everyday Life. New York: Doubleday. The Slash and Burn Politics of the 2004 Presidential Gouldner, Alvin W. 1970. The Coming Crisis of Election. New York: Pearson/Longman. Western Sociology. New York: Basic Books. Scott, Marvin B. 2005. The Racing Game. Chicago: Harper, Charles L. 2006. Exploring Social Change: Aldine de Gruyter. America and the World. 5th ed. Upper Saddle River, Shoemaker, Donald J. 2006 Theories of Delin- NJ: Prentice Hall. quency: An Examination of Explanations of Delin- Hays, Sharon. 2004. Flat Broke with Children: quent Behavior. New York: Oxford University Women in the Age of Welfare Reform. New York: Press. Oxford University Press. Spear-Swerling, Louise. 1997. Off Track: When Poor Herman, Andrew. 2000. The “Better Angels” of Readers Become “Learning Disabled.” Boulder, CO: Capitalism: Rhetoric, Narrative, and Moral Identity Westview. Among Men of the American Upper Class. Boulder, Strauss, Anselm L., and Juliet M. Corbin. 1998. CO: Westview. Basics of Qualitative Research: Techniques and Pro- Hopper, Columbus B., and Johnny Moore. 1990. cedures for Developing Grounded Theory. 2nd ed. “Women in Outlaw Motorcycle Gangs.” Journal of Thousand Oaks, CA: Sage. Contemporary Ethnography 18 (4): 363–387. Trefon, Theodore, ed. 2005. Reinventing Order in Knipfel, Jim. 2000. Slackjaw: A Memoir. New York: the Congo: How People Respond to State Failure in Berkeley. Kinshasa. London: Zed. Kozol, Jonathan. 1996. Amazing Grace: The Lives of