Research Methodology for Investigative Journalism: An Introduction1
Centre for Media and Labour Studies
This paper is about application of research techniques to investigative journalism. To put the topic in perspective, it is considered imperative to first define the key concepts – ‘research‘ and ‘investigative journalism‘.
What is Research?
We live in a world of problems. Life, therefore, is about problem-solving. Individuals, organizations, societies, governments, and so on, are never contented with their current states or conditions of life. They seek to make progress. But the attainment of that goal – attaining improvements in conditions of life – is often constrained by series of problems. Those problems must be solved or removed before any meaningful progress can be made. The research approach is one of the ways by which problems can be solved on a scientific, systematic and enduring basis.
Positing that research is ‘one of the ways’ for solving social problems means that there is recognition that there are alternative ways to solving human problems – the mystic approach, that is reliance on supernatural forces; reliance on ‘authority’; mere reliance on the practical ‘experience’ of a few, particularly reliance on the wisdom of older generation and reliance on ‘reasoning’ or ‘common sense’. While the research process may make use of some of the above, excluding the mystical approach in particular, it is not limited to any. It is more encompassing. Hence, research has been defined as a scientific and systematic thinking strategy which involves planned and formalized procedure for collection, analysis and interpretation of data for problem-solving (Aborisade, 1997:2). Funk and Wagnalls (1946:x) define research as a diligent, protracted investigation of some phenomenon or series of phenomena (Cited in Brynard and Hanekom, 2006:3).
Features of Research
A key feature of research is that it is teleological. This means it is concerned with a predetermined goal or purpose. Research is conducted to achieve a purpose – it is problem-solving oriented. Other characteristics of research have been identified by Mouton (1996, cited in Brynard and Hanekom, 2006:3 – 4). They include:
Epistemological: Research is driven by the search for knowledge with a view to establishing the truth, as much as possible, tentatively, in an aspect of life.
Ontological: Research is a study of the real world, society or mankind. Research takes place with the perspective that facts about mankind can be established and improvements in living conditions attained on the basis of increased understanding of phenomena in the social world.
Methodological: Research applies certain methods and techniques to gather and analyze the facts gathered in the pursuit of valid knowledge.
The Cyclical Research Process
The research process may be broadly represented diagrammatically as follows:
THE CYCLICAL RESEARCH PROCESS
Identification of problems
definition of problem
Formulation of tentative solution to the problem through research hypothese s or research questions , which guide researcher as to data to gather and literature to review. (3)
Gathering of data dictated by research hypothesis and/or research que stion (5)
Analysis of data gathered
Interpretation of an alysis in the light of hypothesis or Research Question , thus co nfi rming or rejecting hypothesis and providing a solution to
identified problem (7)
Review of literature in the light of hypothese s or research questions (4)
WHAT IS INVESTIGATIVE JOURNALISM?
The meaning of Investigative Journalism can be established from various dimensions, including the following five categories:
Self-initiated agenda-setting as opposed to serving as a mere transmission belt for passing information;
empirical facts-based reporting, or making a claim/allegation and supporting it with conclusive evidence;
judgmental reporting as against neutrality;
comforting the afflicted and afflicting the comfortable; and explanatory reporting.
INVESTIGATIVE JOURNALISM AS SELF-INITIATED AGENDA SETTING
Investigative journalism may perhaps be best understood from the point of view of what it is not. Following somebody else’s agenda is not investigative journalism. Acting as a mere transmission belt for passing information or proceedings of meetings to the public or uncritically accepting official versions of accounts on the basis of not wanting to offend authorities or individuals is not investigative journalism. Rather, investigative journalism is based on self-initiated agenda setting. Hence, the Investigative Reporters and Editors (IRE) define investigative journalism as ‘the reporting, through one’s own initiative and work product, of matters of importance to readers, viewers or listeners’ (cited in Houston et al, 2002: viii). Investigation is what distinguishes an independent report from an advertising brochure.
The Question then to be posed is: What kind of agenda do investigative journalists predominantly set? The answer lies in looking at investigative journalism as ‘comforting the afflicted and afflicting the comfortable‘.
INVESTIGATIVE JOURNALISM AS COMFORTING THE AFFLICTED AND AFFLICTING THE COMFORTABLE
According to Houston et al (2002:502), a cliché which is used to define investigative reporting is ‘comforting the afflicted and afflicting the comfortable’. The afflicted are mainly the poor, the vulnerable classes in the society, the deprived, the forgotten and the ‘voiceless’. Milna, a journalist, offers a meaning of investigative journalism when he says the duty of the investigative reporter ‘is to raise things which people in power find uncomfortable’ (Milna, cited in Spark, 1999:6).
Investigative journalism is therefore about looking into complaints and problems confronting the generality of the society and ‘taking up the cudgels on someone’s behalf’ (Spark, 1999).
In their definition of investigative journalism, the Investigative Reporters and Editors (IRE) states that ‘in many cases, the subjects of the reporting wish the matters under scrutiny to remain undisclosed (cited in Houston, et al, 2002:viii). Similarly, Spark (1999:6) submits that:
‘investigative reporting seeks to gather facts which someone wants suppressed. It seeks not just the obvious informants who will be uncontroversial, or economical with the truth, but the less obvious who know about disturbing secrets and are angry or disturbed enough to divulge them’
Investigative reporting therefore subverts the establishment rather than bolstering them or accepting their points of view without questioning. Thus, investigative journalism is about choosing issues that are relevant to people’s lives for investigation.
INVESTIGATIVE JOURNALISM AS EXPLANATORY REPORTING
It is not always that investigative reporting deals with afflicting the comfortable. Investigative journalism may also be explanatory. As Roberts puts it, investigative journalism is ‘not so much catching the politician with his pants down or focusing on a single outrage as it is digging beneath the surface so we help readers understand what’s going on in an increasingly complex world’ (cited in Houston et al, 2002: viii).
INVESTIGATIVE JOURNALISM AS JUDGMENTAL REPORTING
Spark (1999:1) asserts that investigative reporting is not an exercise in impartial balancing of stories in terms of making allegations and giving a right of reply to the victim. Rather, investigative reporting is judgmental. Cook said of his Checkpoint radio programme: ‘This is and always will be a biased programme – biased against fraud, criminality and injustice (cited in Spark, 1999:2). Similarly, the dictum of the Insight Team at Harold Evans’s Sunday Times consisted in the following:
‘All stories are either “We name the guilty man”, “The arrow points to the defective parts”, or “Stop these evil practices now” (cited in Spark, 1999:8).
INVESTIGATIVE JOURNALISM AS EMPIRICAL FACTS-BASED REPORTING
Expression of judgment on allegations made must be based on the facts gathered. Thus, investigative journalism cannot be judgmental unless it is based on empirical facts. Hence, Spark (1999:41) succinctly puts it that ‘the object of a journalistic investigation is to establish facts which put the rights and wrongs of an issue beyond doubt’
Investigative journalism therefore involves making an allegation and producing conclusive evidence. The investigative reporter is thus enjoined to: ‘check facts, never take anything on trust, never take anything on trust from people who have an interest in pushing a particular view’ (Spark, 1999:2)
From the foregoing, an investigative story or report should answer the what, who, when, where, why and how and make readers angry, sad, relieved and/or more informed about a topic that touches their lives. (Houston et al, 2002:525). Thus, we may define investigative journalism as self-initiated, facts-based, judgmental and public-oriented journalism, which seeks to either facilitate the functionality of the existing social structure or assist the process of social change.
Based on the above attempt at defining ‘research’ and ‘investigative journalism’, we can assert that the two concepts are similar. Research means scientific investigation to establish empirical facts and reach new verifiable conclusions about identifiable problems and/or to explain the unknown. Application of research to investigative journalism therefore means journalism that applies the methods of scientific enquiry.
KEY COMPONENTS OF A RESEARCH PAPER
The key components of a research paper, based on the cyclical picture above consist of the following: Introductory Section, Literature Review, Methodology, Data Analysis and Reporting of Findings. The remaining parts of this paper are guided by this framework.
COMPONENTS OF THE INTRODUCTORY SECTION
IDENTIFICATION OF RESEARCH TOPICS OR PROBLEMS FOR INVESTIGATION
The definitions of both scientific research and investigative journalism given above affirm that both are problem-oriented. Houston et al (2002) opines that investigative journalism is based on intense curiousity about how the world works or fails to work: How is something supposed to work? How well is it actually working? Why is it working the way it is? How is it organized so it works the way it is?
Williams, cited in Houston et al (2002:ii) also gives the following advise to assist in identifying topics for investigation:
‘sit and stare at the wall and ask what things happen in this town that affect a lot of people but are never written about?’
Based on the above insights and considering the status of Nigeria as a fast degenerating and under-developed poverty and corruption-stricken society, tons and tons of topics daily cry for attention, at all levels of the society, in both the public and private sectors of the economy. They include:
- Official corruption or waste of government resources
- Environmental pollution and climate change
- Unsafe workplaces
- Bad roads
- Epileptic power supply
- Epileptic petroleum products supply
- Lack of water supply
- Quality of housing units, including unsafe classrooms
- State of hospitals and costs of medical care
- Commercialisation of social services, including education
- Kerosene adulteration causing deadly explosion
- Petroleum products: production, importation, distribution and pricing.
- Politically motivated crimes, particularly murders
- Consumer Issues: manufacturing and distribution of fake drugs and food.
- Arms of government: finance, expenditure profile against appropriation Acts, contract awards, etc.
- Politicians: qualification of public office holders, past records in workplaces, security agencies, etc.
- Rigging of election processes
- Anti-corruption Agencies: Constitution, furniture, powers, independence, penalties for shirking responsibilities or maintaining dead silence on petitions received, checks and balances., etc, etc, etc.
SOURCES OF INVESTIGATIVE OR RESEARCHABLE TOPICS
A reporter-researcher may arrive at investigative or researchable topics through any or a combination of the following sources:
1. Personal Experience/Knowledge of the existence of a problem. Investigative project ideas may be developed from hints or trends perceived from regular (daily/weekly) beats/assignments.
2. Discussion or tips from other people. They could be friends, social acquaintances or even strangers. It could be through listening to conversations in elevators, restrooms, public offices, at bus-stops or in the buses or taxis, courtrooms, libraries, public phone-booths, telephone conversations made in the streets, newsstands or newspaper vendors’ locations.
3. Reading. Reading of local, national and international newspapers, magazines, websites, subscription to relevant electronic mailing lists or listserves.
4. News: News on radio, television. It also includes paying attention to ‘breaking news’ and asking why or how what has been announced happened?
5. Previous projects by other reporters or academic projects. Problems already investigated may be re-looked under the following conditions: controversial findings, methodological weaknesses changes in time and place and suggestions for further investigations, which may be contained in previous studies.
GUIDELINES FOR CHOOSING RESEARCHABLE OR INVESTIGATIVE TOPICS
- Interests: Choose a topic that interests you, that has meaning to you. Interest is the basic for excelling in any human endeavour.
- Duplication: Avoid duplication or re-examining topics which others had previously worked on except the conditions earlier mentioned above (under sources of investigative or researchable topics’) obtain.
- Feasibility: The practicability of carriyng out the investigation should be determined by asking and answering the following questions:
- Are data accessible?
- Is there adequate time to carry out the project?
- Are the variables observable and measurable?
Research is based on empirical data not on the mystical which is not observable and verifiable.
- Are there enough resources to carry out the project? It may involve several long travels to check and cross-check facts from people.
- Do I have the expertise and resources to interprete technical materials/details or could experts be hired?
- Can one reporter carry out the project? If not, are other reporters, editors, researchers, students on industrial attachment, librarians, outside experts/assistants be involved and/or hired
- Could the subject (person or organisation) being investigated apply pressure on people not to talk?
- Any Downsides?
- Could the subject successfully sue?
- Are there possibilities of advert withdrawals?
- What are the possibilities of readership or viewers’ boycotts?
However, considering the possible downsides does not mean abandoning the project; it only calls for necessary caution.
- Significance: The topic should be significant. This means it should be about important or fundamental problems. The idea is to avoid wasting resources on insignificant problems.
- Is the reporter theoretically equipped to conduct the project or could some experts be consulted or interviewed to provide the necessary theoretical insights?
- Does the topic imply a wide scope or does it have specific focus? Topics with wide scope should be avoided. Topics and/or objectives of projects should be narrowed down or delimited to manageable levels.
STATEMENT OF THE PROBLEM
The research problem is to be stated in such a way that readers will be enthusiastic and concerned that there is an urgent need to find a solution to the problems. To this extent, the researcher reporter will be expected to briefly explain the importance of the chosen topic/problem, the current developments, the patterns, the magnitude, and so on.
Asika (1991:100) recommends four main steps in dealing with problem statement.
Step 1: State the problem clearly, and in a single sentence. For example: “This study is concerned with examining the effects of education at NIPSS, Kuru, on the performance of former students as heads of governmental agencies
‘This study is concerned with establishing the causes of corruption in the Nigerian public sector’
Step 2: Show why it is a problem
Step 3: Link the problem with a given theory or known fact(s)
Step 4: Show the implication of the problem to a general or more serious problem.
To be convinced that the researcher-reporter has satisfactorily dealt with step 1 for instance, what has been stated under step 1 should be an answer to the following questions.
‘what problem is my research trying to solve?’ or
‘what situation is my research geared towards improving?’
Thus, a research could be concerned with the cause(s) of a social problem, effects, impacts or consequences of a social problem, impact of an Agency, etc.
The researcher-reporter should pay attention to the following requirements:
- The objectives should be unambiguously, unequivocally or clearly stated.
- The objectives need not be multiple. The study could have a single objective.
- The objective(s) should be relevant to the research topic and flow from it.
- The objective(s) could be introduced along the following style:
‘The sole objective of this study is to assess the relationship between attendance at the Nigerian Institute for Policy and Strategic Studies (NIPSS) and the performance of former students as heads of public Agencies’.
RESEARCH HYPOTHESIS OR RESEARCH QUESTION
A research hypothesis is defined as an informed guess of possible or likely solution to a problem. It could also be defined as a tentative solution to a problem. Research hypothesis and research questions perform the same role – they act as a guide for the researcher to know what types of data to collect in the field in order to ensure that efforts are focused in the right direction.
While research questions are more relevant to qualitative studies, hypotheses are more relevant to quantitative studies. Research questions are by nature interrogative while research hypotheses predict relationships.
FORMULATION OF A HYPOTHESIS
The following steps could be adopted in the formulation of a hypothesis:
1. Write down the research topic or the research objective, e.g.
‘The role of education at NIPSS on the performance of NIPSS graduates as heads of public agencies’.
2. Change the topic or the objective into a question, e.g.
‘What is the effect of education at NIPSS on the performance of NIPSS graduates as heads of public agencies?’
3. Identify the ‘X’ and the ‘Y’ variables in the title or objective, e.g.
X variable = Education (at NIPSS)
Y variable = Performance (of NIPSS graduates)
4. Make use of the variables to formulate a suggested answer to the question stated in step No.2 above, e.g.:
Ho: Chief Executives who have been educated at NIPSS do not perform better than Chief Executives who have not been educated at NIPSS.
HA: Chief Executives who have been educated at NIPSS perform better than Chief Executives who have not been educated at NIPSS.
Based on the above hypothesis, the researcher-reporter will gather data on heads of public agencies that had been educated at NIPSS and those who had not been educated at NIPSS and the quality of their performances. Thus, the hypothesis acts as a guide in determining what data to gather – education and performance.
If the researcher-reporter elects to use research question, the hypothesis in step No. 4 can be changed into a research question as follows:
‘Do Chief Executives who have been educated at NIPSS perform better than Chief Executives who have not been educated at NIPSS?’
The research question above requires the same set of data: education (of Chief Executives) and their performances. The researcher-reporter is to explain how he/she intends to measure ‘performance’.
SIGNIFICANCE/JUSTIFICATION OF THE STUDY
Here, the researcher-reporter should justify why time, money and labour should be expended on the project. Significance of a study comprises two main elements
- Identification of the benefits derivable from the study, and
- the beneficiaries.
Limitations of a research work refer to shortcomings, imperfections or weaknesses which may limit the data accuracy and validity of the findings. The purpose of pointing them out is to convince the reader that the researcher is aware of them and that in spite of those weaknesses, the findings are still reliable.
Literature review is indispensable in ‘backgrounding’ the subject of investigation.
‘Literature’ in the context of this paper refers to the written or printed word -documents and files of public institutions, media publications, books, etc, The researcher-reporter should gather any relevant document, whether printed, handwritten or in any form for review. ‘Review’ means comparing and contrasting the works, views, perspectives and findings of previous researchers. The similarities and differences in their views or findings, etc, will be identified. Whether the researcher-reporter agrees or differs from them or any of them will also be indicated.
Literature review therefore is not a place for full blown expression of the opinions of the researchers but a discussion of opinions, findings, etc of those who had worked in similar areas in the past.
Literature review serves a number of purposes: it provides a theoretical basis for the work and offers the researcher-reporter an insight into the best methods, instruments for data gathering and the statistical tools for analyzing the data gathered, which previous researchers had used.
How do we determine what literature to review? The Key concepts or terms involved in the topic, research problem, objectives, research hypothesis/research questions and so on are the guides to determining which literature to look for.
ACKNOWLEDGING SOURCES OF INFORMATION
It must be stated that there are many styles of acknowledging sources of information. What is important is consistency. The researcher should guide against mixing different approaches in a particular work. This paper explains the broad guidelines as far as the American Psychological Association (APA) format is concerned. But other styles include: Modern Language Association (MLA); the Council of Science Educators (CSE) for those conducting research in the natural sciences; The Columbia Guide to Online Style (COS), and the Chicago Manual of Style (CMS). Each of these has voluminous manuals, which are regularly revised. (See The Rodrigues, 2003).
Acknowledging sources of information is done in two places – in the text of the project/work and in the reference list or bibliography section.
Reference Citation in the Text
Plagiarism should be avoided. Plagiarism is the act of using the work, ideas information, perspective, findings of other people without attributing such to them. Any borrowed ideas, facts, findings, etc ought to be acknowledged.
Following the American Psychological Association (APA) style, three pieces of information are to be provided in the text. They are the author’s surname, the year of publication and the page number, if the idea can be traced to a particular page, particularly where quotations are involved. Two alternative ways may be used in text citation – acknowledging the source ‘before’ or ‘after’ the borrowed idea or finding, etc is presented.
Example: Acknowledging ‘Before’:
Ikhariale (1998) has identified the main problems of environmental pollution. They include ……
Example: Acknowledging ‘After’:
Poisonous chemicals and wastes capable of destroying human organs have been found in dump sites (Okecha, 2000: 54 – 57).
It should be noted that titles such as ‘Prof.’ ‘Dr.’ ‘Pastor’ ‘Chief’ ‘Mr.’ ‘Mrs.’ etc are not to be used. Initials are also not reflected in the text except under certain conditions such as making references to authors with the same Surname or Personal Communications. Personal Communications refer to letters, memos, emails, personal interviews, telephone discussions, and the like. Personal Communications are not to be included in the Reference list or Bibliography because they are usually not retrievable for verification by readers.
REFERENCES OR BIBLIOGRAPHY
The Reference list or Bibliography provides information necessary to identify and retrieve each source cited in the text.
However, while ‘Reference List’ contains only the list of sources that were used in the research, a Bibliography includes both works cited in the text and those used for background or for further reading, which are not referred to in the text directly.
Entries in either the Reference list or Bibliography should have a hanging indent. The entries are to be done in alphabetical order, by the surname of the author or the first author, if there are two or more authors for a particular work.
GENERAL FORM OF PRESENTATION OF BIBLIOGRAPHY FOR PERIODICALS
Periodicals refer to items published on a regular basis such as journals, magazines, scholarly newsletters, newspapers (dailies and weeklies). Entries are to be presented as follows:
Author’s surname, initials. (Date of publication). Title of article. Title of Periodical, Vol. No (if any), p.XX or pp.xx – xxx.
However, the word ‘volume’ or the abbreviation ‘Vol’. will not be used. The information is just to be stated in the appropriate location, as shown above.
MODES OF STATING PUBLICATION DATE
(2007, November). (for meetings, monthly magazines, newsletters)
(2007, November 3). (for dailies and weeklies)
(n.d.). (for work with no date indicated)
(in press). (for work accepted for publication but not yet printed).
GENERAL FORM OF PRESENTATION OF BIBLIOGRAPHY FOR NONPERIODICALS
Non-periodicals include items published separately such as books, reports, brochures, manuals, monographs, audiovisual media, and the like.
Author’s surname, initials. (Year of publication). Title of work. Location or place of publication: Publisher’s name.
Simpson, S. & Fagbohun, O. (1998). Environmental law and policy. Lagos: Law Centre, Faculty of Law, Lagos State University.
GENERAL FORM OF PRESENTATION FOR ONLINE PUBLICATIONS: ONLINE PERIODICALS AND DOCUMENTS.
The major differences are that the date of retrieval and the website address are provided.
Author’s surname, initials. (Date of publication). Title of article. Title of Periodical, Vol. number (if any). Retrieved month, day, year, from….. (indicate website address).
Author’s surname, initials. (Year of publication). Title of work. Retrieved month, day, year, from…… (indicate website address).
MAJOR DIFFERENCES BETWEEN CITATION FOR PERIODICALS AND NONPERIODICALS
The major differences between citation for periodicals and non-periodicals, as may be noted above, are that place of publication and the publisher are indicated for non-periodicals whereas they are not to be supplied for periodicals.
NOTES ON KEY ASPECTS OF ENTRIES
The Author’s surname comes before the initials. Where a work is authored by two or more authors, commas are to be used to separate authors, surnames and initials. However, for a work that is part of a nonperiodical such as a chapter in a book written by several contributors and edited by another person(s), the initials precede the surname of the editor, as follows:
Author’s surname, initials. (Year of Publication). Title of Chapter. In (state initials of the editor without enclosing the information in brackets). Surname of the editor. (Ed.). Title of the Book. (Page numbers of the article or book chapter), Place of Publication: Publisher.
Okecha, S.A. (2000). Environmental problems of the Niger Delta and their consequences. In A. Osuntokun. (Ed.). (2000). Environmental Problems of the Niger Delta. (pp. 51 – 61). Lagos: Friedrich Ebert Foundation.
2. Title of Article or Chapter
The first word of the title and of the subtitle, if any, as well as any proper nouns to be in upper cases while the rest should be in lower cases. The title should not be italicized and quotation marks should not be used.
- Title of Periodicals
The initial letters of the first word and of other key words in the title of the periodical should be in capital letters (i.e. initial caps). The name of the periodical and volume number should be italicized.
Also, if the periodical does not use volume number, the month, season or other form of designation should be used with the year, e.g. (2007, May).
4. Title of Non-Periodicals
The first word of the title and of the subtitle, if any, and any proper nouns should be capitalized. The entire title should be italicized. However, any additional information given for identification and retrieval (such as edition, report number, volume number, put in parentheses immediately after the title) should not be italicised.
5. Publication Information: Nonperiodicals
- The location or place of publication refers to the town or city where the publisher published the work. The state, province and/or country are not to be reflected, unless the city is not well known.
- The term ‘Publishers’ ‘Co’ or ‘Inc.’ are to be omitted but terms such as ‘Books’ or ‘Press’, should be stated if they appear as part of the name of the publisher.
- If the publisher is an institution such as University of Ibadan, The Polytechnic, Ibadan with names reflecting the city or place of publication, the place of publication is not to be repeated in the publisher location.
- If multiple publisher locations are given, it is sufficient to indicate the first stated location or the location where the publisher’s home office is based.
This section is a plan, which outlines the elements of the research and how they are related to each other. The researcher has to provide information on the following before starting the study:
1. What type of research design will be adopted?
2. What is the target population of study and where is the study area?
3. What is the total numerical and strength of the study population?
4. What is the sample size – what portion of the study population would be studied?
5. What sampling techniques will be used to determine the sample?
6. What types of data – primary and/or secondary – will be collected?
7. What research instruments will be used in gathering the relevant data?
8. What measures will be taken to ensure the research instruments are valid and reliable?
9. What statistical methods will be used in analyzing the data?
A brief explanation of some of the above elements follows:
There are different types of research design, corresponding to research typology. The researcher is to reflect and specify the one that is most relevant to the nature of the intended research. Research and research designs have been variously classified. The following may be considered as a form of categorization.
1. Historical Research/Design. It aims at having accurate account of the past, chronologically, and gaining a clearer perspective of the present through a sensitive interpretation of fact events (historiography). Historical account of things, institutions and people might show some trends and relationships.
2. Case Study Research Design. It permits an in-depth study of a given social unit – an individual or group of individuals, community or institution.
3. Causal-Comparative or Ex-Post Facto Research Design (After-the-fact) Research. It attempts at establishing cause and effect relationships, but after the event has occurred, without the possibility of manipulating the independent variables.
4. Experimental Research Design. This involves experiments in order to establish cause and effects relationships by evaluating the influence of a stimulus, which is either introduced or eliminated. It may require dividing the study population into two groups – Experimental and Control Groups.
5. Survey Research Design
In a survey research, the researcher is interested in studying certain characteristics, attitudes, feelings, beliefs, motivations, behaviour, opinions of a population, which may be large or small, without attempting to manipulate any variables.
Survey studies may be divided into the following categories:
1. Panel Survey Research Designs
This involves gathering data on the same set of issues with the same respondents at various points in time, on two or more periods. An example is a public opinion survey, which periodically measures the perceptions and attitudes of a sample of the population regarding their views of the country’s economic performance carried out at intervals.
2. Developmental Survey Research Design
A developmental survey seeks to establish changes in the characteristics, opinions, belief, etc of a population over a period of time.
It may be divided into
- Longitudinal Studies, and
- Cross sectional Studies.
In a longitudinal survey, the same set of people, items or subjects are studied over a given period of time, at different points in time and changes noted and recorded.
In a cross-sectional study, a one-time observation of as many variables as are necessary is done at once. A study of a sample on opinions, values and preferences regarding the state of the economy is an example of cross-sectional survey. Responses of the sample are broken down on the basis of variables such as age, sex, race, education, and so on. The aggregate responses of each of the categories are compared.
3. Descriptive Survey
Descriptive survey designs gather data with a view to describing the features of the phenomena being studied. Examples include employment surveys, manpower and inflationary trends.
4. Correlation Survey Designs
These seek to establish the existence or otherwise of relationships between variables.
5. Public Opinion Survey Designs
Public opinion surveys seek to establish an idea or a feeling of the people in a given place, on a policy issue or event of interest. However, it is difficult to generalize the results obtained to the entire population because random sampling technique is not usually adopted.
The study population is the group of people, households, objects, events, items, companies, institutions that the researcher seeks to make findings about. If the population is large to the extent that a census study is not possible, the researcher has to select a portion of the population for study. The portion of the population selected for study is known as the sample.
There are several formulae for selecting a representative size. But a simple approach is using the attached Appendix (suggested by Stoker (1985), quoted in Brynard and Hanekom (2006) as a guide. It should however be noted that the more heterogeneous the population, the more the sample size should be and the more homogeneous, the less the sample size, to be representative of the population.
Sampling techniques deal with how to select the members of the sample from the study population. Assuming a study population of 3000 persons, items or institutions wherein the researcher seeks to select 300 as members of the sample. How do we select those 300 items, persons or institutions?
There are two main methods of selecting respondents for inclusion into the sample: probability and non-probability sampling methods which are also called random and non-random methods.
In the probability sampling methods, each member, person, thing or element in the study population has an equal chance of being selected as a sample member because selection of participants is by chance and is not influenced by the people who do the selection. On the other hand, all the elements do not have equal chances in the non-probability sampling methods.
Non-probability sampling methods are useful for quick and cheap studies, for case studies, pilot studies and qualitative research. The most important advantage of the random sampling methods is that the results of investigation can confidently be extended to the entire study population.
TYPOLOGY OF SAMPLE DESIGNS
|NON-PROBABILITY (NON-RANDOM) SAMPLING METHODS
|1. Simple Random Samplingi. Balloting
ii. Table of Random Numbers
iii. Systematic Sampling or Fixed Interval Position
2. Stratified Sampling
3. Multi-stage Sampling
4. Cluster or Area Sampling
5. Double Sampling
|1. Accidental or Convenience Sampling2. Quota Sampling
3. Snowball or Network Sampling
4. Purposive or Judgment Sampling
TYPES OF DATA
There are two major types of data in research: primary and secondary. The researcher is usually required to:
- Specify the types of data to be used, either primary or secondary, or both.
- State in detail the specific data that will be gathered under each type, e.g. changes in staff levels, changes in budgetary allocation to an institution may be obtained through secondary data.
- Specify the sources through which each element in the data to be generated will be gathered.
WHAT ARE PRIMARY AND SECONDARY DATA?
Secondary data refer to data which already exist (e.g. admission list of students, budgetary allocations to arms of government, etc) but were not collected specifically for the purpose of a proposed study. From the viewpoint of journalism, we can also define secondary data as information already published or broadcast. On the other hand, information generated expressly for a specific purpose by the researcher is called primary data.
While original information needed may be obtained through primary sources, information from secondary sources has to be verified This is because rejoinders or dissenting letters to the editor might have been written. Apologies might have been tendered and corrections made. Books and other types of publications might have been revised.
SECONDARY SOURCES OF DATA
Depending on the nature, sources of secondary data are multifarious. They include:
- Recorded news broadcasts
- Online versions of stories
- Television & Radio programmes
- Web only publications
- Reference books
- The Internet: listserves, etc.
- Dissertations & Theses: Doctoral and Master’s Students often receive unparalleled access to institutions because of the perception that their research works are essentially for academic purposes which may not be as threatening as those of journalists.
- Depository libraries: National Library and National Archive
- Government Agencies, Ministries, etc.
A research instrument is a device used in collecting data. The researcher is expected to specify the specific instrument or a combination of instruments to use for given projects.
There are three main types of research instruments. The limit and nature of this paper do not give room for an explanation of them. They are:
1. Interviews. Several books and volumes of books exist on interviews.
2. Questionnaire. There are several types of questionnaire. The author of this paper has an unpublished 90-page manuscript on questionnaire construction.
3. Observation. There are two types: participant and non-participant. Two main instruments available to measure or record observation are the checklist and the rating scale. While the checklist determines attributes or items that are present or lacking in a subject of interest, the rating scale measures the degree or extent to which an attribute is present. In other words, the rating scale measures both in kind and degree.
VALIDITY AND RELIABILITY
The assumption in research is that the data gathered and the results/findings made are of the highest possible quality – that they were obtained empirically and objectively analysed. Validity and reliability have to do with the concern to attain this goal.
Validity is concerned with the ability of a research instrument to measure what it is designed to measure while reliability is concerned with the consistency in the results given by the same instrument from all respondents. For example, the clock is designed to measure time. If it measures time correctly, it is a valid time measuring instrument. But if it measures time wrongly, it is not a valid time measurement instrument. But assuming the clock is consistently late or fast by 10minutes at any point in time, it will be said to be reliable but not valid. Thus, validity is superior to reliability. But both are important in determining the usefulness of a measuring instrument.
MEASURES TO ENSURE VALIDITY OF RESEARCH INSTRUMENTS
- Content validity
- Criteria validity
- Construct validity
MEASURES TO ENSURE RELIAILITY OF RESEARCH INSTRUMENTS
- PILOT STUDY
- Training of field researchers
METHODS OF DATA ANALYSIS
The researcher is expected to have given a thought to the statistical tool(s) that would be used to analyse the specified data, before commencing the project. The data to be gathered will be restated and the statistical tool to be used in analyzing each element of the data will also be stated.
STATISTICAL DATA ANALYSIS
Statistics assists the researcher to summarise a whole mass of data succinctly. It is a means of processing numerical facts and interpreting them meaningfully.
However, the worth of a research work is not determined by whether or not it involves statistical analysis. Statistical analysis is more relevant to data and research findings that are quantitative in nature, that is, data that can be represented by numbers. An historian who is concerned with investigating politically motivated murders in Nigeria would be hard-put to produce either a statistical or an experimental study, yet the research may still acquire the status of a scientifically and scholarly study as that of any statistical or experimental study.
BASIC TYPES OF STATISTICS
We can distinguish between two broad types of statistics – descriptive statistics and inferential statistics.
Descriptive statistics are concerned with the description and/or summary of the characteristics/indices of data about a population or a set of data about a sample. It may refer to the population as a whole, i.e. a census study or any other well defined population with clear boundaries (CCS, 2003: 104).
Inferential Statistics are concerned with inferences that can be made about population indices/characteristics on the basis of the corresponding indices obtained from samples drawn randomly from the population. (Welman, Krauger and Mitchell, 2005:236). Put in another way, inferential statistics are the appropriate tools of analysis when it is required to make inferences or deductions from the characteristics of a sample to the characteristics of the population from which the sample is drawn. They establish the extent to which the findings or information obtained from the sample can be assumed valid for the population.
It should be noted however that the distinction between descriptive and inferential statistics is not in the techniques used but whether the analysis is concerned with the population (census) or the information is derived from the characteristics of a sample (CCS, 2003:104).
In this paper, we would briefly look at the following statistical analyses;
- Measures of Central Tendency
- Measures of Dispersion
- Measures of Association
- The Use of the Chi-Square (X2) Analysis
MEASURES OF CENTRAL TENDENCY
The measures of central tendency describe the entire data with the use of one central value, score or figure. The central value used therefore sums up the data.
The three commonly used measures of central tendency are:
- the mean
- the median, and
- the mode
The mean is what most people refer to in plain language as ‘average‘. The mean is obtained by adding up all the values in the data set and dividing the sum by the number of observations.
Formula: X = ∑X
X = Mean
∑ = ‘sum of’
X = Score
n = No. of observations/items
For example, to calculate the mean budgetary allocation to 10 states of the Federation, we add up allocation to each of the ten states and divide by 10.
However, the mean has a major weakness: it is very sensitive to extreme scores. Following our example above, it is possible that no single state has an allocation which is the same as the mean. A good illustration of this is reflected in the GDP – Gross Domestic Product, which is the mean value of goods and services produced by each member of the population in a year. Actual income distribution may not be identical to GDP per capita.
The median is the single value in the mid-point of a set of scores. One half of the total scores fall above it and the other half fall below it. To calculate the median, i.e. the middle value, the scores have to be arranged into ascending or descending order, from lowest to highest or the other way round. If the number of scores is odd, the median is the middle item. If the number of scores is even, the median is the average of the two middle scores.
Mathematically, the formula for the median when the number of items is odd is:
(n + 1)
Where n = number of items
For example, 4, 6, 7, 8, 10, 13, 14 (odd number of items, 7)
Median = 7 + 1 = 8 = 4th item = 8
For an even number of items, say:
2, 4, 5, 7, 8, 9 the median will be the average of the 3rd and 4th items,
i.e. 5 + 7 = 12 = 6
In the example of budgetary allocation to 10 states of the Federation, the median will be the sum of the 5th and 6th allocations when arranged in order of magnitude.
Though the median is less accurate than the mean, it is a better measure where extreme scores are involved. For example, assume the least allocation is
10m while the highest allocation is N 100m. N
The mode is the item that has the greatest frequency, the score that occurs more than any other score. Data sets may also be bimodal and multimodal.
Example: 45, 45, 47, 53, 57
The mode in the above distribution is 45. It occurred twice while others occurred only once.
Example of Bimodal data distribution
45, 47, 47, 53, 57, 63, 69, 69, 77, 84
The modes in the above set of data are 47 and 69 because they occurred twice whereas others occurred once.
Example of Multimodal Data Distribution
Assume a sample size of 15 employees in an enterprise or Government Agency with Hausa: 4; Ibo 4; Yoruba: 4; Edo: 2; and Itsekiri: 1.
The modes are Hausa, Ibo and Yoruba.
The mode has the advantage that like the median, it is not unduly affected by extreme values. However, the weaknesses lie in the facts that not all data sets may have a mode and some data sets may contain too many modes to make the mode a useful measure of central tendency.
MEASURES OF DISPERSION OR VARIABILITY
Measures of dispersion show the extent to which scores in a data set differ from one another. It is a measure of internal homogeneity or heterogeneity. They are used together with measures of central tendency in making a clearer sense of data. The relevance of measures of dispersion can be appreciated in a situation in which two data sets have the same mean but with differing group characteristics.
The GDP is again relevant here for an explanation of measures of dispersion. The GDP is the average value of goods and services produced by each member of a nation’s population in a year. International Financial Institutions often compare nations on the basis of GDP values in order to categorize nations as High-Income, Medium-Income and Low-Income. Countries in a particular category may share similar GDP but income distribution in individual countries may vary greatly. Some countries in the same grouping may be relatively egalitarian while others are greatly inegalitarian.
Another kind of measure is required to explain the differences in individual countries that may belong to the same GDP size grouping. The IMF, for example, uses the Gini Coefficient to measure income inequalities.
TWO MOST COMMON MEASURES OF DISPERSION
The two most common measures of dispersion are:
- the range, and
- the Standard Deviation
The range is the difference between the highest score and the lowest score. The range between the highest salary of about
3.5m basic pay for Mr. President per annum and about N 108,000 minimum pay for a Federal public servant is about N 3.4m. N
However, the usefulness of the range is limited. Two groups of public servants, say, the political officer holders group and career officers group may have the same range, but the degree and pattern of variability in each distribution may differ. Again, an extreme value can give a distorted view of the variability of a set of values. This shows that the range is not sufficiently reliable a reliable measure of variability.
THE STANDARD DEVIATION
The standard deviation is the most commonly used and most reliable measure of variability. It is defined as the average of all variations or distance from the mean score.
The formula for the standard deviation is:
S = √ ∑(X – X)2
n – 1
S = Standard Deviation
∑ = ‘sum of’
X = Score
X = Mean
n = No. of items
|Data Value χ
||χ – χ
||(χ – χ)2
|χ = 70 = 7
||∑(χ – χ)2 = 28
S = √ ∑(X – X)2 = √ 28__
n – 1 10 – 1
= √ 28__
= √ 3. 11
A large standard deviation means a very heterogeneous population while a small standard deviation means a homogeneous population. For example, measures of
variability can be used to assess the performance or ability of a given group, e.g. students. Where the standard deviation is small, it means the difference in inter-individual performance or ability is small and that the students have attained mastery in a unit of instruction. On the other hand, where the standard deviation is large, it means there are wide differences in the students’ performance, suggesting lack of mastery of the unit of instruction.
MEASUREMENT OF ASSOCIATION/RELATIONSHIP
While the measures of central tendency and variability are basic statistical concepts that summarise the distribution of a variable, how the values of one variable change in relation to the values of other variables is determined by measures of association. The measure of the degree or extent to which two variables are related is called correlation. In other words, correlation is the statistical technique for determining the presence or absence of association between variables of interest.
The degree of relationship between two variables is usually expressed as a coefficient called correlation coefficient, ‘r’.
The values of ‘r’ vary from -1 through 0, to +1. Therefore, with correlation analysis, we are not only able to know the magnitude of the relationship, but we are also able to determine the direction of relationship between variables.
A relationship could be strong or high (i.e. close to -1 or +1); or, weak or low (i.e. far away from -1 or +1); positive (i.e. direct relationship, as one variable increases, the other also increases); negative (i.e. as one variable increases, the other decreases). When ‘r’ is -1.0 or +1.0, it means perfect relationship. Generally speaking, a correlation less than or up to 0.2 is regarded as very weak; correlation between 0.2 and 0.4 is weak; correlation between 0.4 and 0.6 is moderate; between 0.8 and 1 is very strong.
There are different correlation coefficients for different types of data. When data are at interval or continuous level, the Pearson Product Moment Correlation Coefficient is applicable. When data are ranked, the Spearman Rank-order correlation coefficient is appropriate.
I do not wish to bore you with calculation examples. Interested participants may explore further.
Correlation analysis is appropriate in issues such as education and income levels, economic growth rate and unemployment levels, tax levels and savings levels; age and crime rate; age and income; size of a city and crime rates, and so on.
However, correlational relationship is not perceived to be causal. In other words, whereas two variables may be related, it does not necessarily mean that one causes the other to occur. For example, crime rate may be related to age but it does not mean that age causes crime to occur. The State of the economy may be related to incidences of strike action but it does not mean the state of the economy causes strike action; it is often a function of trade union leadership.
THE USE OF THE CHI-SQUARE (X2)
The Chi-Square analysis is appropriate when data involving people, events or objects can be divided into categories, frequencies or grouped forms such as male-female, voter-nonvoter, Yes-No, like-dislike, etc.
The Chi-square test can be applied in two broad situations:
- Determination of differences in opinions, attitudes, preferences, performance, etc. on the basis of position.
Example: Sex and smoking, religious affiliation and party preference, religious affiliation and suicide cases, socioeconomic background and attitude to privatization, socio-economic background and attitude to increases in the prices of petroleum products, education and the use of contraceptives, ethnic background and crime involvement, socio-economic background and attitude to removal of Nuhu Ribadu as head of EFCC.
- To find out whether the population of interest is distributed according to a given theoretical distribution model.
Example: To determine whether or not the admission of students into a particular university ‘fits’ or meets the expected criteria; to determine whether the distribution of bank loans to small scale enterprises fits the Federal Government Policy Guidelines.
Thus, the use of the Chi-square is said to be useful as a test of independence and homogeneity or as ‘goodness-of-fit’ test based on the two categories of application stated above, respectively.
REPORTING OF PROJECT FINDINGS
The form in which presentation of project findings takes, the dimensions or aspects reported as well as the depth of the report, all depend on purpose, audience and time. From the standpoint of journalism, project results may be presented in the following ways:
- As tables
- As graphs
- As statistical summaries: means, median, mode, standard deviation, correlation coefficient, etc.
- As selected quotations, e.g. writing representative powerful statements from responses by respondents in interviews,
- As feature articles,
- As editorials, and
- As reviews, etc.
Research ethics refers to the code of conduct of researchers in terms of what is right and wrong. The research report must avoid breaches of ethics, fairness, factual and contextual accuracy (Houston et al, 2002:538). Some of the ethics governing the social researcher, including investigative journalists, include:
- Honesty, absolute honesty in the pursuit of the truth. Data must not be changed or manipulated. Researchers ought to indicate any limitations and constraints in the conduct of the research, including their areas of specialization and degree of expertise, which may affect the quality and accuracy of data and findings.
- Research findings should be reported truthfully without bias. For example, the contents of documents should be reported without amendments or distortion. Facts, quotations or entire stories should not be invented if they do not exist.
- Confidentiality. In principle, no confidential data, which may harm respondents, should be recorded or published. However, the critical consideration in deciding whether or not to publish a confidential document is to ask and answer the question: Can the harm to an individual or institution be justified by the good that flows to society at large (Houston et al, 2002:541).
- Informed consent of participants & Respondents. For example, ‘ambush interviews’ (i.e. taking respondents by surprise) should be avoided. Such an approach may lead to ‘no comments’. ‘No comments’ may not necessarily point to or take the researcher to establishing the truth and investigative reporting is not sensational journalism. Only if documented persistent requests for interviews have failed could ‘ambush interview’ be justifiable.
- Avoid plagiarism. Plagiarism is an act of passing off as one’s own ideas, facts, findings, perspectives or writings which actually belong to other people. Ideas, findings, information, etc should be properly attributed to their appropriate sources.
- Houston et al (2002) also admonish journalists generally, whether project or investigative reporters or beat writers, never to accept tangible gifts, travel or free meals from sources.
- Journalists also ought to avoid investigating areas or issues that may impact negatively on the community/society.
- Exercise cultural sensitivity.
Aborisade, F. (1997). Research Methods: A student handbook. Lagos: Multifirm.
American Psychological Association (2005). Publication manual of the American Psychological Association. (5th ed.). Washington D.C.: American Psychological Association.
Brynard, P.A. & Hanekom, S.X. (2006). Introduction to Research in Management-related Fields. Pretoria: Van Schaik.
Centre For Civil Society (CCS). (2003). An activist’s guide to research and advocacy. Durban: Centre for Civil Society, University of Natal (Now University of Kwazulu-Natal).
Coldwell, D. & Herbst, F. (2004). Business research. Cape Town: Juta
Ganiyu, M. (2004). The Reporter’s Companion. Ibadan: Emgee Books.
Houston, B., Bruzzese, L. and Weinberg, S. (2002). The Investigative Reporter’s Handbook. Boston: Investigative Reporters and Editors.
Leedy, P.D. (1980). Practical research. New York: Macmillan
Nworgu, B.G. (1991). Educational research: Basic Issues & methodology. Ibadan: Wisdom.
Spark, D. (1999). Investigative Reporting: A Study in Technique. Oxford: Focal Press.
Welman, C., Kruger, F., Mitchell, B. (2005). Research methodology. (3rd ed.). Cape Town: Oxford University Press.
DETERMINING SAMPLE SIZES
(no of respondents)
1 Being Paper Presented at ‘Capacity Development Training on Investigative Reporting for Nigerian Journalists’ Organized by Wole Soyinka Award For Investigative Reporting, February 8 – 11, 2008 at the Development Support Centre (DSC), 3, Rev. Oyebode Crescent, Opposite Rotary Club House, off NTC Road, Iyaganku, GRA, Ibadan