Review of the Literature: Sources

At least five sources of literature should be searched: journals, major books on the subject, monographs, relevant collections of images or objects, and dissertations. Today, much of the identification of relevant material in these five classes of publication can be done via computer access to bibliographic databases.

By this time, you should also have ascertained the bibliographic citation form required for your THESIS AND DISSERTATION. You should take pains to record, from every reference you use, the exact information you will need if you decide to cite it in the future. It is much better to err on the side of recording too many references now than to have to return a few weeks or months hence and spend hours trying to relocate some document.

If you are using a computer, we recommend strongly that you store the data for bibliographic citations. Much drudgery can be avoided, too, by using a computer program to structure your bibliography as you build it. Such a data file stores all types of references and can generate citations in more than 200 publishing styles, plus write footnotes in a variety of styles.

The term literature is employed to include anything appropriate to the topic, such as theories, letters, documents, historical records, photos and other images or objects, government reports, newspaper accounts, empirical studies, and so forth. Some of these, like letters or reports, are called “fugitive materials” because, while such items may actually be of key importance to your topic, the originals may be quite difficult to locate. This is another point at which a professional librarian can prove to be a great help.

Searches by computer for images (photos, sounds, drawings, settings, maps, paintings) that do not have text captions pose challenges because current software may not recover such material directly. With conventional databases, computers look only for exact matches of key words.

Software now in development may use artificial intelligence processes that allow the search engines to be sensitive to enough nuances about English (or another language) syntax to broaden or deepen a search and find more matches. Thus “sunny day at the beach” and “warm sand and sparkling surf” could produce a “match” instead of a “no match” response. Examples of Web sites potentially helpful with the above problem search areas are those for Cycorp, Inc. (http://www.cyc.com/), Virage, Inc. (http://www.virage.com/), and IBM (http:// www.ibm.com/). These sites were retrieved on May 19, 2002. See Chapter 3 for home pages of general and specific search engines.

Figure 4-2 contains six subheadings under Review of the Literature. In all six subheadings, if theory and research are each treated in chronological order in the writing , usually a coherent picture of the topic's background emerges. The key is appropriateness.

Extent and Depth of Literature Review

A review of the literature is necessary for every THESIS AND DISSERTATION, but there is disagreement about the detail and depth of the review at the proposal stage as compared with the final stage. Some advisors recommend a short review, hitting the high spots of the literature and anticipating the complete review in the final document. Others insist on a complete review in the proposal. Those who take the former position feel that the whole proposal, as a matter of principle, ought to be brief (up to 10 to 15 pages) and ought to concentrate on a statement of what the researcher wants to do, why, and how, plus how it will add to what has already been done. This, it is argued, permits a maximum of student independence and freedom from committee constraints and reviews during the study.

Those who take the complete review position feel that the student researcher will have a better proposal, clearer procedure, and better final product if the review is thoroughly done before embarking on the study. Further, they say, the review has to be done anyway, so it is not lost effort. A full analysis of the literature beforehand provides opportunities to educate both the student and the committee to some of the pitfalls ahead.

We lean toward the complete review view, but we prefer to focus on what the student researcher should try to get out of the review at the proposal stage. Whatever the depth and detail of the written requirement of this section of the proposal, we encourage students to read widely in the literature, take careful notes, and maintain an organized file and record.

For more ideas on literature review, use your search engine for on-line information. Also, the work of Clark and Oxman (2000) is a source for guidelines and format for systematic review of the literature in the health care field. The works Gash (2000), Hart (1998), and Cooper (1988) are all good sources of information. C. W. Bowen (2000) describes a meta-analysis approach to conducting literature reviews. He illustrates the power of this approach by reporting the effects of cooperative learning on chemistry achievement in high school and college classes. Finding samples of literature reviews in your field would be very helpful to your progress.

Benefits from the Literature Review

The most important benefit to be gained from a review of the literature at the proposal stage is good knowledge of the field of inquiry-what the facts are, who the eminent scholars are, where the parameters of the field are, and which ideas, theories, questions, and hypotheses seem most important. The reviewer at this stage ought to be able to carry on an informed, intelligent discussion about the field with an expert, using references and citing authors and concepts important or critical to the field.

Another benefit from the review at this stage is knowledge of the methodologies common to the field and a feeling for their usefulness and appropriateness in various settings. The reviewer can get ideas of which methodologies are most often used, methodologies appropriate for the proposed research, and when and how these ideas have been successfully used in the field. A careful look at the prevalent methodologies may also convince a student either to alter the topic of inquiry because skills are lacking to do that kind of research or to keep the topic and acquire the needed skills.

A third potential benefit is reinforcement for an earlier hope that the proposed research is really needed. Even if similar research has been published, the literature review often turns up statements such as “It would be interesting to know if Clark 's work could be replicated in other places or with different groups.” Or, “It would be very useful to know if Clark 's pioneering work is still relevant today.” These are clues that an important study needs to be replicated, essential information to know before the proposal is drafted (Lindvall, 1959).

One vexing problem researchers have faced for years is how best to maintain objectivity when reviewing prior research to assess whether further research on the topic is needed. This is especially difficult when there are many published studies reporting conflicting results. Some researchers have attempted to build tables or charts of the various study results to aid in “eyeballing” the prior reports and to estimate which might be considered more credible.

An impressive breakthrough in weighing the evidence from earlier research (Asher, 1990; Bangert-Drowns and Rudner, 1992; Glass, 1977; Hodges, 1986; Hunter and Schmidt, 1990) is now available. Called meta-analysis, the procedure allows one to add substantial objectivity to research reviews. Meta-analysis applies well to reviewing empirical studies with treatment means and control group means compared. Such studies constitute a population with results that are quantified by the literature searcher and put together into a database. The researcher then analyzes the database statistically, much like any other set of quantitative findings. We recommend its use not only to minimize reviewer bias, but also to help determine the nature of the hypotheses or questions and the directions they might take. Moreover, meta-analysis can be classed as a research methodology in its own right, and it is a widely useful one.

There are Web sites that permit free downloads. The University of Kentucky Computing Center (n.d.) citation is an excellent source. The work of Kenny (1999) is another good source, designed to assist the user in computation of statistics during meta-analysis. Another strategy is to enter meta - analysis in your search engine to see how much will be useful information. Finally, always consult your reference librarian for advice and for the names of download sites that are free to the university community.

Fourth , a literature review at this stage often helps to narrow a problem. Some get so overwhelmed with the flood of literature that severe frustration sets in. Where and how does one narrow a topic to make it feasible, yet not cut out the important details that impinge on it and make it more understandable and researchable? This is a common problem, shared by almost all beginning researchers. We recommend parsimony based on criteria agreed to with the advisor.

Try narrowing the scope of the review by employing a three-step sequence. Initially, read widely in the proposed field of interest. Then, think and analyze, attempting always to narrow down and weed out. After that, arrange to spend some review time with an experienced researcher in the proposed field of study and carefully talk through the problems encountered. If your advisor fits all the requirements of a good listener in this case, you are fortunate; use the advisor.

If you still have trouble getting the ideas to add up to a literature review for your proposal, go back to the first step. Each time the sequence is repeated, it will go more quickly. Vary the use of persons as sounding boards. At some point, a light will flash in the mind, and a good review for the proposal will come into focus.

A fifth value we urge students to wring out of the review is the generation of hypotheses or questions for further studies. The more one knows about a subject, the more questions come to mind. To a researcher, there is always a reason why something (person, group, organization, body, material) operates (behaves, works, acts) the way it does, but that reason (or complex of reasons) is simply not known. It may not have been researched enough, the data may be inadequate, or the theoretical constructs may not yet be available to guide and direct further research. From the literature analysis will come a multitude of ideas for further research based on the work that has already been done.

Keep a list of the questions and hypotheses that come to your mind or that are mentioned in what you read. (In the latter case, be sure that you also record where you found them so that you can prop-erly cite their sources later, if you wish.) In particular, that list will prove useful when you are writing the section of the final chapter of your THESIS AND DISSERTATION in which you discuss the implications of your own findings and the additional research directions your work supports or suggests.

Sixth , the topic being researched now could start a long-term interest, particularly if it is frequently updated and maintained with a consistent citation system in a computer file. Obviously, that makes it easily available for further articles and follow-up investigations.

In searching the literature, it is good also to develop a list of subject headings that relate to themes of interest. We recommend working back from the new to the old and from the general to the specific. Thus, a researcher might start with current reference sources and recent texts and research reviews concerning, for example, nuclear energy applications, dental hygiene in old age, public policy on the rights of victims of crime, adoption practices, or the education of gifted handicapped children. Then, working back through the earlier research will provide a depth context and understanding of current problems. Starting with the general topic will provide leads to specific areas of interest and help develop an understanding for the interrelationships of research. For example, the relationship between education of gifted children, handicapped children, mental tests, public policy, equal educational opportunity, inclusion, separation, and mainstreaming would be cases of interrelationships currently topical in psychology and in the public affairs, legal, social work, and education professions.

Currency of Literature Review Data

For those in the stage of reviewing literature, recent journal issues often provide leads to content that should be in the review. The parameters of academic or professional areas, the current thinking in the field, the investigators who are writing in the field, the ideas being discussed, and the references that are most cited and respected in the field are all displayed in recent issues of important journals. At an early stage of preparing the review, hours spent with journals identified as central to the topic may be very cost-effective.

One word of advice about journals: There is wide variation in quality. Some journals are very careful about what they publish; others are not. Some are refereed by top experts; others are hungry for any manuscript that turns up in the mail. Some are read by outstanding scholars; others would have difficulty finding their way into a schol-ar's wastebasket. How does one tell the difference? Ask professors, colleagues who know the field, and librarians who specialize in the field and look through journals. Read a copy of the journal; look at its format, its publisher, its board of editors; examine the qualifications of writers; and review the procedures for determining what gets published in the journal. You can also visit the Web of Knowledge home page (http://www.isinet.com), which will provide access to the major citation indexes (e.g., Social Sciences Citation Index , Science Citation Index , and Humanities Citation Index ).

A similar comment could be made about the library's book collection and the library's databases that give access to the collections of other libraries. As a researcher putting together a review of the additional conceptual structure for students to supplement the outlines in Appendix B. The general model does not parallel exactly the chap-ter or the topic development of this book, but all the items in the boxes in the figure are treated. If referred to in discussions with an advisor and other committee members or consultants, the model can be useful in maintaining the focus of a conference on specific problem areas and in working out solutions to them.

RESEARCH METHODOLOGY

The research methodology part of the proposal should identify for the reader the one or more research methods the student plans to use (i.e., opinion polling, case study, experimental, or other). The student's objective should be to give the reader a capsule statement about the contemplated research methodology, while indicating that details are to be found in succeeding portions of the proposal.

After the research questions and/or hypotheses have been decided, the single most important choice to be made by the investigator is the research methodology to employ. Once selected, the research methodology tends to govern, or at least limit, the range of choices as to how the data will be collected, how it will be analyzed, how results will be reported, and even the nature of the conclusions that may reasonably be drawn from the results.

We found no generally accepted classification of types of research. For example, two texts (Gay, 1996; Slavin, 1992) each list five research types, but they agreed on only three of them. Moreover, one text considers qualitative or ethnographic research a distinct and major classification, while the other subsumes qualitative investigations under another general category of research. Yet, they both agree that qualitative research developed primarily in anthropology and was borrowed as a methodology by other fields. They also agree in defining qualitative research as describing a situation as it exists, without involving formal hypotheses, but focusing on explaining social processes in great detail.

Because there does not seem to be anything comparable to a taxonomy in the classification of research types, we chose instead to simply list and briefly illustrate a variety of frequently utilized approaches to research.

Types of Methodology

For more specific aid to the student, we interrupt our orderly development of the outline in Fig. 4-2 to insert the section below. Types (or methods) are named and illustrated. A review of the types may prove useful in helping the student decide what will be written under the Research Methodology heading in the THESIS AND DISSERTATION proposal.

Accepted research methodology legitimately embraces a wide variety of forms:

In fact, there are many actions that can assist in discovering knowledge, and humanity learned much about human nature long before there was [formal recognition of] science and scientific method. Any actions that lead to accurate statements about nature must be considered as having some methodological legitimacy; the characterizations of contrasting methods are simply arguments that some scientific actions are more effective than others at producing statements of good generality. (Johnston and Pennypacker, 1980, pp. 412-413)

The methods chosen for a study certainly have profound effects on the outcomes. The same holds for how subjects are selected and for how data are collected and analyzed. Most important , though, is the match between methods and the theoretical or hypothetical propositions under scrutiny in the investigation. It is the statement of the problem that properly serves as a launchpad for all the subsequent steps. The correct research methodology should fit smoothly into its place in the research plan sequence to move from what one seeks by doing research to how one goes about seeking it.

These inquiries should govern selection of a research methodology: “What is (a) the most practical, (b) the most efficient, (c) the most promising, and (d) the most readily available way to solve the research problem or answer the research question?” Rarely will one method provide a “Yes” answer on all four points. More often, a compromise proves necessary.

Inevitably, the methodology choice influences the outcomes of the research. What matters most is that the student knows and acknowledges the influence and does not ignore or try to hide it.

In preparing the several editions of this work, the authors found no standardization in research methodology terminology across aca-demic and professional fields. To reconfirm this, we examined five research methodology texts published from 1993 to 1997. Altogether, they listed 15 types of research. All five books agreed on only 2 of the 15 classifications. Every book included at least 1 classification of research not found in any of the other four books.The terminology we decided to use includes all the types of research we could identify and illustrate. Even so, the name of the research type you choose should be reviewed with the advisor before being included in a proposal. The same methodology may be known by a different name in another field.

Examples: All of the examples under Item 8 (Experimental) would be applicable here if they were carried out under conditions in which only partial control was possible of variables, treatments, populations, or other important conditions-the case in many real-life situations for which field and operational studies are the only feasible kinds of professional and ethical codes are to be properly upheld.

These forms of research do not exhaust all that could be listed. Also, as noted, our category names may differ from those that others would use. Evaluation (Item 6, above), for example, may have many subgroups. Some are called outcome based, objective based, and consumer oriented, to name a few (Schalock, 2001). The most significant point is to recognize and employ the method most appropriate for the problem whatever the approach may be called.

Two other points should be noted. First, so far as can be determined, each of the several listed research approaches has resulted in significant new contributions to knowledge. No one deserves higher regard than another. Second, all students should learn about all of the research approaches. Naturally, a student will tend to become much more familiar with the approaches used in that student's THESIS AND DISSERTATION. However, all students should become familiar enough with the listed approaches to know how to use them and to know when an approach is not appropriate.

The proposal should clearly name and briefly describe the research method . Citations from authorities may clarify and support the choice of methodology, although the methodology selected is a responsibility of the researcher. The criterion measure of appropriateness is whether the methodology will yield useful evidence with regard to the statement of the problem. Thus, the choice is always directly related to the problem statement. Ask yourself, Will the method yield the data needed to make an intelligent and useful response to the problem statement? Raise that question with yourself, your advisor, and others who critique the proposal.

Specific Procedures

In the Specific Procedures section, tell the reader, step by step, what will be done in the conduct of the study. This may include correspondence, the design of questionnaires, pilot studies to be mounted before the complete study, the application of some treatment, the conduct of interviews, the distribution of inquiry forms or other instruments designed to gather appropriate data, obtaining permissions, the use of consultation, or other actions. Leave out matters to be dealt with under other headings.

Start with a chronologically ordered list of the procedures to be used. Elaborate each item in only enough detail to let the committee know what will be done. In a proposal, “enough detail” may vary from a paragraph on one item to a full page on another. If you go beyond a page, ask yourself if the writing is as parsimonious and succinct as it should be. Surplus words obfuscate the problem, annoy the committee, and raise the question of how sure you are about what you intend to do.

Research Population or Sample

In the Research Population or Sample section, a few essential points must be covered. What is the population to be studied? Is it a type of flora? Is it a group of research reports on which you will conduct a meta-analysis? Is it a form of virus? Is it a group of people? What are its characteristics? Will the universe (everyone or everything in the population group) be studied, or will there be a sample? If a sample, how will it be selected from the whole? What is the justification for selecting the sample? Is it possible to determine the representativeness of the sample? If not, does that fact constitute a prohibition or just a limitation? How does one gain access to the sample population, and how difficult a problem is that expected to be?

One reminder here-the research population and the sample, if any, should tie in very clearly with the statement of the problem. If the problem indicates a dependence on perceptions of urban residents or workers, for example, one might expect such individuals to be in any research population designated. Similarly, if one wants to be able to generalize to nursery-grown trees in the United States , the sample must be constructed with that goal in mind.

Instrumentation

In the Instrumentation section, detail the relevant data about instrumentation (tests, apparatus, interview protocols, questionnaires, and the like) proposed for the study. The purpose of any instrument should be to help produce or gather data to answer questions raised in the problem statement.

Review the potentialities in contemporary technology for ways to enhance the speed, the accuracy, and the reliability of instrumentation. For example, there are computer programs for administering questionnaires, for displaying stimulus material of a variety of kinds, for administering tests, and for a number of other uses requiring an interface with a subject. Such material comes onto the market in a steady stream as researchers recognize its value in minimizing human error factors.

Choosing the right package, for a particular set of data, among all the alternatives is a difficult task. Some Web pages that perform statistical calculations have links that will help in this task. For example, several university library Web sites will lead to bibliographies, journal searches, links to statistical packages, and on-line help (University of Kentucky Computing Center, n.d.; University of Miami Libraries , 2001; University of Minnesota Libraries , 2002). Several contain an integrated collection of statistical tests combined with a system that helps you to select, and then perform, the most appropriate analysis. Trochim (2002) at Cornell provides an interactive set of Web pages to help in the selection of appropriate analysis for data. It asks a simple series of questions about the data, then makes recommendations about the best tests to use. The work of Kenny (1999) is another good source, designed to assist the user in computation of statistics during meta-analysis. Another strategy is to enter meta-analysis in your search engine to see how much will be useful information. Finally, always consult your reference librarian for advice and for the names of download sites that are free to the university community.

Technological instrumentation employed must be cited, giving its name and where to locate it. As with any other reference or citation, the purpose is to allow a reader to access the material directly to verify that it was represented accurately by the author.

Careful advisors tend to want to see any instrument before it is applied to a research population. At least the specifications for any instrument should accompany the proposal, and it is much better if the entire instrument does.

Pilot Study

If you plan a pilot study, the proposal should describe when, where, and how it will be carried out. A pilot study, by definition, takes place before the actual study to determine feasibility of the study and to work out bugs. It is important, then, that the pilot have a provision for soliciting and gathering formative evaluation from the pilot study population. This pilot population, in other words, ought to have built into it a systematically designed opportunity for the student to learn the points of view of respondents, including any problems or suggestions the respondents may have.

Tell the reader if and how the pilot population differs from the proposed study population and how that difference may affect the study. Be completely honest. A difference may not threaten the validity of the study if recognized and accounted for beforehand.

After the pilot study, the researcher decides whether the proposed procedures need to be revised. The proposal must tell the reader what the researcher intends to do after the pilot study is completed and how it is to be done.

Data Collection

In the Data Collection section, one describes the nature of the data and how the data will be collected, including, for example, the mailing of questionnaires, the gathering of specimens, the scheduling of interviews, the search for documents in libraries in specified locations, or the recording of differences between two groups of subjects. A statement of provisions for follow-up is expected. Whichever type of research is proposed, the committee will want evidence that the researcher has thought about the possibility that data may not come easily and has made plans for the eventuality. Also, if animal or human subjects are involved, the protection of their rights must be indicated (see Index).

The process of collecting data must be appropriate to the research problem and the specific nature of the data. For example, interviews are very inefficient ways to collect large-scale survey information, although they may be used to supplement or to ascertain validity and reliability of such data. Other examples of inappropriate data collection procedures-which may spoil otherwise excellent research studies-are those that unacceptably invade the privacy of respondents or put subjects at risk.

Advances in computer technology now make feasible the firsthand collection and analysis of many kinds of data involving events and behavior. A laptop computer may be used to collect and store data and then to generate a variety of tables and reports, including answering many different questions about probability relationships between and among events. (More about computer utilization is discussed below.)

Treatment of the Data

After the proposed data are collected, what will be done? Will tables and charts be constructed, and if so, what will they display? If the data demand analysis, describe that analysis. Will the analysis be done in concert with some theoretical construct? If so, provide a description. The committee expects to find such content in the Treatment of the Data section.

Usually, the theory, the hypotheses, or the research questions guide the researcher in sifting through a mass of data. They focus the search and provide implicit criteria for the evidence search. A simple analogy might be made to archaeologists at a dig. There is a mountain of stuff to sift through, and perhaps all of it is interesting, but without theory, questions, or hypotheses guiding the search, it is possible that some important artifacts will be glossed over, and some critical relationships missed.

Here, too, computer software can make substantial contributions to the research process. Many colleges and universities maintain centers that furnish consultation to students and faculty on the selection and use of the most appropriate statistical procedures for answering questions or testing hypotheses from a given set of data. Such centers are much like the libraries in the way they operate, being there for use as needed. In some instances, they will store data for the researcher and run analyses as requested.

Doing such work on one's own is made easier, too, by statistical software packages on the market. SPSS and SA are popular sources for a large variety of statistical and graphic analysis packages. In addition, there are others. Statistical and Statistical Graphic Resources

(retrieved May 20, 2002, from http://www.math.yorku.ca/SCS/StatResources.html) provides resources for statistics, statistical graphics, computations related to research, data analysis, and teaching. It contains over 580 links.

The THESIS AND DISSERTATION student must remember, though, that use of a statistical procedure carries with it the responsibility of understanding and being able to explain the reasons that the procedure is appropriate. An American Psychological Association Task Force on Statistical Inference (1997, p. 10) recently reemphasized the need for understanding : “Computer programs have placed a much greater demand on researchers to understand and control their analysis and design choices.” This is part of the prevailing ethics of research in academic and professional circles. If you are the author, you accept responsibility for everything in the manuscript.

Appendices

An appendix may be needed to present drafts of letters to be sent, briefs of related research, as well as drafts of questionnaires and/or interview schedules, tests, or rating forms. This is also an appropriate place to put an appendix that shows a time line or flowchart of how the research will proceed. This could be adapted from Figs. 1-1 and 3-1. In general, an appendix contains documentation or evidence of important points made in the body of the proposal and referred to in the proposal. An appendix is the place for important tables, displays, or other items too long and detailed to be in the proposal's main body. For the proposal itself and its appendices, we recommend parsimony. If in doubt, leave it out. After all, items can always be added if that is the wish of the committee or the advisor.

Bibliography

A thorough, focused, succinct bibliography is mandatory. List only the materials cited in the proposal. This is not the place to list everything you can find on the subject in the library. Remember that the bibliography helps to indicate the authority of the work by the quality of sources cited, not the quantity.

A major purpose of the bibliography is to enable the reader to use the works cited. Therefore, each entry must be complete so the work cited can be found. If you must use documents that others cannot obtain, tell the reader in a footnote what the documents are and why they are important to your study. It will not help the reader if you simply cite such works in the bibliography.

Scholarship requires precision in citing, too. Lalumie` re (1993) distinguishes two kinds of citations. One refers the reader simply to a source of information, and the other kind gives support to a claim. He breaks the latter kind into four types (p. 913) that refer

(a) to an opinion of an author, (b) to conclusions derived from a narrative review of empirical studies, (c) to the method and results of an empirical study, and (d) to the method and results of a quantitative review of empirical studies (meta-analysis).

We join Lalumie` re (1993) in encouraging writers to be specific about the basis for a citation so as to avoid misleading the reader as to the weight attached to a cited report.

When specifying a reference, always seek to cite the actual source. If an article you are reading contains a pertinent quotation from another source, there is a temptation sometimes to use the quotation in your own manuscript as though you had found it yourself in the original source. First, that would be dishonest. It would not give due credit to the author of the article in which you really found it, and it would give the impression, falsely, that you had read the original source and discovered the pertinent quotation yourself. Second, to appropriate material in that way exposes one to dangers of two kinds. Some studies have shown (Adler, 1991) that citations are likely, often, to have been referenced incorrectly in even otherwise well-edited books and journals. Further, if you pretend to have quoted from an original source, you risk embarrassment if one of your committee members knows the original source well and begins to quiz you about what else it contains. Moreover, even if you have been motivated to locate and study the original source of the pertinent quote, it is a professional and academic point of courtesy to acknowledge the author who led you to do so.

MAKE SOFTWARE YOUR SERVANT

Almost every aspect of THESIS AND DISSERTATION work can be enhanced in appearance, done more efficiently, completed more quickly, or made more accurate using readily available software. This applies from the moment one begins to think about possible topics until the THESIS AND DISSERTATION is completed and being readied for journal or book publication.

Where can you find the software you need? First, ask the research librarian. Also, search the Web. For example, the Association for Survey Computing home page (http://www.asc.org.uk/) leads to more than 170 software packages with attributes and suppliers, with a built-in search facility. Information includes statistical design, design analysis, sample size in survey research, multilevel statistical models, and survey sampling routines. Also, follow other sites listed in the references (Clarke and Oxman, 2000; Florida Community Col-lege, 2001; King, 2000; Stempner, 2001; Trochim, 2002; University of Miami Libraries, 2001). To search for appropriate software, one can also use search engines such as Yahoo, AltaVista, or Google. Type buyer's guide and choose computer Internet software to find sources of information. There are also meta-analysis programs (Kenny, 1999; King, 2000) that may allow a free download. These sites often assist the user in the statistical computation and provide an array of computational options. Some have links to other sites for data analysis.

Consider keeping records for a fine illustration of how truly helpful software can be. Make a habit of entering, daily, one's thoughts, musings, ideas, questions, and discoveries in a computer's memory, in the form of one or more files . Such files then become your stockpiles of raw material to which you will probably return again and again as you find new uses for what was originally stored there. Moreover, you will find yourself elaborating on the items already in storage and adding new items.

A key point about filing notes is the value-even the neces-sity-of keeping track of the sources of your notes, especially those that came from reading books or journals, listening to lectures, or participating in discussions. This need emphasizes the importance of being able, later, to cite precisely the source of the information. The

need for complete and accurate knowledge about where one got an idea or a quotation points to the urgency of starting immediately to construct and maintain a bibliography .

A number of students have found it valuable to put the headings in this chapter into outline form right in their own computers so they can then insert relevant material as they learn about it or conceptualize it. Appropriate software not only will allow this, but also will make it possible to rearrange and edit to help clarify your vision of what you want to do and could lead you to make your writing more grammatically correct and compelling.

SUMMARY

This chapter helps the reader prepare the proposal by discussing the most relevant criteria for the process. A suggested format for the overview is presented, with recommendations about how to deal with each section of the proposal. An outline in the form of a Table of Contents is presented as illustrative material. Sections explaining, for example, the problem statement, the review of the literature, and the research design are major parts of the chapter. A list of different kinds of research methodology, with a brief explanation and some examples of each, is included, as are suggestions about maximizing the help available from contemporary technology.