Skip to content

Some Methodological Problems And Relevant Choice of Experimental Instruments in Public Administration


Christopher Ekpu, Ph.D Department of Public Administration University of Benin, Benin City, Nigeria, 

S.O. Uhunmwuangho, Ph.D Institute of Public Administration and Extension Services University of Benin, Benin City, Nigeria, & 

E. O. Akintoye, Ph.D Department of Sociology and Anthropology Benson Idahosa University, Benin City, Nigeria Email: 


The paper examines the empirical research is an integral part of human knowledge. It is as old as the academic consciousness of man. Without empirical research, no academic discipline can survive. As long as man’s urge to improve on existing conditions by actively exploring the unknown through empirical research continues, the onward march of human society to greater heights will not stop. The thrust of this paper therefore is to assert that going by the nature of Nigerian government and public administration which is a logical carry over from the pattern of politics in the third world, the difficulties of experimenting with human beings and of risking waste of public funds should be considered in any fruitful Public Administration research. The consequent of failure in any experiment research, which involves the public administration and the use of public funds are considerable; the political risks to the elected masters would often be greater than they would readily authorize. As our democracy matures, rather than render the empirical research, human beings that constitute the core of every position may become more entrenched and their role more legitimized. This must be especially true in the social sciences research but the application extends quite easily to even Public Administration studies. 

Keywords: Democracy, Empirical, Problems, Research, Demography, Analysis 


The experimental research is a method of data collection and testing of hypothesis through controlled experimentation with people. In this method of data collection, the researcher manipulates and controls one or more independent variable for variation in response to the manipulation of the independent variable (Black, 1996). 

According to Uhunmwuangho (2007, 5), in Public Administration, scientific research approaches administrative phenomena by studying, observing, analyzing, explaining and predicting. Administrative data are gathered, measured, tested and analysed such that findings and conclusions based on evidence are obtained. Considerable research studies have been carried out in Public Administration, using several approaches, perspectives, foci and methods. 

Regular experimental research inquiry began in Public Administration towards the end of the 19th century. The New York Bureau of Municipal Research could be said to have blazed the trail in terms of research methodology, investigation, production of reports and prescriptions of measures for administrative improvement of governments (Aghayere, 1997). Beginning from 1911, Public Administration became concerned with establishing itself as a science with distinct principles, methods, models and theories. However, Social Science research comprises and utilizes certain methodologies. 

The purpose of the experimental method according to Cramer (2004) is to clearly establish the independent effect of the independent variable on the dependent variable. For example, in a non-experimental research, we may proof that economic factors causes labour unrest. This statement however cannot be absolutely certain because economic factors alone may not be the determining forces that cause labour unrest. It might be cheer coincidence that labour unrest occurs when there is an economic crisis or there might be some intervening variables between economic crises and labour unrests. Or it might be that both labour unrest and economic crises are symptoms of some unknown causes. 

These research mysteries can be unfolded only in the experimental setting (Brown, 1980). It is therefore in experimental research that we can clearly see the effect of an independent variable on a dependent variable. However, it is almost impossible to conduct laboratory type of experiment in public administration researches. 

Therefore, experimental research in social science researches is extremely limited as compared with that of physical sciences, and there are four reasons for this. Firstly, theories are not well established in social sciences, and since experimentation is guided by theory, it will be difficult to carry out effective experimentation until our concepts and theories are more clearly and precisely formulated. 

Secondly, the inherent nature of social sciences places substantial difficulty in the way of factors and the extremes complexity of interrelations and reaction in environment in which human beings functions, clearly render it difficult to ensure that the experimenter is not over simplifying reality and is not combining or separating variables in an unrealistic way. It is far from easy if it is possible at all to design an experiment, which follows for a finely balanced interplay of human and social factors, which approximates reality to a satisfactory and measurable degree. 

Thirdly, the reality of experimental subjects and control in social sciences experimentation requires that whilst some variables are being manipulated, others are controlled and held constant. A good deal of doubt may be cast upon the pureness of reaction of human groups; to what extent is the experimenter sure that his experimental group or control group are reacting; “naturally” and wither the former displaying the “hawthorn effect” and the latter is “having the experimenter on”. In the Hawthorn case, the experimental group reacted more to the stimulus of being a special group than to other intended stimuli. 

It is not known, also, for example, for factory groups elaborately even if not maliciously to “pull the leg” of social science researchers. These objections are valid, however, only when the groups know that they are part of an experiment, but failing to inform the group or individuals that they are part of an experiment can raise grave ethical problems. The question of reconciling deception in the name of science with integrity, which is at the root of academic freedom, is a thorny one. 

Lastly, there is the difficulty of menstruation, what units do we use to measure? For example, what units do a researcher uses in measuring creativity in management or the quantum of resistance to change? (Babie, 1973). 


Researchers in the Social Sciences are required to report their research findings objectively. This requirement is not easy to attain. To overcome this problem and meet the objectivity criteria, researchers have to choose the data collection method that would generate data that would at the same time meet the objectivity criteria. Also, the methods of data collection should be reliable and valid. 

These criterions are met partially by the use of experimentation as a veritable tool in public administration research. Experimentation as a tool of data sourcing and testing of hypothesis utilizes experiments with people – their behaviour and attitudes to certain phenomena. 

It has the advantage of high degree of control of intervening variables that is present in other types of data collection methods in social science research. 

Consequently, the intervening questions for this paper include: how valid and reliable is the use of experimentation in social sciences studies? Are the findings through experimentation devoid of flaws or mistakes? If they are, in what ways or what or are the best method for conducting experimental research in public administration? And how do we limit or manage these problems? 

These things are clear even to the okada rider on the roads and the market women in their respective stores. They affect everyone. This is what the author consider the General Problem and it is important to flag off the Problem by reproducing known statistics and consequences related to this aspect of the problem. It is the only way to initiate an awareness of and sensitivity to the need for a policy-orientation to the study from its start. The logic of it is that, if the study eventually presented useful results or findings and recommendations based on these, it is at this public level that their implementation will take place and their impacts felt most. That is the logical ontology-epistemology-methodology circle, which points to the need for a tactical approach. 

Theoretical Exposition 

In an early study into the effect of literature stimuli on tendencies to register as a voter and to vote, H.F. Gosnell selected a number of districts and then matched them for nationality, sex, ratio, economic conditions and existing exposure to stimulus from political parties. The experimental districts received a variety of canvassing stimulus – information notices, specimen ballot papers, cartoon notices urging registration and voting – which the control districts did not receive. Gosnell was able to observe the effects of the different stimuli and to measure them by checking against the register of voters and the party lists of those who voted. This was a fairly rigorous experiment in the classical form, having matched experimental and control groups, and the manipulation of variables, the results of which were accurately measured. This experiment according to Dexter (2004) is capable of wider application, not only in the field of registration and voting, but in any area such as the utilization of social services, the donation of blood and the use of tuberculosis and cancer test units. 

In many parts of Africa, where people live in rural area with low level of education, where the will to change and to react to modern techniques may not bee well developed, the need for experimentation into the best methods of involving voluntary response from the people is great. For instance, any developing country which wishes to embark upon a mass family planning programme would be sure of success if it conducted experiments into effects of information stimuli on the population in question, to determine what sort of propaganda is likely to be most effective, what its intensity, freque3ncy and timing should be. This is a very different matter from commercial advertising experiments – although some of the techniques may be the same – because the changing demographic structures, which results from a successful mass family planning campaign, have repercussion on a very wide range of public administration aspects. For example, changes in birth rate affect the age structure of the population, and this will or should affect public policy and administrative decisions on such matters as health services, schooling, retirement ages and even military conscription (North, 1973). 

Secondly, the author of this paper is aware of only this example and that which probably most closely approaches strict experimental research at the core of public administration is an experiment in district administration, which was carried out in Malawi from 1949 to 1954 (Tufte, 2004). A separate administrative district – Domasi District was created and carved out a larger district. It was carefully selected and was as typical as possible, being neither too far from nor too near to an urban area, having a number of good and poor soils, and in other respects being as like other districts in that part of the country as could be selected. For example, the population density was fairly heavy; there were few non-African residents; communications, water supply, standards of health and literacy were all described as “far” or “average” there was a mixture of tribes and religions ; and there was a general lack of important maps, population and vital statistics, and soils types. Over a period of half a decade, a large number of experiments was carried out in the new district the results were measured and compared with other districts (what in effect were the control groups) and with what had been happening before. The experiments – must but not all of which we deliberate included such matters as increased and decreased staff rations; staff continuity; frequency of visits by administrative and professional officers; the co-operation of traditional leaders; tax registration; census and collection; soil conservation; self-help schemes; community development; adult literacy; and management exploitation of village forest areas (Frund, 1979). 

In the area of office administration, experiments were carried out in providing all field officers in the area with common office staff, accommodation and facilities; and the effects of this pooling on team work, co-ordination, economy and the speed of operation were observed. An interesting experiment was also carried out on the use of card indices to replace – to a large extent or to supplement the usual files of documents. The stimuli and the responses were, wherever possible, measured and the results were recorded. Although the research significance may not have been fully appreciated at the time, here was a deliberately designed experiment in which different and measurable stimuli were applied and the reactions measured, recorded and compared with the control groups. It is regrettable that little of what was learned from this experiment district appears to have been applied elsewhere (Casley, 1990). 

Domasi was an experiment whose methods could quite readily be applied elsewhere and as a type it renders it possible to make the frequent and numerous “trial and error” experiments in public administration much more logical and useful, and the results much more clear. What is needed is that the stimuli applied and the reactions be measured, recorded and compared with carefully selected controls (Harson, 2000). In effect, if the administrator who is anxious to change things would become more systematic in his method and realize the importance of measurement and controls, he would probably produce results which were much more fully understood and which therefore would be of greater value to others. He would, infact, change from being a mere experimenter to being a researcher experimenter, and it is here that the possibilities of a fruitful liaison between the practitioner and the academic become even more promising. The former could contribute his experience of experimental method, and his time-becomes measuring and recording, which a busy administrator rarely feels he is able to afford whilst the latter could ensure that the experiment was realistic and practical, and could point out factors which administrative experience teaches him to take into consideration. 

Methodological Problems in Research 

A good deal of the work of public administration is experimental in the sense that when it is embarked upon it is by no means known whether it will be a success and the initiators are, from the onset, prepared to scrap the work and try something else if it is not successful. This is the “trial-and- error” method (Myrdal, 2004). 

Although established routine administration is not within this category, much that is now can be considered experimental regardless of whether it is smaller or large in scope, whether it is the experimental layout of a new office, or whether it is of the political magnitude and administrative complexity of, say, the Central African Federation. Both are experiments. That is to say, we do not know from the beginning how they will work and whether they will accomplish what we aim at, so we try them, observe them, and if they do not work we modify them or scrap them and try something else. They resemble experimentation in the natural sciences to the extent that the experimental design is consciously established, some of the elements are manipulated as the experiment proceeds, and the results are observed. The absence of strict measurement and control makes them differ from more rigorous experimentation (Nowaezyk, 2008) . 

Some items seem to lend themselves to relatively frequent experimentation; for example, it is not unusual for some functions of government to be tried in a variety of ministerial portfolios before they settle down to their “control” allocation. 

In the words of Osuala (1993, 2), observation method refers to the selection, provocation, recording and encoding of that set of behaviours and settings concerning organisms with empirical aims. Osuala explaining further sees selection as the emphasis on scientific observers. Provocation recognizes the relationship between experimental intervention and observational method. Recording refers to the recording of events through the use of field notes or any other means while encoding refers to the process of simplifying through some data reduction method while empirical aims was only emphasizing the variety of functions observational methods can serve vis-à-vis description, hypothesis or theory testing, generation or hypotheses or theory testing. 

In some countries, according to Olayide (2001), including Britain, this happen with social services and with external technical assistance; and in the newer developing countries the planning functions are tried in the treasury, a separate economic planning ministry, the president’s office, or are scattered throughout a number of separate ministries such as Agriculture, Works, Education and Finance, before they become relatively well-established in one department . 

Another functional area of experimentation is civil service training, and this is particularly clear in Africa. After a long period of inactivity, and with the advent of independence, the need to train national civil servants became recognized as a matter of great urgency. In the absence of analogue situations elsewhere and in the past, which could be copied with a reasonable degree of forms of training were tried – mixtures of residential and on-the-job tuition, new teaching techniques, new teachers and if they were not successful they were scrapped and something, or someone else was tried. The general success of these schemes should not blind us to the fact that they were an experiment or rather a series of experiments (Obasi, 2002). 

Some functional areas are so well settled that, even if defects in their operation are clearly recognized, there is great reluctance to change established practice and to risk a new practice being a failure. Such areas include the administration of title to land, and matters connected with systems of inheritance and succession. In these matters continuity and certainty are of themselves fundamentally important-people need to now what the future is going to be so that they may conduct their affairs accordingly. Similarly, there is a reluctance to experiment with the broad structures of executive government – the structure of all the ministries together (as opposed to a single ministry), and the civil service. But in this case, the reluctance exists probably because the issues are so large and because any major change would be such a vast upheaval, with the result that no one dare to experiment for fear of bringing the whole edifice down about his ears (Collins, 2000). 

Research Into Experiments in Public Administration 

Any research into the type of experiment in the field of social science or public administration as outlined above, comes under this second heading. A good deal of historical research is of this nature, and there is a growing literature of administrative history, based on extensive investigation and dealing with a wide range of administrative experiments. In the word of Loftus (1998), the research project on public administration training by the institute of development studies, University of Sussex, is of this type, and its enquiries into training in India, Pakistan, Kenya and Zambia underline the experimental nature of the training programmes and institutions being studies. 

The value of this type of research according to Horden (1969), lies in the careful analysis and evaluation of the experiment-what went wrong and why, to what extent and why were they successful, how could the various components have been manipulated to provide a greater degree of success, can similar experiments safely be applied elsewhere? What, in short, are the lessons to be learned from the experiments, and how can these lessons be applied elsewhere and in the future. It follows that the more closely in time the analysis and evaluation follow the experiment, the more useful are likely to be the lessons learned. Indeed, there is much to be said for research into experiments being carried on at the same time as the experiments are being conducted, and this is an area in which close relations between the practitioner and the academic could be valuable. 

Experiments Having Public Administration – Implications 

Certainly, the field of experiments, which have public administration implications, is very wide indeed, and these implications are of two kinds. Broadly stated, experimentation in the natural sciences may be said to have the aim of working on the actual objects of administration; with the things which have to dealt with administrative, such as pollution, whereas experimentation in the social sciences often relates to the administrative means and their effectiveness, such as communications and leadership (Ghai, 1999) 

As governments and other public authorities play a greater and greater part in conducting or encouraging general research, having an experimental components, Amartya opined that so will the public administration implications of experimental research increase. The financing, planning and control of such projects are clearly part of public administration, not only because public money is being spent but also because the public has, presumably, a fairly direct interest in the research. 

For example, experimental research into the control of pollution could have quite wide public administration implications (Amartya, 2005). 

The control of pollution on a wide scale is the responsibility of central government upon whom will be brought to bear the pressures of interest groups – health authorities, local councils, fishing industries, tourist agencies; the means of control will usually be the result of experimentation; control may well require legislation followed by a system of public inspection and the expenditure of public funds (which have to be accounted for and audited); a number of administrative units will need to be established to administer the legislation: in short, the matter will become bureaucratized, and will become an integral part of public administration. According to Smith (1976) The bulk of files in the home office and Attorney- General’s office which deal with vivid – section indicates the way in which this type of experimentation has implications for public administration. 

The impact on administration of nuclear research experiments – aspects such as pressure groups, public (and sometimes violent) demonstration, security and secrecy, and international relations – is obvious, and the list of examples, need not be extended here. 

Experimentation in some other fields has import public administration implications of less direct nature. For example, the Hawthorn expedient in the 1930s and their human and social behaviour conclusions, had, or ought to have had, considerable effect on “man- management” in the public services (Bare, 2003). Again, the experiments of advertising researchers size, format, colour, frequency and location of the information being advertised – are important, not only for the departments of information and other public propaganda agencies, but for every governmental unit which advertises staff vacancies in the public press, produces and publishes an annual report, or issues posters and pamphlets about its activities. The experiments of psychologist researchers into small group dynamics are of interest to us because they cast light on behaviour in such environments as the office, the committee room and the classroom of training establishments. The experimental research of organization and management specialists yields a continuous flow of results, which are of direct interest to public administration. Again, examples are well known and we need not extend the list here. 

Much of this experimental research in fields closely allied to public administration, or rather the conclusions, and possible applications of the conclusion, ought to be more closely studied by administrators; we have much to learn from and much to gain from them. 

Strict Experimental Research in Public Administration 

Our pre-occupation so far with matters peripheral to the core of our subject – experimental research in Public Administration – matters such as practical, trial-and-error experimentation in administration, research into experimental and experiments having implications for public administration, suggests that the core is small and possible non-existent. This is because of the nature of our discipline, which like for example, geography – has no well defined boundaries, but rather lies within the area of overlap between a number of related subjects. Just as much experimentation in geography is strictly experimentation in geology, methodology and biology, so much experimentation, which could be said to be in public administration, is infact, in the fields of other social sciences. Therefore, I prefer to place these under the heading of experiments having public administration implication (Jones, 200) 

There is, however, a small group of experiments which, if not lying at the very core of public administration research, does seem to lie a little nearer to the core or centre than other experiments with which we have already dealt. The fact that the experiments is this small group do not have any immediately obvious connection the one with the other, not only suggests that there may be no strict experimental research in public administration, but reflects both the polyvalent nature of our discipline and the absence of any general theories in our subject (Kerlinger, 1973). Consequently, there are two aspects this paper would like to deliberate upon. 


The question of experimental research in public administration places us in a dilemma and as is usually the case, we must try to strike a balance between the opposing factors. 

On the one hand, since the raw materials and the clientele of public administration are the public, since he would be experimenting with human beings, and since he would be spending-and quite possible wasting- public money of which he is the trustee, the public administrator is naturally and quite properly reluctant to experiment in his job. The consequence of “failure” in any experiment, which involves the public and the use of public funds are considerable; the political risks to the elected “masters” would often be greater than they would readily authorize. 

Yet on the other hand, this very reluctance to “play around” with the public and with public money is an argument strongly in favour of experimental research prior to the introduction of any large-scale, complex or costly scheme. The “pilot project”, if carefully designed and conducted in as strict scientific experimental fashion as possible, is a most useful method of ensuring so far as one is able that one’ hypothesis are valid, and that unforeseen factors, relationships and responses come to light. Pilot projects, however, must be treated as serious experimental research, and the various stimuli and responses carefully measured and compared with controls; if they are not treated in this way then it will be impossible to demonstrate satisfactorily and honestly whether the project has been a success, and whether it should not lead to the full-scale project being contemplated. Also, the difficulties of experimenting with human beings and of risking waste of public funds should be considered in any fruitful public administration research. 

R e f e r e n c e s 

Abah, O.J.(2008) An Appraisal of the Application of Experimental Methods as 

an Instrument of Data Collection in Public Administration Research (Unpublished Seminar Paper), Ambrose Alli University, Ekpoma. Aghayere, V.O. and S.O.J. Ojo (1997), Research Methods in Social Sciences 

Ibadan, Stirling-Horden Publishers (Nig.) Limited. Amartya, S. (2005), Development: Which Way Now? Economic Journal, No. 

2 Vol. III Babbie, E.R. (1995) The Practice of Social Research, New York, Belmont 

Wadsworth Publishing Co. Bare, O. et al (2003), Research in Africa, Evaston: Northwestern University 

Press. Black J. and Champion D. (1996), Methods and Issues in Social Science 

Research (New York, John Wiley). Brown, S. R. (1980) Political Subjectivity Application of Methodology in Political 

Science, New Haven, Yale University Press. Cramer, D.(2004), Introducing Statistics for Social Science Research, London, 

Rontledge Casley, D.J. and Lury D.A. (1990), Data Collection in Developing Countries 

Oxford, Clarindon Press, Dexter, A. (2004), Elite and Specialized Interviewing, Evanston, Northwester 

University Press. Frund, J.E. (1979), Modern Elementary Statistics, New Jersey, Prentice-Hall. Ghai, D. (1999) Participatory Development: Some Perspectives from Grass-roots 

Experience. Unrised Discussion Paper, 5th June. Ghai, D. (1999), Participatory Development: Some Perspectives From Grass-root Experience (Unpublished Seminar Paper)Department of Political Science, University of Ibadan. Harson D.G. (2000), Handbook of Political Science Method, Boston, Holbrook 

Press Inc. 

Horden, R. (1969), Interviewing: Strategy, Techniques and Tactics, 

Homewood III, Dorsey. Kerlinger, F.N. (1973), Foundations of Behavioural Research, New York, Holt 

Press. North R.C. et al (1973), Content Analysis, Evanston: Northwester University 

Press. Smith B.L. et al (1976), Politcal Research Methods: Foundations and Techniques 

Houghton Mifflin. Uhunmwuangho S.O. (2007), Some Methodological Insight Into Social Science Research, Nigerian Journal of Citizenship Education Vol. 5 No. 2 Tufte, E.R. (2004), Data Analysis for Politics and Public Policy, New Jersey, 


Published inNumber 1Volume 4