Components of a Research Proposal
A Statement of the Problem and Its Setting
Statements of Hypothesis (or Hypotheses), Definitions, and Assumptions
An Outline of the Procedures You Propose to Use in Testing Your Hypothesis or Hypotheses
A Statement of the Estimated Cost of the Investigation
A Statement of the Schedule Planned for Carrying Out the Investigation
Excerpts from An Invitation to Submit Proposals
Notes on Writing Your Proposal for LIS 397.1, Summer 1998
A research proposal has five main components:
In the first part of a proposal, you should describe the problem you are proposing to investigate, and you should provide basic information about the situation in which the problem has arisen. Your goal in this part is to enable a reader of your proposal to acquire a general understanding of both the problem and its background. You should answer such questions as: What is the nature of the difficulty or of the area of uncertainty? What possible courses of action are under consideration? What decision or decisions need to be made?
In a full-fledged proposal to be submitted to your supervisor or to a potential funding agency (but not in proposals being prepared for LIS 397.1 in Summer 1998), this part of the proposal should contain a selective review of the published literature, if any, relevant to the problem. The review should clarify what distinguishes your problem and your situation from those already treated in the literature.
The hypothesis that is to be tested (i.e., whose truth or falsity is to be investigated) should be stated as simply and positively as possible. It is preferable to label the hypothesis clearly and display it in its own paragraph.
Contrary to a widespread misconception, it is not necessary to phrase a hypothesis in negative terms. It is quite all right to say directly what you hope to be able to prove; indeed, it is preferable to do so. Indirect or negatively phrased hypotheses are often difficult and obscure, and impose an unnecessary burden on your reader.
For example, the two statements, "Male and female directors of academic libraries are equally successful, on the average, in obtaining funds for their libraries" and "Male and female academic library directors do not differ in their mean performance in obtaining funds for their libraries," are logically equivalent, but the first is more readily understood.
You will learn later in this course that it is sometimes necessary to state negatively the statistical hypothesis that is part of a standard statistical procedure that you want to use to solve a particular problem--i.e., to decide whether a certain general, non-statistical hypothesis is true--, but you need not be concerned about statistical hypotheses at the beginning of LIS 397.1.
For clarity, it is preferable to provide a separate hypothesis for each of two or more relationships to be investigated. For example, instead of a single hypothesis of the form "Condition X will be accompanied by Condition Y and Condition Z," it is preferable to use two hypotheses: "Condition X will be accompanied by Condition Y," and "Condition X will be accompanied by Condition Z." Two cautions need to be voiced.
First, try to avoid making causal statements in your hypothesis or hypotheses. It is quite possible for two or more factors (i.e., variables) in your problem to be related without their standing in a causal relationship, i.e., in a relationship such that one of the factors causes the other.
Second, since the words "significant" and "significance" have specific technical meanings in statistics, you should strive to avoid using them in your hypothesis. Suppose, for example, you hypothesize that a certain change in a library policy will be accompanied by a sufficient increase in circulation to be worth (in your judgment) the trouble or cost of the change. Then your hypothesis should say that in so many words, or should speak of a "worthwhile" increase, but should not speak of a "significant" increase. (The technical meanings of "significant" and "significance" will be discussed in LIS 397.1 in connection with the testing of statistical hypotheses.)
The definitions should deal with terms that may be ambiguous in general usage and hence need to have their specific meaning in your investigation pinned down. It is not necessary to define terms that have specific meanings and are widely understood by librarians and information scientists, e.g., "monograph." But if, for example, your hypothesis concerns ways of increasing the "use" of your library, you need to make clear just what you are going to consider as "use" in your investigation and just how you will measure any increase in it.
The assumptions should deal with any factors in your problem which you think might affect your investigation but which you cannot control or examine directly. The point is simply to make such factors explicit in the record of your proposed investigation.
For example, as you are undoubtedly aware, the socioeconomic levels of a branch library's service area affect the use of the branch by people living in that area. Thus, if you were making a comparative study of two branches, you would probably deem it desirable to mention that the most recent census had shown their service-area socioeconomic levels to be comparable, and that you assume that there has been no important change in that comparability between the time of the census and the time of your study. If, on the other hand, you had good reason to think the contrary, then you would need to make a statement to that effect.
This part of your proposal should discuss what data you will collect, how you will collect them, and how you will analyze them. If sampling is involved, procedures for randomizing the sampling should be explained. If a questionnaire or an interview protocol (i.e., the questions to be asked in the interview, also called an "interview schedule") is involved, sample questions should be provided.
What you propose to do in testing your hypothesis must be consistent with the hypothesis itself. For example, suppose your hypothesis says, "The users of the XYZ Library will borrow more books each week, on the average, if the Library starts being open on Sundays during 1 - 6 p.m." Your test of this hypothesis cannot be the conducting of a survey of the Library's users in which you ask them to say whether they would like the Library to be open on Sunday and might borrow more books from it if it were open on Sundays. Such a survey would deal with users' opinions whereas your hypothesis deals with users' actions. The survey might be worthwhile in itself, but not as a test of your hypothesis. On the other hand, such a survey would be part of a proper test of a hypothesis that says, "A majority of the users of the XYZ Library will express an opinion favoring the establishment of open hours on Sundays for the Library."
As to the analysis of the data, a full-fledged proposal going to a supervisor or a potential funding agency requires at least a brief discussion of the principal analytic procedures (e.g., statistical tests) to be applied to the data.
(Note: In order that you may be free to concentrate on thinking about the conceptual--rather than the financial--problems of your proposal, the cost estimate and schedule may be omitted from the proposals prepared for LIS 397.1 in Summer 1998.)
Costs should be estimated as realistically as possible. Most costs you should be able to obtain for yourself, using current prices of materials and typical salaries and wages.
Salary and wage costs should be allocated to the investigation on the basis of the total social cost involved. For example, suppose your investigation involves an estimated total of 40 hours of your time as a school learning-resources specialist. Suppose also that you expect to contribute these 40 hours in the form of a few minutes each day throughout the coming year, so that you expect to squeeze the work into your normal workday without difficulty. Nevertheless, the 40 hours represent an allocation of your time to the investigation instead of to whatever else you might have done in those 40 hours in the way of extra effort applied to your normal duties. Hence, the 40 hours allocated to the investigation need to be charged at your regular hourly salary.
The schedule should set forth the expected starting times or dates and, similarly, the expected completion times or dates of the major phases of your proposed investigation. In doing this, you may (and will probably find it easiest to) use actual dates, such as "September 15, 199x" or "beginning with the fourth week of the Spring Semester, 199x-9x". Be as realistic as you can; try to picture yourself actually doing and/or directing the work involved; make reasonable estimates of the amount of time needed for each of the individual tasks in the proposed investigation; and then total up those amounts.
As an example of the application of the foregoing ideas to actual proposals, here are some excerpts from a set of guidelines prepared by the Texas Library Association (TLA) for persons interested in applying for TLA research grants.
"Submit applications to . . . the [TLA] Research Committee. A cover sheet with background data and a detailed budget should accompany applications.
"The Proposal document should contain the following elements:
2. Problem statement, or assessment of need;
3. Project objectives;
4. Research methods and procedures;
1) Wages and salaries,
2) Fringe benefits;
b. Rental, lease, or purchase of equipment and space;
c. Consumable supplies;
f. Other costs;
7. Future funding;
8. Plans for publication or dissemination of information;
9. Evidence of local support if applicable.
"In assessing requests, the Research Committee will consider the following criteria:
1. Appropriateness of goals for the Association . . . ;
2. Education and experience of the applicant;
3. Feasibility of the endeavor;
4. Reasonableness and appropriateness of budget;
5. Evidence of familiarity with related research . . . or [of] the lack of such work;
6. Significance to the profession and to the community;
7. Ability to evaluate results."
Though emphasizing monetary matters more than do the LIS 397.1 guidelines above, these TLA guidelines offer you a real example of the ways in which sources of funding tell interested persons how to apply for money for research. Each funding agency has its own particular requirements for the organization of proposals to be submitted to it, but all such agencies agree on the essential ingredients of a proposal:
Your proposal should employ an outline structure; i.e., you should use headings, sub-headings, etc. Legal-style outlining (e.g., 1, 1.1, 1.1.1, 1.1.2, 1.2, 2, etc.) is preferred. Make sure that you include the first three components described in these guidelines.
For a real-life proposal, be sure to include the cost estimates and the schedule described above. Also for a real-life proposal, be sure to include whatever other information the potential funding agency requests.
Go to Guide to Course Materials for LIS 397.1
Go to Wyllys Webpage
Last revised 2002 Nov 8