Research of the health status or health needs of individuals and populations is in many ways, an extension of what we do in our routine, day in and day out. For example, if we want to find out about something, we ask friends and relatives about those, watch the TV, hear on radio or read magazine articles or newspaper reports on such interests. Next we evaluate all this information, select the most important and relevant ones, and reach some conclusions for further action. Probably, the only significant difference between the above-stated daily routine of our lives and systematic research is the far more structured and formalized nature of research.
The reasons for doing research could be for the sake of learning more about something (research for knowledge), or for taking some concomitant action (action research) or for highlighting the concerns and needs of a particular group or area (research for advocacy). It can also be part and parcel of a community development strategy. Whatever be the reasons of the research, they must be kept in mind and preferably be specified right from the outset.
Planning the Research
The planning stage is crucial for the success of any study, and the amount of time needed to plan a study is often underestimated. Even if one is planning to do a small scale study on his/her own, still that will involve the carrying out of many of the various tasks described here-forth.4
5One of the most important tasks that pays rich dividends subsequently, is the framing of a time-table or schedule of various stages of the research process. When that takes the form of a Gantt chart, as shown in Figure 1.1 of the previous page, it becomes very useful in sticking to the deadlines. Henry L Gantt developed this pictorial system for use in planning and task allocation. This method was based on the recognition that overall program goals are in fact a series of interrelated steps. The Gantt chart consists of a series of events as bars that cross equal units of time on a chart. It has the advantage of allowing the researcher to have a clear understanding of the timing and interrelationship of all the components of the research. It is both easy to construct and interpret. However, it has the disadvantage that it does not bring out the dependency in relationships between different research steps.
Stages of a Research Study
Any research can be viewed as consisting of a series of stages. First, the preliminary ideas that we have need to be transformed into a formal “research design”. In the design we clarify the aims and objectives and methodology of the research. The aims and objectives of the research have to be rebutted through discussion on its need and supported through review of literature. Subsequently, the research objectives have to be translated into a “series of questions“ that can be investigated empirically within a specific theoretical as well as resource framework. Information is then collected using one or more methods, which is then subsequently analyzed using techniques that are appropriate. The data analysis finally generates the study findings, which has to be presented to a wider audience, through the thesis or reports. These various stages of research have been depicted by a line diagram in Figure 1.2. We shall now discuss in detail about each of these stages.
SELECTING THE TOPIC
Any research would begin with the selection of a researchable issue. But the process of selecting a study topic is easier said than done. While many of the students/researchers have pretty good ideas narrowed down for a study, there are many who really struggle with the idea of generating a study topic. However, most students do have a list of “potential research areas“ at the back of their mind. These are topics they tend to think about or those that raise their curiosity. Almost every day we are surrounded by events, situations, and interactions that make us wonder, stop and think, or bring forth emotions of joy, frustration, relief or anger. This is the potentially rich and fertile ground from which research ideas are born.6
Think about what stirs you up, what you argue about with your friends, family and peers, and what issues are topical in the world, at home or in your workplace. You will soon find that research or study topics abound. Appropriateness is probably, the only word of caution that needs to be exercised while listing probable study topics.
Deciding the Research Question: The research question is the objective of the study, i.e. the currently existing uncertainty that the researcher wants to resolve. To begin with, the research question is broad and of general concern; it has subsequently to be narrowed down to a concrete, researchable issue. That again is no easy task. A particular angle may be provided to the topic, based on personal experiences or prevalent theory or the existence of gaps in current literature. The need to define and redefine the research question during the initial stages of planning, based on extensive search of a body of literature, cannot be overemphasized. For example, the initial research question could be: Should women take hormones after menopause? In order to frame more specific research questions, we may sometimes break the whole question into constituent parts and single out one or two of these to build the protocol around it. For example, such questions could be:
- Do women take estrogen after menopause?
- If yes, how commonly?
- Does taking estrogen after menopause cause harm, or is it beneficial?
A good research question should pass the test of feasibility, interesting, novel, ethical, and relevance. These five essential characteristics of a good research question have been given the acronym of finer by Hulley et al in their book “Designing Clinical Research”.
Feasibility: It is best to select a research question by taking into account the practical limitations and problems in terms of: (i) Adequate number of subjects, (ii) Adequate technical expertise, (iii) Affordability in time and money and (iv) Manageability.
Interesting: A researcher may have multiple interests for picking up a study; these may range from real interest to get to the truth of the matter, or, logical step for building a career, or, even financial support from commercial firms. The study should be interesting not only to the researcher, but also their mentors and experts in the field.
Novel: A good study should contribute new information and not merely reiterate what is already known. The novelty of a study can be determined by a thorough literature review (vide below) and consultation with experts who are familiar with ongoing research.
Ethical: Good research must be ethical; it must not pose unacceptable physical risks or invade too much into privacy. Of course, the final decision is in the domain of Institutional Ethical Committee (IEC).
Relevant: This may be considered as the most important characteristic. Relevance may be determined by considering how knowledge obtained from the research might advance scientific knowledge, influence clinical management and guide further research.
Origins of a Research Question
For an established researcher, research questions usually emerge from his previous experience and study findings. However, for the first time researcher who has not yet set his foot firmly in this field, two sources would provide them with invaluable assistance: (i) Published literature, and (ii) Experienced mentor.
A new researcher must conduct a thorough search of published literature in the area of study using tips discussed later in the chapter. Further, there may be a lag of many years before recent advances are published. These are in fact first presented in conferences of professional associations. It is, therefore, important that first time researchers participate in such conferences and build some relationship with experts in the field. Informal conversations during tea-breaks with experts in a variety of fields can be hugely rewarding.
All the time one should be having a skeptical attitude about prevailing beliefs and practices. Such an attitude can stimulate good research 8questions. One should also try to put into application newer technologies, used in some other field. This can often yield encouraging research questions.
One of the most important points during the planning stage of the research process is to keep on capturing on paper all ideas that keep on cropping up in the mind. Some of the ideas come to us at the oddest hours, like in the middle of the night or while traveling. These need to be noted down in a diary, which one should always have with him/her, if the thoughts are not to be lost forever.
The primary ideas of the researcher are modified and re-modified by repeated drafts and verbal discussions, thus leading to the formation of the study protocol. Study protocol is the device which will be the final outcome of the planning process; it helps the researcher not only for seeking research permission and funding, but also in organizing the research in a logical, focused and efficient way. However, it is warranted that consultation be sought during drawing up the protocol right from the very beginning, from at least two groups of specialists; viz.
- Clinician specializing in the field, and
- Epidemiologist and biostatistician, who are professionally concerned with research design and data analysis, respectively.
REVIEWING THE LITERATURE
As the researcher proceeds in selecting and finalizing the research question, he has to do a thorough review of existing literature on the concerned topic. Review of literature is the summary of current knowledge about a particular issue; it also includes the gaps in the knowledge base. Review of literature, in quantitative research is conducted with the aim to develop as well as implement a research study.
Published information is predominantly cited in the review of literature and these should preferably be from primary sources. Published information from primary sources can be further classified as: theoretical and empirical. Theoretical literature includes concept analysis, models, theories, etc. and can be found in periodicals and monographs. Periodicals such as journals are published over time, while monographs, such as books/booklets of conference proceedings, are usually written once. Empirical literature, on the other hand, includes unpublished studies such as master's thesis, doctoral dissertations and personal communications. Review of literature should preferably be a balance mix of both theoretical and empirical literature.
The sources cited have to be relevant and current for the problem and purpose of the study. Locating the sources or abstracts of the sources 9and reviewing those can be very time-consuming. It is preferable to include sufficient current sources that are not more than 5-10 years old. If the problem studied is relatively new, it may be really difficult to find relevant studies; while if it is a problem that has existed for years, the literature review is likely to include landmark studies that might have been conducted decades ago. It is desired that quality articles from reputed medical research journals are checked before finalizing the topic.
Method for Literature Review
An ability to read research reports and critiquing the literature review of published studies would be helpful in conducting a review of the literature. The major literature review is conducted at the beginning of the research process, and a limited review is conducted during the generation of the research report to identify new studies. A quality literature review logically builds a case for conducting the study under consideration. It provides a basic understanding of the study problem and evidence that the study conducted was appropriate, based on the current knowledge of the problem.
The method for reviewing the literature includes the following (i) using the library, (ii) identifying relevant research sources and (iii) locating these sources. For our purposes, the academic library, one that is located within an institution of higher learning, and contains numerous research reports in journals and books, is the most valuable primary aid for review of literature on the problem. Once a problem has been identified in the clinical setting, one has then to search the literature for studies related to this problem. Firstly, a brief manual search of the literature may be undertaken, and then a detailed computer search is desirable.
Manual search involves examining reference sources such as catalog lists and indexes, by hand to identify relevant sources for an area of interest. Manual search should be brief and aim at clarifying and narrowing down the search topic so as to provide appropriate direction for a computer search.
The catalog identifies what is available in the library. It includes books, journals, conference proceedings, thesis, dissertations, etc. Some individuals overlook the value of books thinking their content to be too old. However, these can be excellent sources, especially on account of their detailed references.
An index provides assistance in identifying journal articles and other publications relevant to the topic of interest. Indexes are organized into two major sections, according to the subject and the author/s. The 10research topic, synonymous terms, and subheadings identified during the manual search would later be used to guide the search through the subject section of the index. Under these subheadings several published articles are listed. Many of these publications are listed under more than one subject heading and subheading for easy access by the researcher. If one is familiar with the names of key researchers for a specific topic, the author section of the index can also be searched to find recent publication by these individuals. The author section is organized alphabetically by the first author's names; though no more than three authors name appear as cross reference to the full citation under the first author's name. The author's name is followed by the same bibliographic content that is included in the subject section.
Index Medicus (IM) is the major index used in the field of medicine. IM is the oldest health-related index that was first published in 1879. This index cites articles from approximately 3,500 medical and related journals. This index is published monthly along with an annual combined listing, covering all aspects of biomedicine and includes allied health fields, such as nursing, the biological and physical sciences, veterinary medicine, and the humanities.
Computer search: With the advancement of technology in computers, software and networking, it has now become possible to conduct a scan of the citations in different databases and identify sources relevant to a research topic. This is what is conventionally called a computer search. After a brief manual search of the literature, clear direction for conducting a computer search would have been obtained. Feeding that information to the computer databases (discussed below) will generate a list of references with complete biographic information. Abstracts may also be readily available for many of the references. Computer searches have the advantage of being comprehensive, affordable, accessible, and are easy to conduct. Even a simple PC with an internet connection can be used to conduct searches. The internet is a worldwide network that connects computer and databases together.
When searching databases, one needs to organize the search according to subject, journal or author. The terms we use in the search determines the quality of the sources identified. Poor selection of terms can result in identification of an overwhelming number or an inadequate number of sources. It is, therefore, worthwhile to spend some time to review a few significant studies for identifying relevant terms and key researchers in the area under consideration.
Medlars (Medical Literature Analysis and Retrieval System) is the computerized system of databases offered by the National Library of Medicine (NLM), USA. This system includes more than 40 data-bases and includes Medline, Avline, Cancerlit, etc.
Medline includes about 6 million articles from approximately 3500 biomedical journals. A medline search generates the following information: author, title, abstract, language, and indexing terms. This database covers the period from 1966 to the present and is updated monthly.
Catline contains about 5,00,000 references for books and serials cataloged at the National Library of Medicine (USA). The information obtained includes author, title, language, indexing terms, and other data such as edition, series, etc. Catline originated in 1965 and is currently updated weekly.
Locating the Sources
Now we are ready to locate the sources identified by the manual and computer search. Locating the resources now further involves the following steps:
- Organizing the list of identified sources,
- Searching the library and the Internet for those sources, and
- Recording the references systematically.
The list of identified sources can be organized in several ways to facilitate locating them on the Internet or within the library. Journal sources might be organized by journal names and year. One can methodically locate all sources within a specific journal and then proceed to the next one. Sources included in the library catalog (such as books, organizational publications, and conferences proceeding) can be organized by author or subject. This organization not only makes it easier to find these sources but also assists in eliminating any duplication of the references.
Some research articles can be obtained using full text retrieval from the internet. Though most of the full text articles are priced these days and the subscriptions for these online journal articles can be paid through credit card. However, many journals permit free downloading of the abstract, and the researcher may decide after going through it whether the priced full article is worth it. Many universities have also purchased rights for use of these by the faculty and students. However, obtaining journal articles electronically is still quite expensive, and many students, therefore, prefer to locate sources in their university/institutional libraries.12
Often studies will need to be read several times to determine their quality and their relevance to the review of literature that we are developing. We then need to select the studies that are of the highest quality and compare their purposes, methods, results and findings. It is recommended that a table including essential information from each study be prepared for making comparisons.
The review of literature, as required for being included in the research protocol (and later in the thesis too) should begin with an introduction, include a presentation of relevant studies, and conclude with a summary of current knowledge. The subheadings and essential content of a literature review may be kept as described below:
- Introduction: This indicates the focus or purpose of the review, describes the organization of the sources along with the basis for ordering the sources, e.g. in decreasing order of importance or from most current to older studies. This section should be brief and interesting enough to capture the attention of the reader.
- Literature review: It should include quality studies that are most relevant for the selected project. For each study, the purpose, sample size, design, and critique of the strengths and weaknesses of the study have to be discussed. The content from studies must be presented honestly, keeping ethical issues in mind and these should not be distorted to support a selected project.
- Summary: This includes a concise presentation of the research knowledge about the selected topic, that is, what is known and what is not known.
FRAMING THE AIMS AND OBJECTIVES
Subsequent to above, refining and clear establishment of the aims and objectives has to be done keeping in view the time, cost and other resource constraints. It is to be noted that ‘aims’ and ‘objectives’ are not one and the same thing. Aims are generally accepted as the broad goals of the study, while objectives are more specific methods of achieving the aims.
PLANNING THE METHODOLOGY
The methodology planning includes a description of (a) study participants, (b) study design, (c) procedures and (d) statistical analysis. The planning for study participants has to include the characteristics of participants, the total number enrolled, how they are going to be chosen and assigned to different study groups, inclusion and exclusion criterion, etc.13
Establishing the inclusion criteria: Under this, we define the main characteristics of the target population. The task of specifying the clinical characteristics often involves difficult judgments about what factors are important to the research question and how to define them. For example, a researcher intending to find out the effect of vitamin D and calcium supplements in the prevention of osteoporosis may include only such subjects who have “good general health” or even “those with diseases not strongly related to osteoporosis“. The task of specifying the demographic characteristics, viz. age, sex, income, education, etc. is even tougher. Sometimes, no single course of action can straightaway be decided as being clearly right or wrong. What we have to keep in mind is that our decisions have to be sensible and there has to be a semblance of balance between costs, efficiency, generalizability, duration of study, etc. The inclusion criterion also addresses the geographic and temporal characteristics of the accessible population.
Establishing the exclusion criteria: Exclusion criterion indicates the subsets of individuals who do not qualify to be included as study subjects. One of the reasons for defining these could be that by including them we might be interfering with the quality of data or with the acceptability of randomization procedures. Exclusion criterion may improve the feasibility of a research study, but this is usually at the cost of generalizability. Therefore, we should use such exclusion criteria very sparingly. Sometimes, especially in experimental studies, exclusions may be necessitated by ethical considerations. However, the researcher must consider whether subjects excluded for practical reasons are likely to threaten the validity and generalization of the inferences.
The planning of the study design would have to be done based on the review of literature and after having deliberated on the various options available (Details in Chapter 2). The procedural planning would need the listing of study variables, framing of operational definitions of diagnostic criteria, deciding the protocol for experimentation, description of treatment and instruments going to be used in the study. One has also to consider the methods that shall be used to collect the required information. Sources of data, methods of sampling, and randomization procedures, if any, have also to be planned in advance.
The methodology planning should conclude with the method going to be employed for statistical analysis. It would include the statistical technique going to be used, number of participants in each group, methods proposed to be used for handling missing data as well as criteria for excluding observations, etc. It is also important to include a resource checklist covering human, financial, equipment, transportation, etc. in the research plan.14
WRITING THE PROTOCOL
The planning stage of research culminates with the writing of the protocol. Once the research question has been finalized, the process for arriving at the protocol can begin. It is generally accepted that at least two versions of the study plan need to be produced sequentially; these are:
- The one page outline protocol: This is the preliminary protocol and includes a checklist of all the components of the study. It has an orderly sequence that helps clarify the researcher's line of thinking on the topic. One such sample prototype protocol is shown in Table 1.1.
- The final protocol: This is a detailed expansion of the above outline and may range from 10 to 15 pages. This document serves multiple purpose; beginning with seeking the permission from relevant authorities (university, IEC, health authority, etc.), funding, etc. It includes details like, procedural techniques, data collection instruments, etc.
Applications for the granting of permission for research usually have to adhere to prescribed contents arranged in a particular format. For example, the guidelines circulated recently by Delhi University to the postgraduate researchers and their guides require that the following elements be included in the protocol:
- Title (1 page),
- A brief statement of the specific aims and objectives of the study (1 page),
- Background and significance of the study. This should discuss the known and unknown areas related to the problem and why the proposed study is worthwhile (2-3 pages),
- Methodology should describe the study subjects, their selection technique, methods used for data collection, diagnostic criteria as well as data analysis methods. Some sample blank tables (dummy tables) showing how the data is intended to be organized and analyzed should also preferably be included in the protocol. The sample size, as well as the ethical issues regarding participant's rights and welfare, also needs to be dealt with (2-3 pages),
- Index of references should follow the Vancouver system of references (2 pages), and
- Appendix, if any, should include ‘Consent Form’, ‘Instrument for Data Collection’ (2-3 pages).
The Board of Research Studies, Faculty of Medical Sciences, University of Delhi. Cautions that the following aspects should especially be taken care of while submitting the protocol:
- Feasibility of conducting the study in view of available resources and time frame,
- Ethical guidelines for undertaking human/animal experimentation as per Institutional Ethical Committee (IEC) guidelines,
- Statistical requirements in respect of sample size and data analysis of protocol must be satisfied (there should be a minimum of 30 subjects in each study group), and
- The proposed work should not have been conducted during the last 5 years (a declaration must accompany the protocol).
SUMMARY
Research is any systematic enquiry into existing knowledge to generate new knowledge.
Planning is the first and one of the most important steps of research.
Steps of a research process include:
Selecting the research question
Reviewing the literature
Drafting the protocol
Selecting the study population
Collection of data
Analysis
Writing the report
Review of literature includes:
Searching the Library through:
Catalogue | Index | Index medicus |
Searching the Computerized Databases
Medline | Pubmed, | etc. |
Methodology planning includes planning the study design, participants, clinical procedures, statistical analysis, etc.
Planning stage terminates with finalization of the protocol.