Management research methodology integration of principles, methods and techniques

This book is designed as an introductory textbook on management research methods. Research is a creative process and the topic of research methodology is complex and varied. The writing of a book like this is beset with over-structuring and simplification. The basic premise for writing this book is that research methods can be taught and learnt. The emphasis is on developing a research outlook and a frame of mind for carrying out research. Competence in mathematical and statistical techniques of analysis is not necessary for understanding this book; however, some competence is required for carrying out research in management

About the Authors

Part A Scientific Method in Management Research1 Scientific Method

Defining ResearchScientific EnquiryScientific Method

Formal Science and Empirical ScienceLogic of Scientific Method

Hypothetico deductive MethodModels

2 Overview of Research in ManagementScientific Research in Management

Research Problem IdentificationResearch Problem Definition

Generation of Hypotheses

Formulation of Research ProblemsResearch Design

Classification of DesignsIssues of Research DesignResearch Design Process

Selection of the Type of Research

Measurement and Measurement TechniquesSelection of Sample

Selection of Data Collection ProceduresSelection of Methods of Analysis

Decisional Research with Mathematical ModelsSome Philosphic Issues of Management Research

Consultative Approach to Management ResearchErrors in Research

SummaryAnnexure 2.1Suggested ReadingsQuestions and ExercisesPart B Research Problem

3 Problem Solving

General Problem SolvingWhat is a Problem?Types of ProblemsProblem Solving ProcessLogical Approach

Soft System ApproachCreative Approach

Thinking ProcessCreative Thinking

Creative Efforts in ResearchBarriers to Creativity

Creative Problem Solving ProcessDevelopment of Creativity

Group Problem Solving Techniques for Idea GenerationIntroduction

BrainstormingDelphi MethodSummary

Annexure 3.1—An Illustration of a Case of Application of SSMSuggested Readings

Questions and Exercises

4 Formulation of Research Problems

Approaches to Management Research ProblemManagement Problem is Posed to the ResearcherInvestigation of an Idea by an Experienced ResearcherPilot Study

Initiatiation of a Novice/Student to ResearchExploration for Problem Identification

Literature SurveySystem Study

Errors of Problem Identification in ResearchHypothesis Generation

Characteristics of a Good HypothesisOrigins of a Hypothesis

Process of Hypothesis Generation

Hypothesis Generation Using Qualitative MethodsFormulation of The Problem

Model Building Context

Decision Maker and His ObjectivesEnvironment

Alternative Courses of ActionScenarios and Structural Modelling

Interpretive Structural Modelling (ISM)Formulation of Effectiveness FunctionSummary

Annexure 4.1—An Example of TaxonomyAnnexure 4.2—An Example for Meta Analysis

Annexure 4.3—An Illustrative Example of Theoretical FrameworkAnnexure 4.4—Examples of Hypothesis Generation

Annexure 4.5—System Study and Problem Formulation–Allocation of Assembly Manpower(Karthikeyan 1986)

Annexure 4.6Suggested ReadingsQuestions and Exercises5 Research Proposal

Research Proposal

Purpose of a Research ProposalTypes of Research ProposalsDevelopment of the Proposals

Formatting the Research ProposalContents of the Research ProposalRequirements of the Sponsoring AgentEvaluation of Research ProposalsSome Implicit ConsiderationsSummary

Annexure 5.1—Sample (Real) Research Proposal (Krishnaswamy et al, 1997)

Suggested ReadingsQuestions and Exercises

Part C Research Design—Types of Research6 Experimental Research

Experimental ResearchPrinciples of ExperimentLaboratory Experiments

Difficulties of Performing Laboratory ExperimentsDesign of Laboratory Experiments

Execution of Laboratory ExperimentsStrength and Weakness of ExperimentsErrors in Experiments

Experimental Designs

Basis of Experimental DesignBasic Designs

Statistical DesignsField Experiments

Quasi-Experimental DesignsQuasi-Experimental Designs

A Comparison of The Two Quasi-Experimental DesignsUse of Quasi-Experimental Designs

Action Research

Defining Action Research

Process of Action Research

Comparison of Action Research with ExperimentsScientific Merits of Action Research

Validity and Reliability of Experiments and Quasi-ExperimentsConcept of Validity and Reliability

Validity in Experimentation and Quasi-ExperimentationValidity of Quasi-Experimentation

Sources of Invalidity of Experiments and Quasi-experimentsChoice of Experimental Design

Analysis Procedures Used in Experimental DesignSummary

Annexure 6.1—A Laboratory Experiment

Annexure 6.2—A Randomised Two-Group ExperimentAnnexure 6.3—Solomon Four-Group Design

Annexure 6.4—Factorial Design

Annexure 6.5—Randomised Block DesignAnnexure 6.6—An Action Research CaseSuggested Readings

Questions and Exercises7 Ex Post Facto Research

Ex Post Facto Research by ObjectiveExploratory Research

Historical ResearchDescriptive Research

Ex Post Facto Research by Nature of StudyField Studies

Survey Research

Qualitative Research MethodsCase Study Research

Participant ObservationEthnographic MethodsCritical Incident TechniqueRepertory Grid Technique (RGT)

Some Additional Qualitative Research MethodsTriangulation

Analysis Procedures for Qualitative DataEvaluation Research

Annexure 7.6—Example of Cognitive MappingSuggested Readings

Questions and Exercises

8 Modelling Research I—Mathematical ModellingIntroduction

Mathematical ModelsWhat is a Model?Development of ModelsPrinciples of ModelingPatterns of Model BuildingUse of Analogy in ModellingModels as ApproximationsData Consideration in ModellingModels as Heuristic InstrumentsSolutions of Models

Questions and Exercises

9 Modelling Research II—Heuristics and SimulationHeuristic Optimisation

Definition of HeuristicsWhy Use Heuristics?Heuristic Methods

Heuristics Problem-Solving ApproachesMeta-Heuristics

Choice of Heuristic MethodsEvaluation of Heuristics

Evaluation of Heuristics in Empirical AnalysisSources of Problem Instances

Performance Measures/Measure of EffectivenessExamples of Heuristic Optimisation

Advantages and Limitations of Heuristic MethodsSimulation Modelling

Meaning of SimulationWhat is Simulation?

Classification of Simulation ModelsThe Process of Simulation

Key Steps in Simulation Experiments

Validation of Simulation Models/ExperimentsSummary

Annexure 9.1—Demonstration of Constructive Heuristics and SA (Simulated Annealing)Annexure 9.2—Illustration of Heuristics

Annexure 9.3—Illustration for Empirical Evaluation of Greedy HeuristicsAnnexure 9.4—Illustration for Monte Carlo Simulation

Annexure 9.5—Illustration for Simulation from Actual ResearchSuggested Readings

Questions and Exercises

Part D Research Design for Data Acquisition10 Measurement Design

Primary Types of Measurement ScalesNominal Scales

Ordinal ScalesInterval ScalesRatio Scales

Judgment MethodsFactor ScalesSummary

Annexure 10.1—Illustrative Example: Content Validity

Annexure 10.2—Illustrative Example: Concurrent and External ValidityAnnexure 10.3—Illustrative Example: Construct Validity

Annexure 10.4—Illustrative Example: Reliability in MeasurementSuggested Readings

Questions and Exercises11 Sample Design

IntroductionSampling Process

Non-Probability SamplingProbability Sampling

Simple Random SamplingStratified Random SamplingCluster Sampling

Systematic Random SamplingArea Sampling

Determination of Sample SizeRequired Size/Cell

Use of Statistical Models

Bayesian Method for Determination of Sample Size

Illustrative Examples of Sample Size DeterminationSummary

Suggested ReadingsQuestions and Exercises

Part E Acquisition and Preparation of Research Data12 Data Collection Procedures

Non-Observation ErrorsObservation errors

Validity and Reliability of Data Collection ProceduresValidity and Reliability of Interviews

Validity and Reliability of Observation

Validity and Reliability of QuestionnairesSummary

Suggested ReadingsQuestions and Exercises

13 Data Preparation and Preliminary Data AnalysisIntroduction

Data PreparationEditing DataCoding Data

Transcription of Data (Transcribing)

New Variable/Functional Combination/Splitting FormData Description

Summarising StatisticsExploratory Data Analysis

Stem and Leaf DisplayBox Plots

Data Mining

Statistical EstimationContent Analysis

Some Recent DevelopmentsExample of Content AnalysisSummary

Suggested Readings

Questions and Exercises

Part F Data Analysis and Reporting

14 Hypothesis Testing—Univariate AnalysisIntroduction

Logic of Hypothesis TestingNull Hypothesis

Research Hypothesis

Errors in Hypothesis Testing

Identification of an Appropriate Test for Hypothesis TestingParametric Tests

F-Test for Analysis of VarianceNon-Parametric Tests

Chi-Square TestMcNemar Test

15 Bivariate Analysis and Hypothesis TestingIntroduction

Simple Linear Regression Model

Fitting of a Simple Linear Regression ModelNon-parametric Methods of Association

Spearman’s Rank Correlation Coefficient (rs)Kendall’s Tau

Contingency CoefficientSummary

Suggested ReadingsQuestions and Exercises

16 Analysis of Experimental DataIntroduction

Analysis of Single Factor ExperimentsSingle Factor Randomised Blocks Design

RBD Model

Latin Square Design

Latin Square Design Model

Completely Randomised 2 × 2 Factorial Design2 × 2 Factorial Design Model

Suggested Readings

Questions and Exercises

17 Multivariate Analysis of Data—Dependence AnalysisMultiple Regression

Variable Selection and Model Building

An Overview of Multiple Regression Analysis ProcedureVariants of Regression Analysis

Discriminant AnalysisIntroduction

AssumptionsThe Method

Testing Statistical Significance of Discriminant FunctionsCanonical Correlation Analysis

IntroductionThe ModelAssumptionsThe MethodSignificance Test

InterpretationPath AnalysisOther Methods

Conjoint Analysis

Automatic Interaction Detection AnalysisSummary

Suggested ReadingsQuestions and Exercises

18 Multivariate Analysis of Data II—Interdependence AnalysisIntroduction

Factor AnalysisIntroduction

Geometric Representation of Factor AnalysisThe Model

Factor Analysis versus Multidimensional ScalingCluster Analysis

Introduction

Methods of ClusteringReliability

Style and Composition of the ReportPrinciples of Thesis Writing

Format of ReportingFormat of DissertationsFormat of Research Reports

Format of Publication in a Research JournalReporting of Qualitative Research

Rules for Typing or Word ProcessingSummary

Suggested ReadingsQuestions and Exercises

Appendix A1—System Concept

Appendix A2—Analysis of Covariance (ANCOVS)Appendix A3—Some Research Findings on Creativity

Appendix A4—Some Further Group Problem-Solving Techniques

Appendix B—Sources of Information of Management and Social SciencesAppendix C—Formulae for Hypothesis Testing

Appendix D—Selected Statistical TablesBibliography

Glossary

Issues of Management Research

The Cone of Science

LEARNING OBJECTIVES

Upon completion of this chapter, you will be able to:Study the nature of management researchUnderstand the process of scientific enquiryDevelop definitions and hypotheses

Appreciate the principles of formal scienceStudy empiricism in scientific methodUnderstand the logic of scientific method

Study the inductive method for hypothesis generationStudy the deductive method for hypothesis testing

Study the hypothetico-deductive approach as the core of scientific methodAppreciate scientific attitude

Understand the current objections to the use of scientific method in management researchEnumerate alternatives to scientific method in management research

Management initiates and determines the activities of an enterprise It makes plants, offices,computers, materials, and equipment productive through human effort It gives competence andeffectiveness to organisations in rendering goods and services to society The aims of themanagement are to motivate the employees in the organisation to achieve a high degree of workperformance in competitive situations, utilise resources efficiently, and to provide high qualitygoods and services In trying to achieve these aims, the manager faces many hurdles andproblems, which he needs to overcome and solve He does this by taking appropriate decisions.

In this book, we assume the management to be synonymous with decision making, withrespect to all activities in an organisation, in all areas pertaining to an enterprise and itsenvironment These activities encompass human resources, technology, supply chain,production, marketing, accounting and finance, public relations, policies and strategies of thefirm They also include managerial functions like organising, staffing, planning, controlling, andinnovating.

Defining Research

Research is defined as a systematic, self-critical enquiry The enquiry is aimed at understanding athing or phenomenon or solving a problem When an enquiry is aimed at understanding, it istermed as basic or fundamental research, which pursues knowledge, and may or may not havepractical or commercial use When the enquiry is aimed at applying the available knowledge forpractical or commercial use, or for solving a problem faced in practice, it is termed as appliedresearch.

Research is a systematic, self-critical enquiry

Research is a systematic enquiry, whether scientific or otherwise Scientific research, on theother hand, employing scientific method, (to be dealt with later in the chapter) has well definedobjectives and methods, generates dependable data, reliable and unambiguous findings, andjustifiable conclusions.

Research which employs scientific method is scientific research.

Management research Research in management is primarily applied research, in the sense that

it is directed towards aiding the manager in his decision-making Research is carried out in theenterprise to solve managers’ immediate problems or help them in their predictive efforts fordetermining the future course of action or tackling an anticipated problem.

Management research is an applied research directed to aid the manager in his decision-makingand in understanding the decision-making process Management research may be reporting,descriptive, explanatory or predictive.

However, management research may be carried out in universities and research institutionswhere the primary objective of the researcher is to understand the phenomena of decision-making processes and their environments In this case, research tends to be basic or fundamental.The manager himself may carry out management research in the enterprise when he makessystematic enquiries Data/information is collected and analysed, depending upon his ownbackground and experience Such research may not be scientific but would be useful in decision-making However, more often, a hired outside specialist (management scientist or consultant), incollaboration with the manager, carries out the research Research in such cases is more scientificand also gives practically useful results The manager tends to check on the objectives, methods,and the terms of research to make it more useful to the firm, and within this framework thescientist makes a scientific enquiry to derive valid results In either of these cases, the researchprocess follows the same general steps However, in the latter, some problems may arise betweenthe manager and the scientist in the conduct of research (This aspect will be discussedin Chapter 4).

Management research may comprise studies, which are reporting, descriptive, explanatory, or

predictive (Cooper & Schindler 2000) Reporting type of research consists of furnishing data,

information, or statistics It may involve considerable skill in obtaining data from sources,abstracting the information from it, and evaluating the information thus obtained.

In descriptive type of research, the researcher may try to describe a single event or characteristic

through distributions or may try to relate a few events or variables through statistical analysis.The results cater to broader decision interests in the organisation, relating to policy,

administration, and the like Explanatory research explains the phenomenon Hypotheses and

theories mark this kind of research Statistical or Operations Research (OR) modelling may be

used in analysis Predictive research uses the type of modelling done in explanatory research to

forecast the occurrence of an event or events under certain conditions arising in the future; forexample, when a capacity addition/expansion of a plant would be desirable with the currenttrends of demand continuing or changing because of technological changes Predictive researchis particularly useful in planning the activities of a firm.

Whether the manager himself researches or depends on an outside scientist, he should have agood understanding of the processes of management research It would be ideal if he is carryingout the research himself, in order to get good and reliable results If he is collaborating with ascientist, it would be fruitful for both to interact closely.

Whether the management research is basic or applied, many diverse disciplines like socialscience, economics, psychology, administration, statistics, and mathematics merge into a theoryof management and decision-making Therefore, research in management tends to be complex.The rigorous natural science modes of investigation tend to become more difficult to apply inmanagement research There are also other factors that impede management research.

The competence and effectiveness of any firm is dependent upon the quality of its humanresources Managing the human component of an enterprise is the most important and centraltask of management (Likert 1967) In this context, social relations in an enterprise present manydifficulties in the application of natural science methods Severe problems arise in measurement,which cannot be rigorously carried out Motivation, attitude, stress, loyalty, cooperation, and soon, are not amenable to precise measurement Further, the openness of results, which is soessential in natural science, becomes well nigh impossible, as considerable resistance oforganisations and customers, coupled with patenting problems, exists Public funding formanagement research is not generally as widespread as for physical, biological, and engineeringsciences For these reasons, the application of scientific method in management research haslagged considerably behind other disciplines.

The main objective of this chapter is to outline the scientific thinking and method necessaryfor good management research It also highlights the concerns of many researchers andpractitioners who seem to advocate alternate perspectives for management research Researchprocesses, methods, and techniques are elaborated in subsequent chapters.

Scientific Enquiry

Definition of science Since science is dynamic, its definition, if one is attainable, must change

over time Therefore, it is more useful to obtain a common understanding of some agreedcharacteristic of science, rather than to attempt to define it rigorously.

The central goal of science is the enhancement of knowledge Bunge (1967) proposes thefollowing goals:

1.Advancement of knowledge and prediction

2.Mapping the patterns of various domains of facts (conceptual mappings)

3.Continuous improvement of its products through a set of partial models, using logical andempirical analysis

4.Metascience, which is the science of science itself.

Science, in the main, aims at developing more and more true patterns of reality gradually Itstarts with simple and partial models representing different aspects of reality—first itscomponents, then the relationships among them It then adds on more and more territories andfeatures Another major objective of science is to sharpen and improve its own methodology andtechniques for gaining knowledge of reality and of predicting it This is referred to asmetascience by Bunge (1967).

Process of scientific enquiry Science deals with nature and has grown out of natural

philosophy Bwad (1923) identifies two models of philosophical activities as a means of enquiry—speculative and critical (analytical) Speculative activity depends much on broad experienceand imagination Analytical activity, on the other hand, requires thoroughness, insight, andconcentration on detail It is clear that each must complement the other In science, as inmetaphysics, both the modes are necessary There is, however, a dominance of analytical activityin scientific enquiry and method.

In the method of redefinition, terms and statements are made more carefully in a language thatclassifies their meaning and can be communicated without distortion Consider, for example, theterm quality When controlling production in a plant it may be defined as the proportion ofacceptance (that is, the quality of production is high when rejections are low) When the qualityof a product is specified, the customer redefines quality as the degree of acceptance The lexical

definition is degree of excellence.

In explication, concepts are specified in a symbolic language so that they are unambiguousand precise, for instance, proportion is often denoted by % (percentage) as in

Worker percentage = (Workers/Total number of employees) * 100 = w%Percentage annual interest = (Interests per year/Investment) * 100 = i %

The third method, that of illustration, holds that the meanings of terms and concepts are to beaccording to ‘use’ Sometimes terms or concepts are defined for a specific study or research Forexample, the term ‘worker’ in this research means an employee of a factory, who is eligible forworkman’s compensations under the Workman’s Compensation Act.

However, sometimes the only way of classifying the meaning of a term or concept is tostipulate one For example, conditional reflex is a reflex action that has been induced by habit ortraining to follow a stimulus that is not naturally associated with it These are usually terms thatform part of the jargon of a particular discipline.

Science uses explanation as a way of understanding some aspect of the world But science isnot just explanation; it encompasses the ability to predict the behaviour of things and to obtaincontrol over them.

The classical view of Victorian scientists’ empiricism and positivism was that the centralcharacteristic of science was argued as the use of induction, as propounded by Francis Bacon andJohn Stuart Mill (Whitney, 1961, page 2 & 220) The strategy was that of accumulating facts andthat broader laws and theories would emerge automatically by the sheer volume of data The riskof being misled by false theories was considered to be minimal.

Subsequently, simple induction gave place to the hypothetico-deduction, that facts would leadto hypothesis (theory) and further investigations would lead to testing the hypothesis, thus,rendering science as a progressive process Popper (1959) introduced the concept of falsificationwhich engendered that scientists should strive to disprove every theory advanced, that is, sciencefalsifies wrong theories but does not prove any theory once and for all In other words, a theoryholds until it is disproved Further, Kuhn (1970) believed that science operates without muchchange over periods within paradigms in which implications of current theories are researched.As anomalies build up, the new progressively destroy old theories in a revolutionary manner.Thus a theory, is possibly only relative.

Science is knowledge, which is the ability to make true statements and to defend them A truthcorresponds with facts and is coherent with other truths, which are already established Theformer gives science an empirical base and the latter a systematic structure, fitting it with othertruths.

Scientific knowledge is gained systematically rather than by direct experience However, theorigin of scientific knowledge is the experience of the scientist or the communication of theexperience of others Knowledge originating from experience generally goes through a cycle ofscientific processes—like observation, perception, thought, language, concepts, classification,definition, constructs, theory and verification—before it is accepted as scientific knowledge (Fig.1.1).

Scientific knowledge is the knowledge gained systematically through a cycle of process:observation, perception, language, thought, concepts, classification, definition, constructs,principles, hypotheses, laws and theory and verification.

The processes of enquiry (Fig 1.1) involved in the transformation of experiential knowledgeto scientific knowledge is now briefly outlined.

Observation Much of what we know is derived from observation, either systematic or

haphazard The starting point of research is often an observation Observation of phenomenonundertaken as a part of scientific enquiry is scientific observation It is purposeful, related toscientific properties, systematically carried out in a phased way, and is subject to checks withregard to validity and reliability Observation may be direct (as in the case of studying individualbehaviour) or indirect (as in interviews and questionnaires).

Fig 1.1 Scientific knowledge acquisition

Concept Concepts are the basic building blocks of thought and communication Any

phenomenon is contemplated by using concepts but a concept represents only one aspect ofreality Concepts help in organising an observation and experiential knowledge When carryingout scientific enquiry, concepts have to be precise Values, leadership, costs, and a machinecentre are all concepts A machine centre, for example, can be thought of as a means ofconversion in a process, as a demand for investment in financial considerations, and as a sourceof failure in system maintenance Each one of these concepts, when investigated, would requiredifferent tools of analysis or methods of investigation Concepts are, thus, products of experienceand perception and are, therefore, inventions of the human mind They make it possible to thinkabout the same phenomenon in different ways.

A concept is a basic building block of thought and communication which helps in organizing an

observation or an experiential knowledge It is a product of an experience and perceptionrepresenting a degree of abstraction.

Concepts also represent different degrees of abstraction Consider the human resource conceptof a manufacturing company At the highest level of abstraction we have the human resourceconcept and at the lowest level we may consider a lathe operator [Fig 1.2] There are many othertypes of machine operators like millers, press operators, and so on A higher level of abstractionis a machine operator There are skilled workers in the company like assemblers, inspectors,viewers, and so on The next higher level of abstraction is a skilled worker Further, we havesupervisors to supervise these skilled workers, engineers and designers, who are technicallyqualified, to support them Thus, the next higher abstraction is technical manpower In additionto technical manpower, we have administrative, legal, human relations, purchase, and sellingmanpower This sums up the human resource of the organisation at the highest level ofabstraction In management, concepts are particularly useful in providing insights andunderstanding of perceptions, roles, and viewpoints of various actors in an organisation.

Classification No two objects are identical, but they may bear likeness to one another This adds

to the complexity of a phenomenon A fundamental philosophy of science is to reduce thiscomplexity so that a few general fundamental principles can explain them To reduce thiscomplexity, objects are classified together based on what they have in common, such as geneticsimilarity (having similar origin, for instance, sea foods, and cotton textiles), structural similarity(having similar constituent parts, for example, process layouts, and line organisations), orfundamental similarity (having similar behaviour, like schizophrenia).

Fig 1.2 Human resource concept—an example

Definition With each classification obtained, the members of the classification will provide a

definition of the term denoting that classification When a definition aims at showing themeaning of a term by indicating the application to which it refers, but not through other terms, itis called ostensive definition For example, we may ask which companies, from a list ofcompanies, have good industrial relations between their management and staff, and get it marked

by a trade union leader By its very aim, ostensive definition is of the lowest form and leavesroom for ambiguity Some definitions are internal, in the sense that they belong to the samelanguage system, for example, the language of statistics When the definition extends outside thelanguage system it becomes external All ostensive definitions are, thus, external Definitionsthat depend upon intentions are internal Ostensive definitions are not generally useful inscientific research but may be used in the initial stages Whether or not internal to a languagesystem, a statement whose truth is asserted but not considered liable to empirical challenge is adefinition.

The above definitions can be descriptive or lexical, operational or mathematical Table1.1 briefly outlines the functions, value, and use of various definitions.

Table 1.1 Definitions

Descriptive definition: When one term is defined using other terms, the definition is verbal.

Concepts may be defined in terms of other concepts Such definitions are descriptive definitions,for example, a vendor is a person or organisation supplying goods and services to anestablishment Concepts may be defined in terms of other concepts at a lower level or a higherdegree of abstraction, but usually definitions using lower level concepts are more useful andmeaningful In lexical definitions too concepts are defined in terms of other concepts (synonyms)

at the same level of abstraction For example, Firm – partnership for carrying out business;factory building – buildings with plant for manufacturing, and so on.

Operational definitions: These are mostly used in research and help in making measurements.

They are stated in terms of criteria for measurements so that they are unambiguous and precise.They must be capable of being counted or measured, or it must be possible to gather informationon them in some way The object to be defined may be physical (amount of investment) orhighly abstract (attitude), the first may be measured in monetary terms and the second mayrequire a set of questions to measure the multidimensional construct.

An operational definition is stated in terms of criteria for measurement so that it is unambiguousand precise, which is mostly used one in research.

Mathematical definitions: are expressed in terms of symbolic expressions.

For example: Inventory holding cost for an item = (Q × c × i/2)

WhereQ = quantity purchased in a cycle

c = cost per unit of the item

i = interest rate on capital locked up in the inventory.

A conceptual scheme, a precise language, and definitions are required to enable scientificobservation and classification of information collected But all these operate together and in acyclic manner.

Construct Refinements and redefinitions of familiar concepts to suit a particular discipline in

order to describe the operations of the phenomena relevant to the discipline, lead to ‘constructs’.Thus, constructs of a particular discipline are concepts, which are clear and precise after beingshorn of ambiguity and vagueness by rigorous redefinitions and refinements Constructs linkedto the perception of scientists are called observables Observables can be enhanced and enrichedby instruments used for extending the range of perception of the scientist (for example, atelescope for an astronomer or a motion picture camera for making time and motion studies inindustrial engineering) A construct may not have a direct link to perception or observation andmay be purely speculative Such constructs are termed theoretical constructs.

A refined and redefined familiar concept to suit a particular discipline is called a construct It isclear and precise A construct not directly linked to perception or observation, is termedtheoretical construct.

Hypothesis Based on enquiry or insight or a limited observation of phenomena, a scientist may

make a proposition A proposition is the meaning of a declarative sentence used for the assertionof some relationship between concepts.

A hypothesis is a declarative sentence or proposition in which at least one empiricalgeneralization follows and states the existence of the size, form, or distribution of somevariables The relationship between a fact and its cause is expressed as hypothesis, which mustbe capable of being experimentally verified and must have a definite practical consequence.

If a series of observations are made on objects O, in order to determine whether or not theobjects in this class exhibit property P, and if in each case of O and so on, that is, O1, O2, O3,and so on, the property P is observed, the scientist moves from a declarative statement withrespect to each observation of O to a universal statement about class P, like all Os are P Such ajump is known as generalisation.

Hypothesis is any declarative sentence in which at least one empirical generalisation followsbut whose contradiction does not take the form of a protocol sentence Hypothesis is aproposition that typically states the existence, size, form, or distribution of some variable (Emory1976) In this form, a proposition can be tested and becomes a hypothesis If a declarativesentence whose consequences, when tested empirically, result in reality not leading to them itbecomes necessary to reject it What is scientific is, therefore, relative to the status of knowledgeat the time of making the hypothesis Hypotheses are rejected not because they are false butbecause they are irrelevant Usually, generalisations that are not confirmed are called hypotheses.They are only tentative, need to be confirmed, and are only just ‘working hypotheses’.Generalisations emerge but hypotheses have to be invented (Caws, 1965) In any case, ahypothesis that is confirmed indicates confidence in the repeatability of observations.

Any statement whose truth can be tested by observing the world is an empirical proposition Ahypothesis is a proposition that can be experimentally verified, and has a definite practicalconsequence Most hypotheses have the following features (Weatherall, 1968).

1.Logical relationships likeAll, some, or noneIf, unless

Greater than, equal to, less thanProportional to

2.Quantitative relationship

3.Spatial relationship

Descriptive hypotheses: These describe properties.

1.“Current unemployment in India is greater than 15 per cent.”

2.The marketing manager and the financial manager of a firm do not agree on the quantity offinished products to be maintained as safety stocks.

Relational hypotheses: These describe the relationship between two variables.

1.“Parents with higher education spend more on the education of their children,” while anassociative relationship is implied, a cause-effect relationship is not implied.

2.The greater the employee welfare measures provided by the management of a company, thesmaller the labour turnover of skilled workers.

Explanatory hypotheses: These indicate a cause-effect relationship The direction of relationship

is important and should be interpreted properly.

When the salaries of government employees increase, their families spend more on theirclothing The direction of relationship is: increase in salary → purchase of clothing But thereverse is not true, that is, it cannot be said that by purchasing more clothing salaries can beincreased But in all cases the direction of relation is not so obvious.

Generally, good hypotheses and associated experiments are designed to win both ways.Hypotheses may be defective because they are indefinite, narrow, semantically faulty, logicallyfaulty, and related to unattainable conditions In developing hypotheses, argument by analogy isa great spur to imagination, but has no justification (cf inductive inference).

In confirming a hypothesis, tests of statistical significance measure the strength of evidencethat a difference between the hypothesised and actual relationship is not fortuitous However, itgives no information about the importance of the difference The errors that are usuallyassociated with hypothesis testing are Type I error-insignificant is taken as significant or Type IIerror—lack of significant difference is taken as evidence of no difference.

Laws An empirical generalisation, either affirmative or conditional, accepted as true becomes a

law Laws are based mostly on partial and intermittent observations and, therefore, a certaininexactitude is the price to be paid if the law covers a wide territory A statement whose truth isasserted and is empirically significant is a principle if it has theoretical terms or else it becomes a

law Principles are a priori to the scientific system like in physics but they may be feasible in

sciences like the social sciences Principles and hypotheses are accepted as suitable startingpoints of theoretical work but are not observable All generalisations accepted as true are lawsand all hypotheses accepted as true are principles.

A theory is a set of asserted universal propositions communicated in a set of universal sentences(by several researchers) which are derived by observation and empirical evidence capable ofagreement and corroboration.

Theory Theory is a set of asserted universal propositions communicated in a set of universal

sentences Each scientist may have his own set of propositions However, several sets ofassertions of several scientists will be fully or partially isomorphic with each other Therefore, aunion of such sets would be a theory in an all-inclusive sense A theory is nothing but an outlook,systematically acquired and derived by observation and empirical evidence, capable ofagreement and corroboration.

Hypotheses, laws, and theories play a vital part in the scientific enterprise of explanation Thisis depicted in Fig 1.3 Scientific knowledge develops through theory building Theory andempirical evidence constitute every scientific effort Theory as a set of definitions, constructs,and hypotheses, which are systematically related and are used for explaining and predictingphenomena through testing against observable data The inferences thus drawn provide thenecessary evidence to accept the theory, modify it, or add to it as further research addshypotheses and constructs to the existing ones A theory, generally, constantly changes as newlaws and new propositions replace old ones At any point of time a theory serves as a guidelinefor the useful ways in which phenomena can be investigated.

Fig 1.3 Theory development process

Consider, for example, the business systems theory Ludwig Von Bertanalaffy (1969) set forthconcepts of open systems and a general systems theory He based his theory on living systems Aliving system is not a conglomeration of separate elements but a definite system having

organisation and wholeness It is an open system maintaining a constant state, while matter andenergy, that enter it keep changing (called dynamic equilibrium) The organism is influenced byand influences its environments and reaches a state of dynamic equilibrium This is analogous toa business organisation, which is a manmade system and has dynamic interactions with itsenvironments, customers, competitors, labour organisations, suppliers, government, andtechnology.

A business organisation is a system of interrelated parts working together with each other inorder to achieve business goals and individual goals Compare a business organisation to ahuman body, the skeletal and muscular systems correspond to operating elements, the circulatorysystem to staff function, the nervous system to the communication system in an organisation, andthe brain to the top management Such a viewpoint of business theory could provide a frameworkfor rational decision-making, as propounded by Simon (1960), and would indicate the desiredfocus for scientific enquiry on proper points for decision-making, for the benefit of both theorganisation and the individual (Johnson et al, 1980).

Formal Science and Empirical Science

Formal science (for example, mathematical sciences) is based on the axiomatic method Anaxiom is postulated as true and further axioms are deduced from it Thus, an axiom is ananalytical truth The only truths that logical terms deal with are analytical and a scienceembodying only such terms is called a formal science In formal science some of the followingterms are relevant.

A formal science is a mathematical science based on axiomatic method (an axiom is ananalytical truth) It embodies only logical and analytical terms and also has theorems.

Logical calculus is a system within which formal properties and relationships can be calculated.

(for example, Calculus) A deductive system is one that has interrelated statements out of whichsome follow from the others deductively An axiomatic system is either an axiom (not following

deductively from any statement) or an axiom has to result in a useful calculus and must be

independent Axioms, in general, should be consistent A theorem is a sentence arrived at by

using a set of axioms put together by means of acceptable transformation rules, inference rules,formation rules, or rules for specification of terms A series of sentences starting with axiomsand ending with a theorem is the proof of the theorem (Bullock Alan, et al (1990), pp.64-65

Empirical science draws inspiration from natural sciences (like physics, chemistry, geology,astronomy, and so on) and is based on facts The scientific method used in developing scientificlaws and theory uses the hypothetico deductive procedure (Popper, 1959) This procedureinductively evokes hypothesis by experience or by the study of a phenomenon and tests itdeductively using information from the phenomenon Science, in this sense, has empiricalmeaning.

The scientific method uses, both descriptive terms and logical terms, and axioms both logical,like in formal science, and syntactical (axioms that are empirical) Arguably, the best way to goabout science is to separate the mathematical form and the empirical meaning and argument(Braithwaite, 1973).

Logic of Scientific Method

A satisfactory scientific explanation is one that logically classifies the relationships between factsand not one that is merely psychologically acceptable Logic has two values, like simpleswitches, connoting either a Yes or a No In a logical system, deriving logical truth from onesentence to another is called inference In logical inference there is a need to start with a truesentence The relationship, which can be inferred in such a process, is called implication.

Deductive logic If the inferences in a logical system are certainties, they are called

demonstrative, and if they are probable, they are called non-demonstrative Demonstrativeinferences are deductive—drawing particular conclusions (true inferences) from generalprinciples The set of sentences deriving conclusions (last sentences) from other sentences iscalled an argument The set of sentences, reasons, or premises must be true for the conclusion tobe true, then the argument is valid Propositions are true or false; reasoning or argument is validor invalid Deductive logic is the study of validity and not of truth The following examplesillustrate this.

Inferences drawn from a general principle to particular conclusion constitute deductive logic.The set of sentences deriving conclusion is called an arguments Conclusions are trueinferences Deductive logic is the study of validity and not of truth,

Raman takes decisions

— Premise 1 (P1) True (T)— Premise 2 (P2) True (T)

— True (T)

Both the premises are true and the argument is valid and, therefore, the deduction is True.

2 All industrial organisations manufacture goodsIndia Systems is an industrial organisation

One of the most important applications of deductive logic is in the testing of hypotheses If weknow that the consequences of hypothesis are true then the hypothesis is true.

h > gg follows from h is true

g/hif g is true then h is true

This is the most frequently used argument in testing scientific hypothesis But this argumentsuffers from the fallacy of affirming the consequence (the consequence of one hypothesis may bethe consequence of another).

h > gg follows from h is true

g/hbut g is false, therefore, h is false

Inductive logic Francis Bacon introduced the use of inductive logic In induction, empirical

evidence or fact is the starting point Inference is drawn from the evidence in the form ofconclusions, which explain the evidence or facts An inductive conclusion may be one of themany possible explanations of the fact and is, therefore, only tentative It may also explain factsother than those observed as evidence Therefore, it is usual to refer to the relation betweenevidence and inductive conclusion as supporting, that is, the evidence supports the conclusion.When contradictory new evidence is observed, the conclusion has to be abandoned Take forinstance the following example:

In inductive logic inference is drawn from the evidence in the form of conclusions, which explainthe evidence or facts Inductive conclusion is one of the many possible explanation of the factsand therefore only tentative.

Statement: If productivity is high then the workers are motivated.

Deduction: Productivity is high in organisation X, therefore, the workers in organisation X, are

Inductive-deductive thinking Any research effort uses both inductive and deductive thinking.

The formulation of a hypothesis, not much emphasised in scientific method, involves the limitedobservation of a fact (evidence) and inference from that tentative cause The relationshipbetween fact and presumed cause is expressed in the form of a hypothesis that should be