DISSERTATION OUTLINES

G. David Garson

 

Forming an Outline

Forming a workable paper, thesis, or dissertation outline is half the battle of writing. An outline is part of the roadmap which tells the reader where you are going with the story you are attempting to tell. It is also an important tool in your own reasoning process where establishing the foundation for your research and eventual written work. A bad outline will lead you down endless mazes and yield an unreadable final product. A good outline will ease and encourage you along the path to victory in your chosen project.

Of course, it is difficult to write an effective outline without having a clear idea about the content of your writing project. Often your outline will include much of the generic sample below. Be prepared to write a paragraph or so about each point. It is often advisable to leave the introduction and the conclusion to last. By writing a paragraph or more under each outline topic the researcher is carrying forward into the outline the results of the brainstorming process discussed in Chapter 1 and efforts to frame an analytic topic discussed in Chapter 2.

:


Sample Outline

Title page

Acknowledgments page

Abstract page, with keyword list

Table of contents

Table of figures

Introduction

Statement of the research problem

Contextual background

Theoretical relevance

Policy relevance

Research statement

Research questions

Hypotheses

Definition of terms

Body

Literature review                                                                                  

Differentiation of planned research from existing literature

Methodology to be used

Research design

Population, sampling, and data acquisition

Instrumentation, including assessment of validity and reliability


Qualitative and/or statistical procedures                        

Hypotheses

Hypotheses related to theories and models

Operationalization of variables

Preliminary evidence on relationships

Evidence with use of control variables

Consideration of alternative models

Conclusion/Discussion

Scope of generalization, limitations, delimitations, and threats to validity

Summary of findings, related to previous research/theory

Implications for theory and/or practice

Future research agenda

Endnotes

References


The sample outline above is a standard outline. It flows in a logical order from introduction to evidence to conclusions. However, it should be noted that on rare occasions the researcher may require a hierarchical outline. The hierarchical outline has a tree-like structure which branches into finer and finer categories, ultimately ending up with Aleaves@ which contain conclusions pertinent to that path along the tree. Such an outline might be appropriate, for instance, when the purpose is taxonomic (placing observed phenomena into an exhaustive set of categories). Field guides to identification of plants, rocks, or celestial phenomena are often arranged hierarchically, based on empirical but non-statistical observation. A statistical  methodology which organizes data into hierarchical trees is inductive expert systems, which rely on such algorithms as ID3 from information theory. In general, however, hierarchical outlines are rare in quantitative writing and the researcher is well advised to stick to the standard outline unless there are compelling reasons not to.

 

3. Standard Outlines                                                                            

Each of the elements of the typical standard outline just described will now be explained in more detail. Keep in mind, however, that the format discussed below is a template, not an iron code of scientific investigation. Some sections may be inappropriate for some types of research and in other cases additional sections not included below may be needed. Moreover, the outlining process is often best undertaken when not conceived as a before-the-fact initial writing effort but as an ongoing process wherein the researcher continually goes back and reconsiders the logic and story-line of his or her writing and redrafts the outline accordingly.

Standard Outline Elements

Title page


Your  title should include critical key words (ex., your dependent variable). The words in your title have an important function in electronic reference retrieval and may influence the impact of your work on your discipline. Strip off unnecessary words like "A study of ..." and any redundant words. If your work is later published in book form, then you may entertain a "sexier" title with a more catching but cryptic nature, but avoid this temptation in academic work. Your academic department or publisher may well supply rigid guidelines, but in general the title page typically includes the full title of your dissertation, thesis, or paper; your full name; the degree, course, or publication for which the work is submitted; and the date of submission. If contact information is not on the title page, it should be on a subsequent separate page: name, address, telephone, e-mail, fax, and web address, if any.

Acknowledgment page

Be sure to explicitly acknowledge anyone who has helped you with the substance or mechanics of the research. This will include your advisor and members of your committee. If you have benefitted particularly from the ideas of one or more persons, acknowledge this. Acknowledge everyone who read your drafts and gave you feedback. It is better to err on the side of having a lengthy acknowledgment page.

Abstract page, with keyword list


This page is more important than it may seem. It sets the tone for and expectations of your readers. Moreover, in this era of electronic indexing what you put here will be widely incorporated by various search and bibliographic services and in Dissertation Abstracts in the case of dissertations. Your work may even be cited by others on the basis of the abstract alone. Supply a keyword list of your choosing in order to prevent clerical people in reference service organizations from choosing your descriptors for you. About 150 to 300 words is typical. It need not be limited to a single paragraph (though that is common) but may extend to two or three as necessary. Do include your actual main findings, not just a description of the areas your research covers. It is customary to put verbs in the past tense, to reflect research already completed, in contrast to abstracts in the proposal stage of writing.

Table of contents

The table of contents will follow in a standard outline. Of course, the outline discussed here, while typical, is only one of a great many variations on a theme. Whatever variation you choose, do include page numbers. Hint: learn how to get your word processor to compute these for you automatically -- to do so you will have to enter markers as you type in chapter, section, and subsection titles, so learn this before you start typing.

Table of figures

Have separate listings for figures, charts, and tables. As with the table of contents, learn before you start to write how to get your word processor to generate these listings automatically.

Introduction


The introduction gets disproportionate reader attention and deserves polishing and re-polishing by the writer. Keep in mind that what you say at the beginning is apt to be the standard by which you are judged at the end. Typically, the introduction contains a statement of the research problem, discusses the problem's context in terms of practical policy and/or academic theory, and sets forth a statement of the proposed research, to be elaborated upon in subsequent sections. The introduction is also the place where the writer clearly identifies the audience for the writing, and indicates whether it is assumed that that audience is already cognizant of certain knowledge pertinent to the research at hand  B  information which, for a different audience, might otherwise have to be set forth in the paper, thesis, or dissertation.

Statement of research problem

The statement of the research problem is where the researcher sets forth the subject and makes the case that it merits the reader's interest. Without overburdening the researcher with too many details to be presented in subsequent sections, the statement of the research problem should indicate the general nature of the hypotheses to be investigated. It should also demonstrate that these raise questions about which researchers might disagree -- that there are policy and/or theoretical issues involved. That is, the statement of the research problem defines the scope and general direction of the proposed research and presents the motivation for interest.

Contextual background


The section on the context of research is not the literature review section though it is directly related to it. Without reviewing the literature in detail, in the contextual background section the researcher makes a fuller presentation of the objectives mentioned in the problem statement. That is, here there is a fuller presentation of the theoretical and practical importance of the research. However, the presentation is still a summary and not the full literature review. Often, after completing the literature review, the researcher may be in a better position to refine the contextual background section..

Theoretical relevance

While purely policy-oriented writing is common and important, if at all possible the researcher should relate the proposed research to theory reflected in the literature of the field. This prefigures and frankly is a little redundant with the literature review, but in the section on theoretical context the researcher is only hitting the highlights and main points. If there is no such theory, of course, the researcher is well advised not to force an inappropriate framework onto his or her topic simply for the sake of "being theoretical." However, there is also an opportunity here to set forth original typologies, frameworks, and theories of one's own.

Policy relevance

While purely theoretical writing is quite acceptable, if the research has policy implications these should be highlighted here. This prefigures and is somewhat redundant with the findings and conclusion sections. However, in the section on policy context, the researcher is only suggesting how findings of the research at hand may relate to existing policy debates or existing or proposed legislation . If there are no policy implications, of course, the researcher should not force an inappropriate framework onto his or her topic simply for the sake of "being relevant."


Research statement

The research statement is a relatively short concluding section to the introduction in which the researcher recapitulates the research problem and its theoretical and policy importance, and then summarizes where the rest of the dissertation or article is going, summarizing the general nature of the literature review, types of hypotheses, data collection strategy, and methodological procedures to be invoked.  Key terms are also defined at this point. In essence, the research statement is an extended abstract of the entire body: some redundancy is expected but comprehensive treatment is avoided in the research statement.

Body

A good introduction has given the reader a beginning-to-end overview of the entire work. The body goes back and starts at the beginning again but presents each element in much greater detail. In writing the portion of the outline which forms the body the researcher may find it helpful to write out one sentence (not just a heading) for each of his or her main points. Although headings are the type of outline many used in their high school days, outlining using full sentences is much less ambiguous and makes for clearer argumentation.

Literature review


It is easy and common to write bad literature reviews. Bad reviews lose the reader in details and irrelevancies, give the impression the writer is meandering, and tempt the reader simply to skip ahead to the section on methodology. Good reviews relate the literature to the hypotheses section which will follow and keep the "story line" of the work in the forefront. This requires omitting much perfectly fine and hard-gotten literature research which, while interesting and important in other contexts, is simply not very relevant to the research questions at hand. In general, it is a bad idea to arrange the literature review in chronological order or in journalistic order (the inverted pyramid from most important to least important). Reading the literature will help the researcher refine his or her hypotheses, which in turn points deeper into the literature, and so on in an interative process. After a few iterations the researcher should know what hypotheses are to be investigated, and the literature should be organized according to clusters of hypotheses. Formal hypotheses are not presented in the literature review, but it is still possible to organize in parallel. For instance, later one set of hypotheses may deal with economic independent variables and another set with educational variables, so the literature review would have sections paralleling this.

Differentiation of planned research from existing literature

In this section the research makes explicit how his or her proposed research confirms, disproves, qualifies, extends, or is an original addition to the literature of the discipline. These are not mutually exclusive categories as, obviously, the research may confirm one set of theories, disprove others, or provide the basis for synthesis of seemingly contrary findings in the existing literature.

Methodology to be used


The methodology section is particularly crucial in doctoral dissertations. Often this section is the one most thoroughly discussed with one's doctoral advisor and committee. The researcher does not want to invest substantial research time when he or she lacks agreement on the methodology to be used. There is no one perfect methodology. Strong consideration should be given to a multi-trait, multi-method approach supplemented with qualitative research. That is, use multiple indicators for each variable, confirm the results of one method (crosstabulations based on survey research, for instance) with those of another (ex., regression models based on archival data), then increase validity of one's findings by supplementing with qualitative research examples (ex., case studies).

Research design

This section sets out the research design, whether experimental or observational. Analysis of variance alone has dozens of research designs but the objective of all designs is to come as close as possible to controlling for all variables which might reasonably be thought to influence the dependent variables being investigated. For this reason, the research design section must highlight the treatment of potential control and modifier variables which may affect the main relationships of interest.

Sampling and data acquisition


In this section the researcher describes the types of data which are to be collected. Often, random sampling protocols are discussed here. Response rates should be reported along with a demographic profile of one's sample and, if available, of known characteristics of the population. Nonresponse should be analyzed, as by wave projection methods. One should also describe quality control procedures such as tests of inter-rater reliability, triangulation of respondents, recoding of recorded interviews, and so on.

Instrumentation, including assessment of validity and reliability

More often than not, previous researchers have examined the same variables which are the focus of the proposed research. Sometimes scales have been developed to measure a concept, have been validated in large populations, and are accepted instruments in the profession. In this section the researcher discusses past measures and tests of their reliability and validity. If using original measures, the researcher should discuss validity and reliability concerns and tests related to them. If there is a separate section on operationalization of variables, as in this example outline, specific items are not discussed in this section.

Qualitative and/or statistical procedures

Whatever statistical procedures are invoked, all of the assumptions of the selected procedure should be evaluated in relation to the proposed research. Where appropriate, tests of assumptions (ex., tests of multivariate normality, homogeneity of variances) should be reported. Nearly all research reports significance levels of findings, but in doctoral research it is also important to discuss statistical power and alpha (significance level) reduction procedures when, as is usual, multiple tests are conducted. In addition to assessment of significance of relationships, the effect size (magnitude) of relationships should also always be reported.


Hypotheses

This section is the kingpin of the entire dissertation or article structure. Care should be taken to make sure there is parallelism between the hypotheses section and the earlier introduction section's research statement, with the literature review, and with the conclusion.

Hypotheses related to theories and models

This section is where one formally lists the research hypotheses, grouping and introducing them in relation to the theoretical model one is testing. "Formal" means that hypotheses are expressed in terms of their statistical test, which is often a null hypothesis (ex., introduction of community policing in a police district will not affect the burglary rate). One should also state the alpha (significance) level that will be used for testing hypotheses (ex., .05). Though optional, it is desirable to state the expected direction and general strength of association (weak, moderate, strong) expected on the basis of one's theory or model.

Operationalization of variables


The list of hypotheses gives the list of variables involved in the research. All should be listed and formally defined. Ideally, each variable should be measured by multiple (ex., >3) indicators. The researcher should discuss each indicator in terms of whether it is an item with intrinsic high reliability and validity (ex., gender); a component of an accepted, validated instrument (as discussed in the instrumentation section above); is an indicator whose reliability and validity has been confirmed by the author (discuss how); or is an indicator whose reliability and validity has not been tested. Actual instruments containing all items, along with full coding instructions, should be reproduced in an appendix.

Preliminary evidence on relationships

The first task of the researcher is to determine if any relationship exists connecting the primary independent and dependent variables of interest. In graphical terms, the researcher asks for each arrow in the model, "Is the relationship significant, in the expected direction, and of the general level of magnitude predicted?"  Some authors separate a "Results" section from a subsequent "Findings" section with data in the one and conclusions in the other. Unless required by one's department or journal of publication, however, this practice tends toward confusion for the reader and is deprecated here.

Evidence with use of control variables

If the originally posited relationships are upheld by the evidence, the researcher must then consider for each relationship in the model if there are additional variables (hopefully ones which have been anticipated and for which measures have been taken!) which modify the relationship because they are intervening or common anteceding variables. Even relationships which are statistically insignificant overall may have a significant relationship for some ranges of a modifier variable.

Consideration of alternative models


The researcher must assume that his or her data will support more than one model. Plausible alternative models should be investigated to determine if they fit the data as well as or better than the initial model. These alternatives may be suggested by the literature, by the researcher's original judgment, and by statistical coefficients based on the original model (ex., modification indexes in structural equation models). The researcher must also balance tests of model fit against the preference for model parsimony. Simpler models may be preferred to more complex ones unless the fit of the latter is substantially better.

Conclusion/Discussion

The conclusion should be fully consistent with the introduction and should flow from the presentation of evidence. The central purpose of the conclusion is to show how the researcher has completed the tasks set forth in the introduction. There is a tendency among researchers to feel that on completing writing of the body, the conclusion is a minor effort to be tacked on. Avoid this viewpoint as the concluding section is apt to be the most-read, most-influential section of the work. Moreover, the evaluation of the work will have much to do with the consistency of the concluding section with the introductory, methods, and literature review sections which came at the beginning of the work. Writing the conclusion is a major effort to which the researcher must allocate significant time. Note that this section is usually labeled "Conclusion" or "Discussion," not "Summary."

Scope of generalization, limitations, delimitations, and threats to validity


The most common error in writing the conclusion is making generalizations which are not supported by one's data. The researcher should remind readers, and him- or herself, of the exact boundaries of generalization. Delimitations are boundaries set by the researcher him- or herself for reasons of scope, while limitations are boundaries inherent in the nature of the data. For instance, if one has random sample data on Arizona city managers then one does not want to generalize to American city managers unless one has demonstrated those in Arizona are similar in important respects to national averages for city managers. In fact, all limitations and threats to validity should be recapitulated at this point. The limits may be extensive enough that the researcher should label findings exploratory or preliminary results. One should err on the side of humility.

Summary of findings, related to previous research and theory

This section is the culmination of the article, thesis, or dissertation. It is an extended summary of the work's story line, relating each main point to the evidence presented in the body of the work, with brief mention of relevance to previous research and theory discussed in the introduction and literature review.  Overall, the discussion section summary need not be comprehensive of all findings in the body of the study. Rather its purpose is to highlight and consolidate the story line of the thesis. No new findings should be presented here -- all such analysis should have been in earlier sections. The researcher should address whether findings are supportive of the thesis in effect size as well as in significance and direction. The researcher should also discuss findings inconsistent with or only partially supporting his or her thesis.


It is important in this section to "stick to the facts," avoiding the often-strong temptation to generalize beyond what is actually proved by data in earlier sections. In this discussion section summary the researcher should keep in mind that the findings have already been presented in the body of the paper, thesis, or dissertation. Therefore, while a summary in the discussion section is needed and appropriate, it should not be unnecessarily redundant. Instead, the discussion section summary is just that -- discussion -- and leads directly into the next section on implications for theory and practice.

Implications for theory and/or practice

In this section the researcher discusses at greater length the significance of the work's findings for confirmation, modification, or rejection of current theories and/or practice related to the subject at hand. In this section the researcher will hark back to the review of the literature section, showing how his or her study supports, extends, revises, or disconfirms earlier literature in the profession. The more applied the thesis, the more this section will focus on implications for professional practice in the domain studied.

Future related research agenda


It is common to end the conclusion with a discussion outlining the main elements of a future research agenda growing out of the present research. These are suggestions directed toward the discipline, and are not necessarily commitments of the author for future work. Such a section is most effective if it elaborates on one, two, or three research agenda elements most strongly associated with the researcher's own work. Such elements are typically logical and direct next steps rather than long-range research needs. However, a rationale should be apparent why such next steps were not incorporated into the researcher's own work in the current paper, thesis, or dissertation. At the same time, a laundry list of a large number of possibilities for future research should be avoided as it distracts from the "story" of the current work and diffuses the actual helpfulness of the future research section as a whole.

Endnotes

The "Endnotes" section is not for citations, which go in the "References" section. Rather, "Endnotes" is used for elucidating comments and providing relevant supplementary information whose insertion into the body of the text would be distracting to the story line of the work. They are listed automatically by number by most word processing programs, which will automatically renumber when additional endnotes are inserted in the middle.

References


A "References" section lists only references cited in the work, as opposed to a "Bibliography," which lists all references consulted whether cited or not. Most publishers and academic departments do not want a bibliography. Your publisher or academic department will very likely specify the reference format style to use. One of many advantages of using bibliographic software, discussed in a separate section of this guide, is that with almost no effort you can generate the references in one format for your academic department and in any number of other formats according to publishers' specifications. As a default, the APA reference style used by the American Psychological Association is probably the most widely used.

Bypassing the outlining process may well lead to a disorganized paper, thesis, or dissertation. Outlining forces the researcher to clarify things in her or his own mind. It forces the researcher to consider the logical order of his or her argument and whether there are portions which are missing, redundant, or irrelevant to the arguments that are being made.  The outlining process may prompt the researcher to conduct additional investigation to fill in holes in the outline or may prompt him or her to throw out material which is Aleft over@ and has no proper place anywhere in the outline.

.

Outlining Checklist

* Have you examined some examples of outlines for similar dissertations, theses, or articles, to determine what seems most applicable to your own context? Hint: look at Dissertation Abstracts Online, available on Dialog Information Services through most reference libraries (see http://library.dialog.com/bluesheets/html/bl0035.html).

* Have you obtained guidelines from your academic department or publisher so as to assure your outline conforms to stylistic requirements which vary from discipline to discipline?

* Does your title contain key words which form a "mini-abstract" of your paper, thesis, or dissertation, and which make clear what your dependent variable is?

* Do you avoid abbreviations in your title?


* Does your abstract include all relevant key words for electronic search purposes?

* Does your abstract include your main findings, not just an indication of the area you are studying?

* Does your outline start with a description of a problem to be studied and a clear explanation of why the problem is important in the profession, in theory and/or in policy terms?

* Have you clearly identified your target audience (ex., practitioners, theorists, agency policy-makers)?

* Have you avoided giving equal weight to main and supporting ideas in your outline?

* Have you outlined the methodology you propose to use to investigate the issues you have selected, and have you demonstrated how triangulation (a multi-method approach) might enhance your research design?

* Is your outline arranged in a way that the story line stays in the forefront, and the story builds to a climax in the conclusion?

* Would an impartial reader understand the logic of your research from your outline (give it to some friends and test this out!)

* Do you follow familiar organizing principles, such as arranging your coverage in  subsections from general to specific or from abstract to concrete?

* Does the content correspond to your heading levels in your outline, such that the higher the heading level, the more important the topic? Or are important, key parts of your story line hidden away in obscure subsections of your outline?


* In associating content with heading levels, have you observed the principle of coordination (that similar levels of generality should have similar heading levels). For instance, if outline level AA@ is AReproductive System@ and AB@ is ANervous System,@ it would be a violation of coordination to let AC@ be AThe Heart,@ Rather, AC@ would be ACirculatory System@ and AThe Heart@ would have to be a subtopic of AC.@

* Does your outline serve as a summary of your work?

* Will your conclusion correspond to your introduction?

* Have you let the evidence speak for itself, using a professional tone which avoids normative or dramatic adjectives like "unprecedented," "obvious," "overwhelming," or "striking"?

* Have you given space to discuss contrary, partially supportive, and negative findings as well as findings in support of your thesis?

* Have you arranged for peer review feedback for your outline?           

* In terms of the sheer format of the headings in your outline, have you observed parallelism? That is, are headings at the same level expressed similarly in similar terms (ex., you might use terms like campaigning, voting, and electing; or you might use Ato campaign,@ Ato vote,@ or Ato elect,@ ; or you might use these terms as adjectives -- but you would violate parallelism if you Amixed and matched@ types of grammatical form).

 

c. Outlining Examples                           


Numerous sites on the World Wide Web provide examples of good outlining However, as content on the web changes so rapidly one may be better off simply to enter terms like Awriting|outlining,@ Aoutlining skills,@ Astandard outline,@ or Ahierarchical outline@ into a web search engines.

Past award-winning dissertation titles provide good examples of outlines that may well be worth examining. Abstracts are provided through Dissertation Abstracts Online, a service of Dialog and other vendors, available through most research library reference departments (see http://library.dialog.com/bluesheets/html/bl0035.html). The dissertations themselves are available free in many cases through your research library's inter-library loan department and are almost always available for a fee through Proquest (formerly University Microfilm International. UMI) at http://www.proquest.com/brand/umi.shtml. This supplies full text of dissertations in Postscript Document Format (.pdf), which one can download online, for a typical cost (at present) of about $20/dissertation, depending on length. The .pdf format will require you have Acrobat Reader, a browsing utility available for free download at the website http://www.adobe.com/products/reader/?promoid=DJDXD.


The professional association for your discipline also gives awards for best dissertations. For instance, the Leonard D. White Award is given annually by the American Political Science Association for the best doctoral dissertation completed in that year or the previous year in the general field of public administration, including broadly related problems of policy formation and administrative theory. See http://www.apsanet.org/section_275.cfm. Check with your own discipline=s professional association for a similar list of award-winning dissertations which may provide guidance in your outline-formation efforts.

 

Bibliography

Davis, Gordon B. and Clyde A. Parker (1997). Writing the doctoral dissertation : A systematic approach. Barrons Educational Series, 160 pp., Focuses on the mechanics of dissertation writing. One author is a professor of management.

Gelfand, Harold and Charles J. Walker (1990). Mastering APA style: Student's workbook and training guide. American Psychological Association, 238 pp.