-----------------------------------------------------------------------Dissertation Craftsmanship -----------------------------------------------------------------------Charles T. Kerchner © Claremont Graduate University Revised Spring, 2005 This short paper is written to aid students in planning and executing their dissertation research. There are parallels to other kinds of research, such as institutional research or grant and contract work, but dissertations are a particular research form with special requirements that deserve to be addressed directly, and thus I leave to the reader the determination of whether the protocols discussed here are useful in other areas of research. Much of what follows can fairly be called folklore; I will try to explain what I think research is and how one can go about it and retain at least partial sanity. (Earlier versions of dissertation craftsmanship included guides to standard designs of research, some checklists for evaluating your research ideas and plans, and bibliographic references to some of the classics and gems in the field. These have become dated and this section is being revised.) -----------------------------------------------------------------------What is a Dissertation? Dissertations have many purposes. Sometimes they are a person's most significant academic work, an important critical statement or seminal academic contribution. But more often they mark the transition to scholarship. Even when they don't save the world, good dissertations stand as lively and interesting products, and more importantly their authors carry with them new skills and insights as a result of the process. The authors say, "This changed my life," and smile when they say it. Some also become extremely useful guides to policy and practice, but to insist that dissertations have immediate and dramatic application is often to ask too much. At the core, a dissertation is a demonstration. Dissertations, first of all, demonstrate scholarship. Different writers and professors explain it variously as a "contribution to the field" or "the discovery of new knowledge" or "the building of theory" or the "testing of an idea," all of which suggest that the writer has gained an appreciation for scholarship and the ability to discern and create it. Second, dissertations demonstrate research skills. The way in which a dissertation research project is carried out tells the committee (and often future

prospective employers) what level of technical or artistic skill a student possesses and how well organized he/she is. A dissertation in which there are serious mistakes in research design or one that took five years too long to complete makes an eloquent statement about the author. Third, and most fundamentally, dissertations are about intellectual craftsmanship. They are about creating a series of thought processes by which the author comes to engage and make sense out of the world. One's conceptual lenses get ground. In education, this process is probably more difficult than in most other areas of study, for scholarship in education does not rest on a single discipline. Although the vast majority of educational studies have roots in psychological thought and method, many people who consider themselves educators carry with them the ways of thinking of economists, political scientists, sociologists, anthropologists, and a few mathematicians or attorneys. Importantly, each has formed a way of looking at the world during graduate studies. -----------------------------------------------------------------------When Should I Start to Think About a Dissertation Topic? Right now. There are real dangers in prematurely deciding on a research topic for it is to close one's mind early in graduate work to intriguing and interesting experiences that would lead to professional and intellectual growth unimaginable before. However, there is a greater danger of not training oneself to think as an inquirer, a researcher. Trying to put things in researchable terms is a very good exercise throughout graduate work. -----------------------------------------------------------------------Selecting a Research Topic This is the hardest part. Students often spend frustrating weeks or months trying to decide on a dissertation topic. The effort, though not the frustration, is in most cases well directed. Trying to find the right problem and to define it is worthy of great consideration, and a well-chosen problem makes the process and the results more worthwhile. Some questions to ask yourself: 1. Can the topic be expressed in terms of a concept or idea (adaptation, social exchange, conflict, maturation), and can I express the research questions in conceptual terms? 2. Do I know enough about past research in this area to relate my proposed study to those that have been done in the past (is it an extension of the work

done by Jones and Smedley, a different approach to the same problem, a verification, a different population)? By the way, never tell a professor that no one else has ever done work on this subject. 3. Can you write a brief statement of the problem incorporating its linkages to past research? 4. Can you describe the significance of the problem with reference to: a. b. c. d. e. f. g.

its timeliness relationship to a practical problem relationship to an important audience filling a gap in the research allowing generalization to broader principles of social interaction creating a new instrument for observing or analyzing data allowing access to a difficult-to-study population.

5. Is it possible to describe your problem in terms of hypothetical statements? In the simplest terms a problem is a question that asks about the relationship between two or more variables, the answer being what is sought in the research. An hypothesis is an operational statement of a problem that carries clear implications for testing the presence or absence of the conditions being investigated, e.g., time spent studying increases amount learned. Not all studies are amenable to hypothesis testing, but hypotheses are very powerful guides to the researcher because once formed they tell the researcher what to do; they define the research. 6. Do you have an idea about how to go about doing the research? At the point of sorting out research topics, one is not expected to have a completed plan of research in mind, but one of the tests of a good research question is whether one can think of some way to make it operational. 7. Does the research seem feasible? Ask yourself: Do I think that I have the resources to do this? Can I gain access to the persons or information necessary? Can I finish the work in a reasonable amount of time? (A good rule of thumb for a tight dissertation of good quality is that it ought be capable of completion in a year—from proposal to final draft and defense—by a person working at it substantially full time.) 8. Does the faculty have the necessary expertise to help me with a study in this area and if not do I have access to "readers" from other institutions that would help direct my research? Whereas CGU faculty tend to be generalists, it is important not to ask chairs to serve in areas where they have little experience. It's not fair to them or you.

9. Will this dissertation allow you to avoid working with people you don't like? My dog thinks I'm great all of the time; my wife and children most of the time. After that the percentages go down. Don't do your dissertation with someone you can't stand. Conversely, don't be off put by strangers. If there is a faculty person whom you think has interests similar to yours or whose style of research you identity with, then spend some time talking with that person about your dissertation ideas. The time in exploration is well worth it. 10. Can you visualize problems that are likely to come up in the research and what you would do about them? 11. Is this an area of genuine interest for you? Ask yourself if you would be happy doing work in this area for 5 years. Dissertation research often turns into an occupational specialty. 12. Is this consistent with your career plans? If the answer is "no" and you feel a compelling pull toward the research topic, then maybe you have the wrong career plans and the right research topic. There is no single technology of research idea gathering. Some people's minds center on neat, highly researchable, topics. Most don't. One of the best ways is to read other research in the area of potential research. Don't read about research, read the research itself. Ask yourself what are the conceptual underpinnings of the research, how much you know about the content area involved, and whether you could do a project like the one undertaken in the research you are reading. Another possibility is to attempt to write the topic. See if it can be expressed in a sentence or a paragraph. When this is done to the point you don't cringe too badly at your own words, try discussing your idea with other students, faculty you think might be interested, and the types of people who might be your research subjects if you actually did the research. Don't be afraid to discard or modify. The most direct way to test your ideas about design is to wade into the field you plan to study. If you plan to study the job mobility of college deans, go talk to half a dozen. If you plan to study how teachers allocate their time, invest a day sitting in the back of a classroom. If you want to study disadvantaged youth, go spend several afternoons at a sandwich stand or doing yard duty at an elementary school. Most students tend to make their research designs neater than the world they are attempting to investigate. Academics, by the way, aid and abet this problem. Their journal articles and research reports make it seem that the entire research process was organized and conceptualized in one bite of the mind. It's very seldom the case. -----------------------------------------------------------------------The Research Proposal

The research proposal is the rock upon which you build your work. It defines what you are going to do in operational terms, and it helps make clear the faculty's expectations of you. All dissertations depart from the dissertation proposal, some in substantial degree, because the researchers found that for one reason or another parts of the research needed to be modified. But having a clear proposal is of great importance. It gives you something to stand on if you get in deep water later. Generally proposals should be about 20 pages long and should contain the following: 1. A brief statement of the problem in operational terms. This is the honed and polished version of the problem statement you've been working on. (1 page or less) 2. A statement of the importance of the research to theory, practice and policy. This statement must show the importance rather than just assert or claim it. (e.g., this research involves a reinterpretation of the concept of alienation.) (2 to 3 pages) 3. A brief review of the literature and its application. This section has two very specific and important uses, neither of which has anything to do with impressing your committee with the fact that you have read a lot of books in the topic area: (1) it establishes the relationship of your research to the concepts you are using, and (2) it establishes the relationship between your work and the empirical research of others. (5-7 pages) 4. A statement of specific objectives and hypotheses. (1-3 pages) 5. A description (in 7 to 10 pages) of how the research will be undertaken including: a. a description of the population, the study site(s) or samples and the means used to choose the sample. b. a list of the key variables and how they will be operationalized. c. a discussion of the research design--exploratory, descriptive, experimental, and the strengths and drawbacks to that design. d. a discussion of how one will acquire the data, and if primary data are to be gathered the feasibility of questionnaire, interview, observation or whatever methods are to be used. Why this means of collecting data is preferable to challenging methods. (Draft copies of any instruments should be included in the Appendix of the proposal.) e. a discussion of the statistical or other analytical methods that will be used. Include blank charts or tables that you expect to represent your output.

6. A time table for the dissertation research showing both the chronology of the research and the number of person days involved in each step. (1 page) --pilot study and pretests of instruments --gathering field data or getting secondary data in workable form --preliminary analysis --contingency plans for return to gather additional data if problems appear with the first --analysis of the data --writing the report --review by chair and committee --revisions --review of the final draft --proofreading and editing --final approval by chair and committee --oral examination --rewriting as suggested by the dissertation committee --submission of final copies -----------------------------------------------------------------------The Institutional Review Board Like most research organizations, CGU maintains an Institutional Review Board that examines all proposed research, including dissertation proposals, against the standards for protecting human subjects. IRBs began because of abuses of subjects in medical research, and they have grown to consider all kinds of risk that subjects may encounter, particularly psychological risk. Research involving children is particularly scrutinized. In recent years, our IRB and its procedures has become more user friendly. Forms and instructions are available on line, and the staff can assist you. Each student must prepare an IRB form, which is then submitted to the dissertation chair for signature before it is sent to the review board. -----------------------------------------------------------------------The Dissertation Outline Opinion about the architecture of dissertations varies from professor to professor, and indeed from topic to topic--and with some of us from moment to moment. Here is my best sense of a standard personal preference. Chapter I has two primary functions that make it both the easiest and hardest chapter to write. It is an introduction in which the goals of the research are stated, the background of the problem sketched and the research itself is introduced and its importance established. That's the easy part, and it directly

flows from the dissertation proposal. The second and more difficult function of Chapter I serves as a net to capture the reader, to summarize enough of the research or telegraph its direction so that the reader will be able to understand the data and literature that you will be presenting in the following chapter. This requires asking yourself: "What facts or data does the reader have to understand first in order to get ready to read chapters 2-5?" Suppose you were doing a dissertation on equity in education involving the social regard that teachers had for students of low SES and high. You had drawn a sample, reviewed the literature on social modeling, scored teachers on the amount and quality of attention that they gave different students, and ultimately come to the conclusion that teachers were not systematically biased against poor children but that variation was a function of the teacher's attitudes. You conclude that one of the ways to redress the problem of inequality in teacher perception would be to let children and their parents choose the teachers with whom they studied, and so you advocate a voucher system to change schools. Now, that's traveling a lot of territory. You need to get the reader ready to understand your conclusion that moves from a psychological concept, "social regard," to a conclusion involving school finance and structural reform. The reader has to understand the connection: What is really going on here? What is going on is that you are engaged in an expansion of the concept of equity as applied to school policy. Most equity studies have been concerned with sameness in the provision of resources, largely dollars, between school districts and sometimes within them. What you are asserting is that human regard and status is a resource. The reader has to understand that first before it is possible to make the conceptual turn to your conclusion. Once the reader understands that status is considered as a resource, then he or she is prepared to accept a line of reasoning that suggests that no amount of in-service education or human development workshops can be expected to inure all teachers to poor children, but the best defense of poor children would be to arm them with the ability to seek those teachers who did apparently get along well with poor students. Students often try to write Chapter I first, and sometimes it helps them create a functional outline for the dissertation. More frequently, however, it is easier and more functional to write Chapter I last after one has thought through and written the rest of the dissertation. Chapter II is a review of the relevant theory and empirical literature whose function expands upon, and is parallel to, that of the same section in the proposal. It presents the argument of the dissertation. Chapter III is a presentation of the methods employed in the research: everything from the proposal plus a more detailed presentation of the analytical techniques used, why they were preferable to others, and explaining any unusual or problem circumstances encountered and how they were dealt with. Also

include a brief description of the sample researched. Copies of any instruments should go in an Appendix. Chapter IV contains the presentation of the data and the analysis thereof. There is great variation in these chapters depending on the research style used, an experimental design looking a great deal different from an ethnography, but the functions are the same. The first job is to get the reader to understand the nature of the sample. In descriptive studies, basic descriptive statistics are introduced. In an ethnography, one would describe the institution, the people in it, those aspects of the environment that will prove important later. The second function is to address the specific research questions that were introduced in the proposal and in Chapter I of the dissertation. this is where you really show your analytical stuff. In an ethnography it might involve showing how the sample divides into two belief systems over a particular topic. In a descriptive study it might involve the use of discriminate analysis to show that votes on certain types of issues separated Democrats from Republicans. (In some dissertations it is appropriate to have more than one chapter of data presentation. Ask your chair.) Chapter V should include a discussion and an interpretation of the meaning of these results. It should begin with a summary, and end with recommendations for additional research or for policy and practice. In the Appendices include copies of instruments used and tables and charts too lengthy and tangential to be included in the main report, some descriptive statistics or interview transcripts for instance. -----------------------------------------------------------------------Submitting Chapters of the Dissertation There is a great variation in the working arrangements that exist between dissertation chairs and students working on their research, and thus the reporting arrangement that follows is much more a guide than a rule: 1. Immediately after the proposal is written (if not during the writing process), undertake a pilot study of the dissertation research. Modify the instrumentation according to the results of the field experience. 2. Submit Chapter II as revised, the argument of the dissertation. 3. Submit Chapter III as revised, the methods employed. 4. Report on the status of field research gathering about half way through the process. 5. Turn in a rough draft of Chapter IV, the data and its analysis.

6. Revise Chapter IV. 7. Write Chapter V, the conclusions. 8. Confer over the internal flow and fit of the dissertation taken as a whole. 9. Write Chapter I. 10. Submit whole dissertation and prepare for oral defense. In the Appendices include copies of instruments used and tables and charts too lengthy and tangential to be included in the main report, some descriptive statistics or interview transcripts for instance.