Summary Research Methods for Business Students (Saunders et al.) - English version
Deze samenvatting is gebaseerd op het studiejaar 2013-2014.
Table of Contents
Chapter 1 – The nature of business and management research and structure of this book
Chapter 2 – Generating the research topic
Chapter 3 – How to critically review literature
Chapter 4 – Research philosophies and approaches
Chapter 5 – How to formulate a research design
Chapter 6 – Access to data and research Ethics
Chapter 7 – Sample selection
Chapter 8 – Secondary data usage
Chapter 9 – Collection of primary data by observation
Chapter 10 – Collecting primary data using interviews
Chapter 11 – Collecting primary data with questionnaires
Chapter 12 – Analysis of quantitative data
Chapter 13 – Qualitative data analysis
1 - The nature of business and management research and structure of this book
This book teaches the different steps one should take when conducting business and management research. It will help you to undertake a research project by providing a range of approaches, strategies, techniques and procedures. Throughout this book the term methods and methodology will be used. However some may think these terms refer to the same thing, they actually have different meanings. The term ‘methods’ refers to techniques and procedures used to obtain and analyse data, while ‘methodology’ refers to the theory of how research should be undertaken.
1.2 What is research?
People conduct research to systematically investigate things in order to enhance their knowledge, but research is not merely collecting data. Research is conducted when:
- Data are collected systematically
- Data are interpreted systematically
- There is a clear goal: to discover new findings
In order to systematically conduct research based on logical relationships, a researcher must provide an explanation of the methods used to collect data, prove why the results are meaningful and outline any limitations to the research. The goal of research is not only to explain, describe, criticize, understand or analyse something, but also to simply find a clear answer to a specific problem.
1.3 The nature of business and management research
Management research is different from other kinds of research because it is transdisciplinary (multiple studies are involved with it) and it is a design science. Moreover, it has to be theoretically and methodologically accurate, while at the same time being of practical relevance in the business world. The researcher Michael Gibbons has introduced 3 modes of knowledge creation:
- Mode 1 – creating fundamental knowledge
- Mode 2 – creating practical relevant knowledge, with emphasis on collaboration
- Mode 3 – creating knowledge that is mainly relevant to the human condition
Research that only emphasises Mode 1 ways of creating knowledge which only focuses on understanding business and management processes and their outcomes is called basic, fundamental or pure research. Another type of research is called applied research where the emphasis is more on Mode 2. In this case research is only being conducted direct relevance to managers and is presented in ways these managers can understand and act upon. Pure and applied research are two extremes, in order to successfully conduct business and management research there has to be a balance between the theoretical (Mode 1) and practical (Mode 2) part of research. The characteristics of pure/basic and applied science are summarised in figure 1.1 on page 11.
1.4 The research process
When doing research on needs to go through several stages, usually involving: formulating and clarifying the research topic, reviewing the literature, designing the research, collecting the data, analysing the data and finally the writing. However, it is not always necessary to pass through these stages one at a time. More frequently the stages in a research process will cross-refer to other stages, meaning that there is no linear line in the research process. Therefore it’s important to have a strong research topic and to revise ideas many times. See figure 1.2 in the book on page 14.
Chapter 2 – Generating the research topic
2.2 Characteristics of a good research topic
Before generating ideas for a research topic it is always useful to address the assessment criteria. The topic of research should be something that really excites the researcher and it should lie within his capabilities. These capabilities depend on constraints on time and financial resources, possession of the necessary skills and access to the relevant data. Moreover, it is useful for a researcher to have knowledge of the literature associated with the topic and to be able to provide bright insights.
It is important to have a symmetry of potential outcomes, which means that the result will have to be of similar value whatever you find out. If this is not the case there is a chance you find an answer of little importance. Also consider your career goals, consider how this research could be useful in your future career.
2.3 Generating and refining research ideas
There are many different techniques that can be used to generate research ideas. They can be divided into those techniques that involve rational thinking…:
- Examine own strengths and interests, choose a topic in which you are likely to do well
- Explore your university staff research interests
- Analyse past project titles of your university such as dissertations (projects from undergraduates) and theses (projects made by postgraduates)
- Discuss with colleagues, friends or university tutors
- Search through literature and media (articles in journals, books, reports). Review articles in particular, since they contain a lot of information about a specific topic and can therefore provide you with many ideas
…and those that are more based on creative thinking:
- Noting ideas down in a notebook
- Exploring preferences using past projects (see page 28 to know how)
- Exploring relevance of an idea to business using the literature, articles may be based on abstract ideas (conceptual thinking) or on empirical studies (collected and analysed data)
Most often it is a combinations of these two ways of thinking that leads to a good research idea.
There exist different techniques for refining research techniques, one of which is the Delphi technique. This approach requires a group of people who are involved with or share the same interest in the research idea to generate and pick a more specific research idea. Another way to refine a research idea is to is to turn it into a research question before turning it into a research project. This is called preliminary inquiry.
The integration of the ideas from the techniques is an important part of a research project. This process includes ‘working up and narrowing down’, which means that each research idea needs to be classified into its area, its field, and ultimately the precise aspect into which one is interested.
2.4 The transformation from research idea to research project
Writing research questions
It is very important to define a clear research question at the beginning of the research process. A research question may be:
- Descriptive – question usually starts with ‘When’, ‘What’, ‘Who’, ‘Where’, or ‘How’
- Evaluative – question may start with ‘How effective…’ or ‘To what extent….’
- Explanatory – question mainly starts with ‘Why’ or has this word in it
Do not make the research question too simple or too difficult to answer. The ‘Goldilocks test’ may be helpful to determine if a question is too big (when it demands too many resources), too small (provides insufficient data), too hot (when it is a sensitive subject) or ‘just right’. It is also essential for a research question to provide new insights.
Writing research objectives
Research questions can be used either to produce more detailed investigative questions or as a starting point for research objectives. Writing objectives is more generally accepted as a way to specify sense and direction in a research project than research questions. This is because they are more precise in displaying what one would like to make clear. Research objections operationalize the research question, which means that they show the steps that are required to take to answer it.
What is theory and why is it important?
Theory is concerned with causality. This means that it regards the cause and effect relationship between two or more variables. For example, theory explains why and how a promotion influences employee’s behaviour. Logical reasoning is essential here to explain in a clear way why this is the case. The role of theory is to explain the relationship between variables and to make predictions about possible new outcomes. Advising on how to take research in a certain way (Variable 1) is based on the theory that this will eventually create effective results (Yield B). By undertaking research it is possible to collect data with which new theories could be developed.
A research project is designed to either test a theory or to develop a theory. When someone is taking a clear theoretical standpoint and wishes to test this through the collection of data one is using a deductive approach. An inductive approach is used when someone builds a theory from the collected and analysed data. There exist three kinds of theories:
- Grand theories – Newton’s gravity theory, Darwin’s evolution theory etc.
- Middle range theories – these are significant, but they don’t change the way in which we think like grand theories do
- Substantive theories – focussed on a particular, setting, group, or time theories
2.5 Writing a research outline
The research proposal is a structured outline of a research project. Making a research proposal demands that you think through what you want to do and why. It helps you to guide the project through all of its stages. When producing a proposal think of these general criteria:
- A research project needs to be coherent, which means that all the different components of the project need to be in relationship with each other.
- It needs to be feasible as well. This means that the project should be possible to achieve.
Structure of a research proposal
- Title – This should summarise the research question
- Background – This is an introduction for the reader to the problem or issue, it gives answers to the questions ‘what is going to be done’ and ‘for what purpose?’. The background also shows the relationship between a theory and a particular context and it should demonstrate the relationship between the research and what has been done before in this subject area.
- Research question and objectives - the background should eventually lead to a statement of the research questions and objectives and the observable outcomes
- Method – This is the longest section and reveals how the research will be conducted. It consists of two parts: Research design and data collection. Research design is an overall overview of the chosen method and provides the reason for choosing this method. Here you will explain the choice for a certain research strategy and determine an appropriate time frame for the project . The section ‘data collection’ will specific how and where the data will be collected and will explain the various analysis techniques that will be used during the research.
- Timescale – In this section you will divide the research into different stages and explain how much time each stage will approximately take.
- Resources – In this facet of the proposal certain resource categories such as finance, data access and equipment will be taken into consideration. This section will also include the expenses that may be involved with these categories.
- References – This section consists of the literature sources to which you have referred to.
Chapter 3 – How to critically review literature
3.1 What is a literature review?
A literature review is a review in which one makes reasoned judgements about the value of pieces of literature. When doing this, it is necessary to organise valuable ideas and findings. There are two kinds of reviews. The first kind of review goes along with the initial search for research ideas, because that’s when you browse through pieces of work and judge which ones are relevant and which ones are not. The second kind of reviews are referred to as critical reviews. To be able to show the significance of a research project it is necessary to understand the subject field and its concepts, ideas and key theories. One is ‘critically reviewing literature’ when one chooses those pieces of literature that are relevant to the research.
3.2 Critical review
A critical review should be a constructively analysis that critically develops a transparent argument about what the chosen literature tells you about a research question. It should not simply summarize what a piece of literature is about. Rather, it is necessary to evaluate what is significant to the research project and what is not.
The goal of a critical review
Reviewing literature critically enables you to generate the foundation on which a research is based. The exact goal of reading literature depends on the approach one is wishing to use in a research. A deductive approach is when you develop a theoretical or conceptual framework which you afterwards test using data. An inductive approach is when one analyses the collected data to subsequently develop theories from them and relate them to the literature. The difference with an inductive approach is that you don’t start with predetermined theories and conceptual frameworks. There are three ways of using literature:
- Use literature in the initial stages of a research, when making research proposal
- Use literature to provide the theoretical framework and context
- To help place research findings within the wider body of knowledge
When a critical review is successful it will provide new insights about a subject area that no one has ever thought about. It is necessary to show how the new findings and developed theories relate to other literature about your subject to demonstrate that you are familiar with what has already been said about the subject.
Adopting a critical view of your reading
In order to read effectively it is necessary to master various skills, which include:
Previewing: Browsing the text to find out what its purpose
Annotating: Conducting an analogue with yourself, the author and the issues at stake
Summarising: Be able to explain/state the text in your own words
Comparing and contrasting: How has your thinking been altered by this reading?
Use review questions: Questions which you ask yourself during reading which are linked to your research questions.
Content of a critical review
The critical review will eventually have to appear in a project report. It has to include an evaluation of the research that has already been done in the subject area, demonstrate and discuss the relationships between published research findings and refer to the literature in which they were reported. Moreover, a critical review must present the key points and trends in a structured way and show the relationship with the research. By doing so the readers of a research project will have background knowledge to the research questions. When considering the content of a critical review one needs to:
- Include the key academic theories within the chosen research area
- Demonstrate that your knowledge of the chosen area is up to date
- Through clear referencing, enable those reading your project report to find the original publications which you cite
How to be ‘critical’
Being critical means that one needs to make reasoned judgments about a particular text, by evaluating a problem with good use of language. This means that one ’s own critical stance needs to be based on clear arguments and references to the literature. Being critical also means making a clear and justified analysis of the key literature of a research project.
The structure of the critical literature review
A literature review is a critical analysis and a description of what other writers have written. It is helpful to address to a literature review as a discussion of how far existing literature goes in answering the chosen research questions. One should therefore point out the limitations of the existing literature. It is also helpful to look at the way the review relates to the chosen research objectives. There are three common critical review structures:
- One single chapter
- A series of chapters
- Throughout the project report while tackling various issues
Every project report should refer to the key issues from the literature in the discussion and conclusions. Don’t let the review become an uncritical listing of previous research! It is easy to be critical when constantly tries to compare or contrast different authors and their ideas. The review should be a funnel in which you:
- Start at a general level and narrow it down to research questions and objectives
- Provide a brief overview of key ideas and themes
- Summarise, compare and contrast research of the key writers
- Narrow down to highlight research most relevant to the research
- Provide detailed findings and show how they are related to the literature
- Highlight those aspects where your research is providing new insights
- Lead the reader to subsequent sections of your project
3.3 Available literature sources
The available literature sources can be divided into three categories: primary, secondary and tertiary sources (See Figure 3.3 on page 82).
Primary literature First occurrence of a piece of work. Includes public sources as reports and documents, but also unpublished work such as letters and memo’s.
Most of the times this kind of literature is very detailed, but not easy to access, therefore it is sometimes referred to as grey literature.
Secondary literature Is aimed at a wider audience, easier to locate and better covered by tertiary literature. This includes books, journals and newspapers.
Tertiary literature Also called search tools, to locate primary and secondary literature. They include online search tools, databases, and dictionaries.
Especially journals are a essential literature source for virtually any research, since they provide a researcher with information which focussed on his subject area. Nowadays it is easy to access journals via online databases. Refereed academic journals only publish articles which are evaluated by academics before their publication. These articles are therefore characterised by their quality and suitability. Professional journals are made for their members by various organisations. Their articles are usually more of practical nature than those of refereed academic journals.
3.4 Planning a literature search strategy
If one starts his search for literature it is important to have clearly defined research questions, objectives and outline proposal. This prevents information overload. One should make a search strategy which includes:
- The parameters of the search
- The key words and search terms
- The databases and search engines you’re going to use
- The criteria to select relevant and useful studies
One way to start searching for parameters is to browse lecture notes and course textbooks and make notes for research question.
Generating search terms
It is important to read articles from key authors as well as recent review articles in the area of research. This will help generating key words. Recent review articles are sometimes also helpful to refine search terms, plus they will sometime refer to other work which may be relevant to your project. The identification of search terms is an essential part of planning a search for relevant literature. The definition of search terms is: basic terms that describe research questions and objectives and shall be used to search the tertiary literature. Different techniques for generating search terms are:
- Initial reading, dictionaries, encyclopaedias, handbooks and thesauruses
- Search on Google with ‘define: (enter term)’
- Relevance trees: constructed after brainstorming - see box 3.11 on page 96
3.5 Conducting literature search
While most it is very tempting to start a literature search with using a search engine such as Google, this must be handled with care, as the research project should be an academic piece of work and hence must utilise academic sources. Search should therefore be used to provide access to academic literature. Conducting literature search can be done by:
- Using Tertiary literature sources
- Obtaining literature referenced in books and journal articles you already read
- Using Internet: see Table 3.4 and figure 3.3
- Scanning and browsing secondary literature in the library
- Searching the Internet
3.6 Obtaining and evaluating the literature
Box 3.15 on page 108 displays a checklist of what should be done to evaluate the literature.
3.7 Recording literature
It is important to make notes of the literature one has read, because it will help thinking though the ideas in the literature in relation to the research. When making notes there are three sets of information one needs to record:
- Bibliographic details
- Brief summary of content
- Supplementary information
Chapter 4 – Research philosophies and approaches
4.1 Why is philosophy important
The way one chooses to collect data belongs in the centre of the research ‘onion’, as displayed below. The research onion depicts the aspects underlying the choice of data collection techniques.
4.2 Why research philosophy is important
Research philosophy is a term that describes the development of knowledge and the nature of that knowledge. Understanding research philosophy is important because the very purpose of research is also to develop new knowledge. It is not true that one philosophy is better than another, but they might be suited to achieve different things. Two major ways of thinking in philosophy are: ontology and epistemology (See table 4.1 on page 129).
A pragmatist is someone who thinks that concepts are only relevant where they support action. He believes that one philosophical position could be more likely lead to the answer to his research question than another. In addition, a pragmatist also believes that it is possible to work with multiple philosophical positions. According to a pragmatist there is not one way of thinking.
Ontology is a philosophical position that refers to the nature of reality. One aspect of ontology is objectivism. This means that things exist with a purpose independent of those social actors concerned with their existence.
Another aspect is subjectivism, which holds that social occurrences are created through the perceptions and consequent actions of the involved social actors. People who adopt a subjectivist way of thinking find it is necessary to explore the details of a situation to be able to understand what is going on. This is termed social constructionism.
Objectivists think that the culture of an organisation is something that an organisation ‘has’ while subjectivist tend to view the culture as something an organisation ‘is’. Management theory is leaning towards the objectivist way of thinking.
Epistemology regards what constitutes acceptable knowledge in an area of study. It addresses the questions: ‘What is knowledge?’, ‘How is knowledge acquired?’ and ‘What do people know?’.
The philosophy of positivism refers to the philosophical stance of a natural scientist. This philosophy holds that collecting data about an observable reality and searching for regularities and causal relationships will lead to the creation of a new theory or new generalisations. Other characterizations of positivism are:
- The researcher is independent of the subject of the research, he is value-neutral (his feelings are included in the research)
- Cyclical relationship between hypothesis testing and theoretical development
- Quantifiable observations that lend themselves to statistical analysis
Realism claims that whatever we sense is reality: objects exist without concern of the human mind. Therefore realism contradicts idealism, which states that only the mind and its contents exist. Just like positivism, realism also assumes a scientific approach to the development of knowledge. There exist two kinds of realism:
- Direct realism – what you see is what you get, what we perceive and experience with our senses displays the world in an accurate way.
- Critical realism – what we experience are sensations, images of existing things in the real world, not the existing things themselves. What we experience are mere illusions.
There is a difference between these two kinds of realism regarding the capacity of research to change the world. A direct realist would state that the world is relatively unchangeable whereas a critical realist would claim that the researcher’s understanding to that which is being studied could be changed. Many researchers claim that what we explore is just part of the bigger picture. Thus researchers usually adopt a critical realism point of view.
Interpretivism claims that it is necessary for researchers to understand the differences between humans in our role as social actors. We interpret our daily social roles in accordance with the meaning we give to these roles. Interpretivism stems from two intellectual heritages
- Phenomenology considers the way in which we as humans make sense of the world around us
- Symbolic interactionism: we are all in a continual process of interpreting the social world we live in and we interpret the actions of the people that interact with us. These interpretations lead to adjustments of our own meaning and actions.
It is important for a researcher to understand the world of his research subjects and to understand the world from their point of view.
Axiology is a strand of philosophy that studies judgments about value. This includes values in the fields of ethics and aesthetics. One’s own values play a crucial role in all stages of the research process. Our values are the guiding line for all our actions (Heron 1996).
The term paradigm is frequently used in the social sciences, but it often leads to confusion because of its many meanings. Here we define paradigm as a way of examining social occurrences from which particular understandings of these phenomena can be gained and explanations attempted. In Figure 4.2 on page 141 there is an image of how the four paradigms can be arranged:
- Functionalist paradigm – this paradigm is frequently used in business management. Functionalists assume that an organisation are rational entities, in which rational explanations will provide solutions to rational problems.
- Radical structuralist paradigm – this paradigm is concerned with understanding structural patterns within organisations (hierarchies for example) and reporting relationships and the extent to which these relationships may produce dysfunctionalities.
- Interpretive paradigm – when adopting this paradigm one is concerned with understanding the fundamental meanings attached to organisational life. Instead of rationalities this one wishes to discover irrationalities. In this paradigm being involved in the everyday activities of the organisation in order to understand and explain what is happening is more important that to try to change things.
- Radical humanist paradigm – this dimension adopts a critical perspective of organisational life. It emphasises the consequences of one’s words and deeds on others. Working with this paradigm one wishes to change things.
4.3 Research approaches
Two main research approaches can be adopted when conducting research: deductive and inductive approach. Deduction is the development of theory and hypotheses which are tested by using a research strategy. Deductive reasoning is done when a conclusion is logically derived from a set of premises (stellingen). The conclusion will be true when all these premises are proven to be true. There are 5 stages in an inductive research:
- Forming a hypothesis to form a theory
- Deduce testable premises
- Examine these premises and the logic of the argument that produced them, relate it to existing theories
- Testing the premises by collecting data to measure variables or concepts
- Analyze the results, If they are not consistent with the premises the theory is false and should be rejected, or modified. If the results are consistent that a new theory I s formed.
There exist four general characterizations for deduction
- Reliability. Every research should use a highly structured methodology, so that it is easy to replicate. If this is the case the research is reliable.
- Concepts need to be operationalized in such a way that enables facts to be measured.
With inductive reasoning it is not true that when a set of premises are true that a clear conclusion can be formed. This is because the conclusion is based on observations made by humans, and humans make mistakes. A conclusion is therefore never guaranteed.
A third approach, called abduction, starts with a conclusion: a surprise fact. With a set of premises one subsequently tries to prove the conclusion. An abductive approach does not move from theory to data (deduction) or from data to theory (induction), but rather moves back and forth between the two, combining deduction and induction.
Chapter 5 – How to formulate a research design
A researcher must be able to explain why he chooses a particular research design. This justification should be based upon the research questions and objectives and should also be consistent with his research philosophies.
5.2 Choosing a research design
A research design is a general plan of how one will answer research questions. It includes clear objectives derived from the research question, it displays the sources from which data will be collected and it will explain how these data will be collected. This chapter will describe the different aspects of the formulation of a research design:
- The research strategy: qualitative, quantitative or multiple methods
- The nature of the project: explanatory, descriptive or exploratory
- Methodological choice and related strategies
- Determining the time horizon of the research
- Ethical issues regarding the project
5.3 The research strategy : qualitative, quantattive or multiple methods
Quantitative research design
Often, the term ‘quantative’, is used to refer to a way to collect data or a procedure to analyse data that generates or uses numerical data. Some characteristics:
- This research method is often associated with positivism. But may also be associated with interpretivism when data is drawn from qualitative numbers.
- Quantative research is generally associated with a deductive approach, which means that the focus is on using data to test a certain theory or certain theories. However, it could be associated with an inductive approach in some cases.
- This method explores the relationships between variables after which they are measured numerically and analysed using statistical techniques.
Qualitative research design
‘Qualitative’, is a term frequently used as a synonym for a way to a data collection technique or a procedure to analyse data tat generates or uses non-numerical data. Some characteristics:
- This research method is often associated with an interpretive philosophy, because researchers need to make sense of the phenomenon being studied. Qualitative research is often referred to naturalistic research since it needs to be conducted in a natural setting, in order to gain trust, participation and access to meanings and in-depth understanding
- Qualitative research can either be started with an inductive or a deductive approach. But in practice, an abductive approach is frequently used.
- When conducting qualitative research, participants’ meanings and the relationships between them are studied using data collection techniques and analytical procedures, to develop a conceptual framework.
- It is usually associated with action research, case study research and ethnography.
Multiple methods research design
Many management and business research designs are likely to combine qualitative and quantitative elements. This is because some data derived from qualitative research may be analysed quantitavely, or may be used to inform the design of another questionnarie. Quantative and qualitative research may be seen as two ends of a continuum. Characteristics:
- Often associated with critical realism, since this philosophy advocates that while there is an objective reality to the world we live in, the way in which each of us understand and interpret it will be affected by our own social conditioning. It could also be associated with pragmatism .
- This method may use either an inductive or a deductive approach. Frequently both approaches are used.
Figure 5.2 on page 165 shows an image of the different methodolocial choices one could make:
- Mono method: choosing either a quantitative or qualitative study
- Multiple methods: choosing both qualitative and qualitative study
- Multi method: more than one data collection technique is used but this is restricted to either qualitative or quantitative design
- Mixed methods: both qualitative and quantitative design are mixed in a research design
5.4 Recognising the nature of a research design
There exist three research designs one could adopt when conducting research:
1. Exploratory study
This kind of study is a valuable way to ask open questions to discover what is going on and gain new insights about a subject of interest. Conducting exploratory research is useful when one wishes to understand something or wants to assess phenomena in a bright light. A view ways to conduct exploratory research are:
- To search literature
- To interview experts
- Conducting focus group interviews or individual interviews
2. Descriptive study
The purpose of a descriptive research is to acquire an accurate profile of happenings, people or situations. It is possible for descriptive, explanatory and exploratory studies to coexist in one research project, where they might extend one another. When conducing descriptive research one should be cautious, because descriptive study may become too descriptive and may therefore lead to worthless outcomes. This is also the reason why most descriptive studies are often combined with explanatory studies: after describing something the research will provide a valuable explanation. This is referred to as descripto-explanatory study.
3. Explanatory study
When performing this kind of study one wishes to determine causal relationships between certain variables.
5.5 Research strategies
Generally, a strategy is a plan of approach to achieve a certain goal. A research strategy could therefore be defined as the various steps a researcher has to take to answer his research question. The choice of a research question should be guided by one’s research question(s) and objective(s), the cohesiveness with which these link to the research philosophy, research approach and purpose, and to more pragmatic concerns such as the extent to existing knowledge and access to participants and other sources of data. The following strategies will be discussed in this chapter (along with the research design that is linked to them):
Quantitative research design only
Qualitative research design only
Quantative, qualitative or both
The experiment is a type of research that has been used frequently by natural scientist. The goal of an experiment is to examine the probability of a change in an independent variable causing a change in another, dependent variable (Hakim 2000). See table 5.2 on page 174 for the different variables and their meanings. Instead of research questions, an experiment uses hypothesis (predictions). There are two kinds of hypothesis in an experiment:
- Null hypothesis - which predicts that a significant difference or relationship between the variables does not exist
- The alternative hypothesis - which predicts there is a relationship or difference
When performing an experiment, the null hypothesis is tested statistically. The null hypothesis will be accepted when the probability that there is no statistical difference is greater than a prescribed value (most of the times 0.05). In this case the alternative hypothesis will be rejected.
There are various experimental designs:
- Classical experiment – a group of participating people is selected and randomly assigned to either a control or an experimental group. The experimental group will test a manipulation or intervention (storing) and in the control group no such intervention is made. Because the control group is influenced by the same external influences as the experimental group any changes to the dependent variable will have to be caused by the intervention.
- Quasi experiment – also uses an experimental and control group, but the participants will not be randomly assigned to a group. Matched pair analysis is when a participant in a control group will be compared to a participant in the experimental group based on matching factors such as gender, age, occupation etc. this is to create an even greater possibility that the intervention is the cause of change to the variable.
- Within subject design/repeated measures design – this design uses only one single group to determine change in a variable. Every participant will be subject to an intervention of the independent variable. Before the intervention, all participants will be observed, a pre-intervention, to establish a baseline (or control), after which a planned intervention of the independent variable and observation and measurement of the dependent variable will follow. This research design requires much less participants than others, but the side effects may be that the participants become tired or familiar with the experiment.
Internal validity is the extent to which the findings of the experiment can be attributed to the interventions instead of any flaws in the research design (such is the case with a laboratory experiment. External validity is a lot more difficult to establish (when conducting field-based research).
This research strategy is usually associated with the deductive research approach. It is often used for exploratory and descriptive research. Because most surveys use questionnaires it is easy for people to understand and to explain. This is the reason why this kind of research design is so popular. Besides through questionnaires, data for a survey strategy could also be collected through structured observation and structured interviews. With a survey, quantative data is collected and be analyzed quantatively using descriptive statistics. When using a sample one needs to be sure that the sample is representative to the whole population.
An archival research strategy uses administrative records and documents as the main source of data. Not only historical but also recent data documents could be collected and analysed when adopting this strategy. With use of an archival research strategy (research) questions with focus upon the past could be answered. These questions may be exploratory, descriptive or explanatory.
A case study allows one to explore a research topic or phenomenon , within its context or within real-life contexts. With a case study there is no clear boundary between that which is being studied (the phenomenon or topic) and the context within which it is being studied (the real-life ‘case’). This approach is useful when one wishes to gain a better understanding of the research and a certain phenomenon , especially when one wishes to explore existing theory. Below are some characteristics of case studies:
- It is most often used in exploratory and explanatory research.
- Case study research could combine qualitative and quantative methods such as questionnaires and interviews. The use of different data collection techniques within one study to be sure that the data are telling you what you think they tell you is called triangulation.
- It is possible to use multiple cases within one case study, this is termed literal replication. The cases will be chosen in such a way that similar results are predicted to be produced from each one. Theoretical replication is when a contextual factor is deliberately different in a certain set of cases. This approach of case study is done deductively.
- Holistic case study – when research is focussed on the organisation as a whole
- Embedded case study – when research is focussed on sub units within an organisation and the case will involve more than one unit of analysis.
This approach is used to study particular groups of people. When conducting ethnographic research one wishes to explore and analyse people in groups who share the same space (this could be the same street, work group, organisation or even society) and who interact with each other. Cunliffe distinguishes three ethnographic strategies:
- Realist Ethnography – this is an objective, factual strategy which wishes to identify ‘true’ meanings. People are being observed through facts or data about structures and processes, routines and norms, practices and customs, artefacts and symbols (Cunliffe 2010). A realist ethnographer writes in third person, which displays his role as impersonal reporter of facts.
- Impressionist/Interpretive Ethnography – this ethnographic strategy is in contrast with realist ethnography since impressionist ethnography focusses on subjectivity rather than on objectivity. Participants are treated like people rather than just subjects, this is the reason why Tedlock (2005) calls the interpretive ethnographic strategy ‘the observation of participation’. Since the Instead of one definite meaning, an interpretive ethnographic believes that it is likely that multiple meanings exist. The interpretive ethnographer writes his report in first person and uses method such as personalisation, dialogues, quotations, dramatization etc.
- Critical Ethnography – this strategy is designed to analyse and explain the impact of power, privilege and authority on the people who are subject to these influences.
This type of research strategy designed to develop answers to real organisational problems by using a participative and collaborative approach which uses various forms of knowledge. Action research will influence the participants and the organisation beyond the research project. As Greenwood and Levin said: action research is a social process in which a researcher works with members of an organisation to enhance their situation and their organisation. This type of research has 5 themes:
- Purpose – the purpose of action research is to promote organisational learning to produce practical outcomes through identifying issues, planning action, taking action and evaluating action (Coghlan and Brannick).
- Process – the process of action research starts with a particular context and with a research question, but because it moves through several stages (See figure 5.4 on page 183) the focus may change as the research develops. Each stage of the process begins with diagnosing or constructing ideas, planning, taking action and finally evaluating action. This cycle will be repeated several times.
- Participation – this component of action research is critical. For Greenwood and Levin action and participation are essential parts of an Action Research process. One of the reasons why this is the case is because the members of an organisation need to cooperate with the researcher and enable him to study their existing work. Moreover, the participants are required to participate in the form of collaboration though the cycles to allow any improvement in the organisation to occur. Without participation this type of research would not be able to work. Action Research enables bottom up culture change, because organisational members are more likely to implement change they have helped to develop. Therefore, members of an organisation become more engaged and more willing to make decision.
- Knowledge – different forms of knowledge: theoretical, propositional, lived experiences of participants and knowing-in-action knowledge. The last type of knowledge comes from practical application. All these kinds of knowledge will be incorporated into each of the stages of the Action Research process.
- Implications – One of the implications of Action Research is that participants will raise their expectations about their future treatment (since they are so involved with the organisation). Another implication is that the organisation will develop and its culture will change. Also researchers could use the results from this research and use it to develop theory to inform other contexts.
Grounded theory is a theory developed from a set of data (using an inductive approach). It was developed as a way to analyse, interpret and explain the meanings that social actors construct to make sense of their daily experiences in particular situations. There are three stages:
- Open coding – reorganising data into categories
- Axial coding – determining relationships between categories
- Selective coding – the integration of categories to produce theory
During all these stages of coding the researcher is constantly comparing each item of data with others. Constantly coding involves moving between inductive (data to theory) and deductive (theory to data) thinking: while discovering relationships between codes and interpreting them the researcher is thinking inductively (he develops his own theory from the relationships between codes). This interpretation needs to be tested by collecting data from other cases, which means that the researcher is thinking deductively because he tests his ‘theory’(interpretation) with other data. This is known as the process of abduction.
With the Grounded Theory strategy, sampling is not meant to achieve representativeness but rather to focus the research on a core theme, relationship or process. This approach is known as theoretical sampling which ends when theoretical saturation/conceptual destiny is reached. This happens when the data collection does not continue to reveal any new properties relevant to a category, where categories have become well developed and understood and relationships between categories have been verified (Straus and Corbin 2008).
The term narrative means story or a personal account which interprets an event or sequence of events (Saunders 2012). Narrative inquiry refers to a research strategy where a researcher believes that the experiences of his participants can best be accessed by collecting and analysing these as stories. Narrative inquiry preserves any chronological connection and sequence of events as told by the participant. In this way the reader may find it more easy to understand the report and the researcher is able to provide his interpretation of the events.
With Narrative Inquiry the participant is the narrator of a story about an event, work project, managing or setting up a business, or organisational change. It may be used in combination with other strategies as complementary approaches.
5.6 A Time Horizon
Cross Sectional studies
When a research is more like a ‘snapshot’ taken at a certain time then it’s called cross-sectional. These kind of studies often use the survey strategy, where an incidence of a phenomenon may be described or where may be explained how different factors in different organisation are related.
Is the research reported more like a ‘diary’ which represents the events over a specific period than it’s called longitudinal. The advantage of this time horizon is that it’s able to study change and development.
5.7 The ethics of Research
When choosing a subject for a research project one needs to consider the ethics. Some topics have more ethical difficulties than others. One needs to be sure that ethical issues will not be disadvantageous (harmful, embarrassing, painful) to participants. Moreover, the participants need to be aware of that they are subject of research.
5.8 Quality of Research
To ensure quality in any scientific research one needs to consider the ‘canons of scientific inquiry’:
- Reliability – a reliable research is reproducible, meaning that the data collection techniques and analytic procedures would produce the same findings if they were repeated by some one else or another time. In order to be reliable one has to work in a structured and methodological way.
- Construct validity – the extent to which the research measures actually measured what the researcher intended them to assess.
- Internal validity – this is the case when the research displays a causal relationship between two variables.
- External validity – Concerned with questions such as: “Are the research findings generalised?”, “Would a researcher find the same in other relevant settings or groups?”.
Alternative criteria to asses quality of research
The ‘canons of scientific inquiry’ are most suitable with quantative, positivist methods. Researchers who undertake a qualitative research may find it difficult cover the four criteria listed above, because they may not be suitable to their kind of research. As an answer to this problem, Lincoln and Guba, have formulated new names for the ‘canons of scientific inquiry’: dependability instead of reliability, credibility for internal validity and trasferability instead of external validity. Moreover, Lincoln and Guba have also developed a new set of criteria named ‘authenticity criteria’.
5.9 The role of the researcher
Full-time students usually adopt the role of an external researcher. This is someone who needs to identify an organisation within which he conducts his research. The researcher is external to the organisation therefore he has to negotiate with its members to be able to access the organisation and to collect its data.
An internal/practitioner researcher is one who works in an organisation. The advantage of this is that the researcher has easy access to the organisation. Plus, he is also likely to have knowledge of the organisation and therefore understands the complexity of what is happening in that organisation. This may also be a disadvantage since the external researcher’s assumptions and preconceptions may be different from reality.
Chapter 6 – Access to data and research Ethics
Before conducting research, a researcher need to be sure he will have access to the data he needs and he has to think carefully about the possible ethical difficulties he might face. Because business and management research will inevitably make use of human participants ethical concerns will almost always rise.
6.2 Gaining access
There are different types of access to data:
- Traditional access- this may refer to face-to-face interaction, conversations, correspondence or visiting data archives.
- Internet-mediated access – this involves the use of a computer, or computer technologies such as the Web, email and webcams to be able to gain access to questionnaires, discussions, experiments or interviews or to gather secondary data.
- Intranet-mediated access – a variant of internet-mediated access where one gains virtual access as an organisational employee or worker using its intranet.
- Hybrid access – this type of access combines traditional and internet-mediated approaches.
The levels of access may vary because the depend on the nature and depth of the access one wishes to achieve: physical, continuing and cognitive access. Physical access may be difficult because it not all organisations are prepared to engage in activities which are not necessary for them, since time and effort is required. Sometimes the gatekeeper (the person who keeps data and decides who may have access to it) does not allow people to undertake the research (because the organisation does not receive value from it or the topic is to sensitive).
Many people see access to data as a continuing process and not just one single event. One of the two reasons for this is that access may be an iterative (herhalend) and incremental (stapsgewijs oplopend) process. After gaining access to one particular set of data one might seek further to achieve other data in order to conduct another part of the research. Another reason why access is a continuing process is because those people from whom one needs to collect data may be different to those who agreed to your request for access (gatekeepers).
Physical access to data from of an organisation will be granted in a formal manner, though an organisations management. Therefore it is useful to gain trust from the organisational members. This type of access is named cognitive access.
Negotiating access is likely to be an important if one wishes to gain personal entry to an organisation and to be able to have cognitive access to allow one to connect the necessary data. Therefore it is important to consider the project’s feasibility (determine whether it is practicable to negotiate access for a research project) and sufficiency (whether one is able to gain sufficient access to fulfil the research objectives).
Issues of being an external researcher
Researchers need to negotiate access at each level: physical, continuing and cognitive. Because an external researcher lacks status in an organisation or group in which he wishes to conduct research he will face difficulties at every level of access. Therefore external researchers rely on the goodwill of the organisational members. To be able to gain goodwill from members a researcher needs to be able to communicate his competence and integrity and explain the importance of his research project clearly and precisely.
Issues of being an internal/participant researcher
Even though an internal or participant researcher is familiar with the organisation and vice versa, he is still likely to face problems of access to data. The status of an internal/participant researcher who wishes to gain cognitive access could cause suspicions. This is because other organisational members may not know what the internal/participant researcher will do with the data. Here it is also important for the researcher to be able to communicate the purpose of his research.
6.4 Strategies for gaining access
This section will discuss the different strategies one could use to obtain access to data for traditional as well as internet-mediated means. The applicability of these strategies will differ in relation to the researcher’s status as an internal or external researcher. Table 6.1 on page 217 displays a number of different strategies.
Familiarity with the group and sufficient time
Before trying to obtain physical access it is very important for a researcher to familiarise himself with the organisation or group. It may take a lot of time (weeks or even months) before physical access could be gained (if access is even granted at all). This is the reason why a researcher needs to plan sufficient time for the data access part of a research project. One also needs to consider the time needed for a participant to respond to the request of participating in a research.
Using existing contacts and/or developing new ones
When a researcher is able to use existing contacts it is easier to gain access. Because these contacts have knowledge of the researcher means that they can trust him and his intentions and are therefore likely to gain access to the data.
Providing clear account of requirements for participators
Researchers need to be aware that they provide a clear report of their requirements to allow their participants to know what will be asked of them. Without clear requirements, participants may act cautious since the required amount of time they have to put in their participation may seem to be disruptive. This is why an introductory letter to the participants will have to provide them with an outline and goal of the research and what the participants will have to do.
Overcoming organisational concerns
Organisations may be concerned about:
- The amount of time or resources involved in the request for access - the less the better
- The sensitivity about the topic - negative implications are less likely to lead to granting access, thus highlight positive approach.
- The confidentiality of the data and the anonymity of the organisation need to be ensured.
Possible benefits to organisation granting access and using suitable language
Organisations and their members may find it helpful to discuss their own situation in a non-threatening, non-judgemental environment. Therefore it may be helpful to provide a summary report of one’s findings to those who grant access. One should also be aware that the use of language is important and should depend on the nature of the people who participate.
Facilitating replies, developing access, and establishing credibility
Different contact methods could be used to write requests for access (phone, skype, fax, email), but these may not be suitable in all cases. When using an incremental strategy (from minimum requirements of participants to more requirements) one is able to obtain access to a certain level of data and a positive relationship with participants will rise. As one establishes credibility he can develop the possibility of achieving a fuller level of access.
6.5 Research ethics
Ethics refer to the standards of behaviour that guide your conduct in relation to the rights of those who become subject of your work, or are affected by it and usually accompanied by social norms(Saunders 2012). A social norm is an indicator of the type of behaviour an individual should adopt in a specific situation.
Two conflicting philosophical positions have been identified with regard to ethics:
- Deontological view – following rules to guide researcher’s conduct. When one acts outside these rules it can never be justified
- Teleological view – deciding whether an act is justified should be determined by its consequences and not by predetermined rules.
‘Codes of ethics’ were developed to overcome ethical dilemmas arising from various social norms. Codes of ethics are a list of principles which outlines the nature of ethical research and a statement of ethical standards.
Many universities have research ethics committees to ensure that research conducted by students is non-controversial and pose minimal risk to participants. Research ethics committees review all research conducted by those in the institution that involves human participation and personal data. Table 6.3 on page 231 and 232 provides the general principles developed to recognise ethical issues.
Ethical issues associated with internet-mediated research
A list of general ethical issues associated with Internet-mediated research is listed below:
- Scope for deception
- Lacking respect and causing harm
- Respecting privacy
- Nature of participation and scope to withdraw
- Informed consent
- Confidentiality of data and anonymity of participants
- Analysis of data and reporting the findings
- Management of data
- Safety of data and reporting findings
The term netiquette refers to ‘net etiquette’, or in other words the social standards which one should use online. Netiquette concerns the use of emails and messaging since they may be poorly worded and may seem unfriendly or unclear to the receiver.
6.6 Ethical issues during the specific stages of the research process
Figure 6.1 on page 236 sums up the different ethical issues that could rise at specific stages of the research process. Most ethical issues can be predetermined and dealt with during the design stage of a research process. One should be sure that the intended research is in line with the ethical principle of not causing any harm to participants. When seeking access a researcher should not put any pressure on the members of an organisation, since this might be unpleasant to them.
When approval for access is given, one should ensure the participants that personal data will be handled anonymously and confidentially. There is of continuum for the range of approval that could be given.
- The continuum starts with complete lack of consent – this is the case when participants may fear deception from the researchers part
- Through inferred consent - the participant makes an agreement which states that he has control over the way the data is analysed, used, stored and reported
- And ends with informed consent - participants are fully informed and may ask questions whenever they want
Moreover, the participants need to be fully aware of the information that is asked from them. A researcher needs to inform them of this formally with the use of a participant information sheet. It has to include the requirements and implications of participating, the nature of research, how the data will be analysed, reported and stored and who to contact when any concerns rise. A more detailed written agreement could be established as a consent form, which both parties should sign. Consent forms help clarifying boundaries of consent.
Ethical issues during data collection
Once participants or organisations have approved to take part in a research they maintain certain rights. These include that they may withdraw from the research at any point they want and that the researcher should keep the aims of his research project that he agreed, if this is not the case he is deceiving the participants. Moreover, a researcher should not ask of participants to participate in something that will cause them any arm or intrude their privacy.
Another ethical principle during the data collection stage of the research is that the researcher should be objective and should ensure that he collects data accurately and fully. The importance of this also relates to the validity and reliability of a research.
When using observational techniques one should avoid attempting to observe behaviour related to personal life (phone calls and etc.), because this will cause the relationship between researcher and participant to break down. Moreover, the issue of reactivity may raise here as well. Reactivity refers to the reaction of the participants to the researcher and their research instruments. The solution of this problem may be to undertake a covert study so that the participant are not aware of the fact that they are being observed. But the problem of reactivity may diminish when participants are becoming more and more accustomed to the researcher’s presence. This adaptation is referred to as habituation. After observation has taken place a researcher has to inform the participants about what has occurred and for what reason. This is called debriefing.
Ethical issues during the analysis and reporting stages
A researcher needs to maintain his objectivity in order to not misrepresent the data he has collected. He should not collect that identify participants where it is not needed to do so. Confidentiality and anonymity play an important role in this stage of the research as well.
6.8 Principles of data protection and data management
In order to manage data ethically and lawfully one should consider the principles of data protection and data management. Personal data are data that relate to a living individual which allow that person to be identified, maybe in combination with other sorts of information known to the controller of the data. Anyone who controls this type of data is subject to provisions of data protection legislation of the country one lives in. The list below provides a summary of a couple of principles a researcher should consider. Personal data should be:
- Processed fairly and lawfully
- Adequate and relevant
- Obtained for specified explicit and lawful purposes
- Accurate and up-to-date where necessary
- Kept no longer than necessary
- Kept securely
- Processed in accordance with the rights for data subject by the Act
- Not transferred to a country outside the European Economic Area
A further category of personal data, known as sensitive personal data, covers information about a participant’s racial or ethnic origin, political opinions, religious beliefs, physical/mental health, sexual life or any proceedings or sentence related to an alleged offence.
Chapter 7 – Sample selection
When conducting scientific research one should always consider the use of sampling. If it’s possible to obtain and analyse data from every possible case or group member it’s termed a ‘census’. However, this may not always be the case because one might face some restrictions: time, money and access. This is the reason why sampling is often used.
A sample should always represent the full set of cases in a way that is meaningful and which we can justify (Becker 1998). The full set of cases from which the population is taken is called the population. The population does not necessarily signify people, it could also point to Chinese restaurants or electric cars in a specific region for example. There are a number of reasons why sampling is a better option than a census:
There is no budget to survey the entire population
There is no time to survey the entire population
It is not practicable to survey the entire population
Researchers such as Barnett argue that using a sample leads to higher overall accuracy than a census. This is because the researcher focusses on a smaller number of cases to collect data from and therefore has more time to design and pilot the data collection methods. Moreover, data collected from fewer cases means that the information is more detailed.
There two types of sampling techniques (see figure 7.2 on page 261):
- Probability/ representative sampling – the chance of each case to be selected from the population is already known and is usually equal for all cases. This is when you want to prove something statistically.
- Non-probability sampling – the chance of each case to be selected from the population is not known and it’s impossible to make statistical interferences about the characteristics of the population.
7.2 Probability sampling
Often, probability sampling is associated with survey strategies when one needs to make interferences from a sample about a population to answer the research question and to meet objectives. Henry (1990) advises against probability sampling for researches that use less than 50 cases. Because this amount may not be representative of the entire population. The process of probability sampling passes four stages:
- Identify a suitable sampling frame based on research questions and objectives
- Decide an appropriate sample size
- Choose the most suitable sample techniques and select the sample
- Check if the sample represents the total population
The ‘sampling frame’ for a probability sample is a complete list of all the cases in the population from which the sample is drawn. It is not possible to select a probability sample without a sampling frame. When no suitable list exists and you still want to use a probability sampling technique, you will have to compose your own sampling frame and ensure that it’s valid.
The way in which a researcher defines his sampling frame also raises implications to the extent to which he can generalise form his sample. If a sampling frame is a list of all customers of an organisation one can only generalise to that population. Thus, you should not generalise beyond your sampling frame. This is a mistake many researchers make; they don’t place clear limits on the generalizability of their findings.
Suitable sample size
The larger a sample’s size the lower the likely error in generalising to the population . The choice of sample size is governed by
- The confidence one has in the data (whether you are certain that the sample is representative of the entire population)
- The margin of error one tolerates (the accuracy for estimates made from the sample)
- The type of analyses one is going to undertake
- The size of the total population
In order to ensure that faked results cannot be present, the analysed data must be normally distributed. The larger the absolute size of a certain sample, the closer its distribution will be to the normal distribution. This relationship is known as the ‘central limit theorem’ and it also occurs if the population from which the sample is drawn isn’t normally distributed. It is proven that any sample size larger than 30 will usually result in a sampling distribution for the mean that is very close to a normal distribution. This is the reason why Stutely (2003) advises to use sample with a minimum of 30 cases.
The process of making conclusions about a population on the basis of data describing the sample is called ‘statistical interference’. The ‘law of large number’ holds that large samples are more likely to represent the population from which they are drawn than smaller samples. Moreover, their means are also more likely to be equal to mean of the population.
It is essential for a probability sample to be representative of the population. A perfect representative sample is one that represents the population from which it is taken exactly. There are four levels of response to questionnaires and structured interviews:
- Complete refusal – no questions were answered
- Break- off – less than 50% of the questions were answered
- Partial response – 50% to 80% of the questions are answered
- Complete response – all the questions were answered
Reasons why people don’t respond may be because they refuse to participate to the research, they are ineligible (they don’t fit the requirements) or they may be unreachable. A research report should always include the response rate of the research. This could be calculated by the following formula:
total response rate= total number of responsestotal number in sample-ineligible
A more common way of calculating the response rate excludes ineligible respondents who were unreachable. This is the active response rate
active response rate= total number of responsestotal number in sample-(ineligible+unreachable)
It is important to estimate the likely response rate and increase the sample size accordingly to ensure that you will be able to undertake the analysis at the level of detail required. Once the estimated response rate and the minimum sample size are determined one could calculate the actual sample size with the following formula:
na = actual sample size required
n = minimum sample size
re% = estimated response rate expressed as percentage
One way to estimate the response rate is to analyse the response rates achieved for similar surveys that have already been undertaken and subsequently base the response rates on these.
There are five main techniques for selecting a probability sample:
- Simple random
- Systematic random
- Stratified random
- Multi stage
See figure 7.3 for a guideline for selecting the appropriate probability sampling technique.
Simple random sampling
This is done by selecting the sample at random from the sampling frame using a computer or random number tables. You do this by numbering each of the cases with a unique number (Starting with 0) and select cases using random numbers until your actual sample size is reached. This is done without replacement so that no number could be selected twice. This type of sampling is best used when one has an accurate and easily accessible sampling frame that lists the entire population. The sample that is eventually selected can be said to be representative of the entire population, because the numbers were chosen without bias.
Systematic random sampling
This involves the researcher selecting the sample at regular intervals from the sampling frame. This is done by numbering each of the cases in a sampling frame (starting with 0), then selecting the first case using a random number, calculate the sampling fraction and finally select subsequent cases systematically by using the sampling fraction to determine the frequency of selection. The sampling fraction is the proportion of the entire population that one needs to select and could be calculated using the following formula:
sampling fraction= actual sample sizetotal population
When the sampling fraction is ¼ one needs to select every fourth case from the sampling frame. Using this technique one needs to be sure that the list does not contain periodic patterns since this may disturb the results.
Stratified random sampling
This is a modification of random sampling in which you split the population into two or more relevant and significant strata (lagen) based on one or more attributes. In other words, the sampling frame is divided into various subsets after which a random number is drawn from each of the strata. By dividing the population into a series of relevant strata the sample is more representative because one can ensure that each of the strata are represented proportionally within the sample.
To do this a researcher chooses the stratification variable(s) and divides the sampling frame into the discrete strata. Then he numbers each of the cases within each strata with a unique number (starting with 0), after which he selects the sample using either simple random or systematic random sampling.
This technique is similar to stratified random sampling as it is required to divide the population into discrete groups prior to sampling. The groups are called ‘clusters’ and can be based on any naturally occurring grouping (for example by manufacturing firm or geographical area). Instead of the individual cases within the population, with this technique the sampling frame consists of the list of clusters. The technique has three stages:
- Choose the cluster grouping for the sampling frame
- Number each of the clusters with a unique number (beginning with 0)
- Select the sample of clusters using some form of random sampling
This technique leads to a sample that represents the whole population less accurately than stratified random sampling.
This is a development of cluster sampling. Just like cluster sampling, multi-stage sampling can be used for any discrete group, including those that are not geographically based. With this technique one modifies a cluster sample by adding at least one or more stage of sampling that also involves some form of random sampling. The four phases of multi-stage sampling are depicted in figure 7.4 on page 279. Because this technique relies on various different sampling frames, one needs to ensure that they are all suitable and available.
7.3 Non-Probability Sampling
Non-probability sampling provides alternative techniques for selecting samples. There are no rules for deciding the sample size, but it is important to choose a size that represents the population adequately.
Quota sampling is a non-random approach for sampling and is often used for structured interviews. This is a type of stratifies sample in which the selection of cases within strata is totally non-random. To select a quota sample one needs to divide the population into groups, calculate the quota of each group (based on relevant data), give each interviewer an ‘assignment’ (this states the number of cases in each quota from which to collect data) and combine the collected data from the interviews to provide the sample.
Purposive sampling/judgemental sampling
This type of sampling involves the researcher to use judgment to select the cases that are most suitable for answering the research questions and to meet the objectives. There are a number of strategies which one can adopt to use purposive sampling:
- Extreme case/ deviant sampling – focus on unusual/special cases that will provide answers to research questions and enable the researcher to meet objectives
- Heterogeneous/maximum variation sampling – uses judgement of the researcher to choose the participants with sufficiently diverse characteristics to generate the maximum variation possible in the collected data.
- Homogeneous sampling – focuses on one specific subgroup in which all the members are very similar (age,occupation)
- Critical base sampling – selects critical cases because they are either important or can make a dramatic point.
- Typical case sampling – these enable the researcher to generate an illustration of what is typical to those who will read the research report and are unfamiliar with the topic.
- Theoretical sampling – sample selection is dictated by the needs of the theory being developed. Thus the sampling occurs during the research as more participants are needed.
Snowball sampling is a type of sampling where participants volunteered to participate in the research, instead of being chosen by the researcher. Self-selection sampling on the other hand occurs when the researcher asks the participants to volunteer in the research.
This is a type of sampling where sample cases are selected without an obvious relation to the research questions.
Chapter 8 – Secondary data usage
8.1 Secondary data
Secondary data are data that have already been collected for some other purpose. These data include published summaries as well as raw data and therefore include both quantitative and qualitative data. Secondary data is used to provide additional or different knowledge, interpretations or conclusions.
8.2 Types of secondary data
Secondary data that may be analysed further is called raw data, this is data which hasn’t been processed. Compiled data, that have received some form of selection or summarising, may also be analysed further. The three main subgroups of secondary data are (see figure 8.1 on page 307):
- Documentary secondary data – these are data often used in research projects that also collect primary data. Documentary secondary data include texts such as books, journals, magazine articles, newspapers notices, correspondence, minutes of meetings, reports to shareholders, transcripts of speeches and conversations, diaries, administrative and public records and web page texts.
- Survey-based secondary data – these are data collected for some other goal using a survey strategy, usually questionnaires. They are available through compiled data tables or as a downloadable matrix of raw data. These data can be collected by censuses (participation is obligatory), continuous/regular surveys (censuses repeated over time) or ad hoc surveys (specific in their subject matter).
- Multiple sources – these data can be compiled from documentary or survey secondary data. Different data sets have been combined to form a new data set prior to your accessing the data. Multiple-source data can be compiled by extracting and combining particular variables from a couple of surveys to provide ’longitudinal data’. Alternatively, data can be compiled for the same population over a period of time using a series of snapshots to form ‘cohort studies’.
8.3 Finding secondary data
To find relevant secondary data you need to pass two interlinked stages:
- Establish the likely availability of secondary data (for example with the use of tertiary data)
- Locating the data you require (for example via online databases or journals)
8.4 Good and bad sides of secondary data
Collecting secondary data saves much of a researcher’s time and money. Because the data is already collected the researcher has much time left to interpret and analyse the data. Moreover, the fact that the data has already been collected makes it much less obtrusive to use, this is a benefit which is likely to be visible in sensitive situations.
Another advantage is that secondary data may provide a good comparative and contextual background for a researcher’s findings. Reanalysing these data can result in unforeseen discoveries.
Secondary data has been collected for a specific purpose different from your research questions and objectives. The consequence of this is that the data might not be suitable for your research. Furthermore, it may be difficult or expensive to gain access to the data.
8.5 Evaluating sources
When using secondary data one needs to be sure that the data enables one to answer the research questions and to meet objectives, the benefits of the data are greater than their costs and that it is possible to gain access to the data.
To test the suitability of any data set is helpful test the measurement validity. Make sure that the data you use provide you with the information you need to answer the research questions and objectives. To ensure that the data collected is reliable and valid it is useful to have a clear explanation of the techniques that were used to collect the data.
Another way to test the suitability of secondary data is to determine to what extent the data covers the population about which you need data. This is to be sure that unwanted data can be excluded and to ensure that sufficient data remain for analyses to be undertaken once those unwanted data have been excluded.
When looking for data one needs to be sure that measurement bias is not present. Measurement bias may occur for two reasons:
- Deliberate or intentional distortion of data – occurs when data are purposefully recorded inaccurately.
- Changes in the way the data are collected
Chapter 9 – Collection of primary data by observation
9.1 Why observations matters
When one’s research questions and objectives concern the things that people do, it is obvious to discover this by watch them do it. Observation involves the systematic observation, recording, description, analysis and interpretation of people’s behaviour. This chapter examines two kinds of observation:
- Participant observation – qualitative, derived from social anthropology
- Structured observation – quantitative, concerned with the frequency of actions
Other research approaches than observation usually focus on two distinct kinds of people that participate to the research: respondents (those who simply complete questionnaires) and participants ( those who take part in most types of qualitative research). These labels do not work with observation, since the researcher is actually participating in the environment of the people he observes. Thus, in observational research those who are being observed are called ‘informants’.
9.2 Participant observation
When one conducts research through participant observation one attempts to participate fully in the lives and daily activities of informants and thus becomes a member of their organisation, group or community. This makes it possible for the researcher to share their experience not only by merely observing what is happening but also by feeling it.
One of the purposes of participant observation is to discover the delicate nuances of meaning behind the respondent’s comment which are not likely to be found using questionnaires for example. By immersing himself in the research setting the researcher is able to achieve this goal.
Participant observation finds some of its roots in symbolic interactionism where an individual derives a sense of identity from interaction and communication with others and through this process constantly adjusts his or her understanding of reality. Participant observation enables the researcher to get to the bottom of the processes by which individuals constantly construct and reconstruct their identities.
Types of participant observation
There exist four types of participant observation:
- Complete participant – researcher becomes a member of the group in which he is performing researcher. Doing so he doesn’t reveal his true purpose to the members (this may raise questions of ethics).
- Complete observer – researcher does not reveal the purpose of the activity to the members of the group, but he does not take part in the activities of the group.
- Observer-as-participant – researcher only observes (does not participate in) the activities of the group, although his purpose is known to those whom he is studying.
- Participant-as-observer – researchers takes part in the activities of the group and also reveals his purpose as a researcher.
Data analysis and collection
It is important to make notes when conducting observation-based research. Four types can be distinguished:
- Primary observations: data that explain what happened or what was said at the time (like in a diary).
- Secondary data: statements by observers of what happened or what was said; this involves interpretations.
- Experiential data: data on perceptions and feelings as the researcher experiences the process he is researching.
- Contextual data: data related to the research setting.
Robson (2011) argues that the process of data collection always starts with ‘descriptive observation’. This is when the researcher concentrates on observing and describing the physical setting, key informants and their activities, particular events and their sequence and finally the attendant processes and emotions involved.
After stage the researcher is able to write a ‘narrative account’ in which he develops a framework of theory that will help him to understand and explain to other what is going on in the research setting. In order to do this it may be helpful to focus on particular events through ‘focused observation’.
Just like other qualitative data, those collected form participant observation will start to be analysed from the minute one collects them. In other words, data collection and analysis occur simultaneously.
Reliability and validity
Participant observation has high ‘ecological validity’ since it is concerned with studying informants and their activities in their natural settings. Still, this research method may lead to a number of issues related to reliability and validity: observer error, observer bias, observer effect.
Observer error occurs when a researcher’s lack of understand of, or overfamiliarity with, the setting in which he is trying to operate as an observer may lead to unintentional misinterpretations of what is happening. Observer error does not occur intentionally or deliberately.
Observer bias occurs when the observer uses his or her own subjective view or disposition to interpret happenings in the setting he or she observes. This may the case when the observer is already a member of the group he wishes to observe ( an employee in an organisation) and therefore fails to interpret the setting objectively. When using covert observation (when the purpose of research is not known to the informants) it is difficult to check with the informants whether your interpretations were valid. Using overt (informants are aware of the purpose of the research) observation enables the researcher to ask the informants to read some of the secondary observation that relate to them. This is referred to as ‘informant verification’.
The observer effect happens when the presence of the observer affects the behaviour of those being observed, which could lead to unreliable and invalid data. A solution to this is to act covertly or to achieve minimal interaction (when the researcher melts in the background) . But informants may become familiar with the researchers presence, minimizing the observer effect; this is termed ‘habituation’.
9.3 Structured observation
Unlike with participant observation, structured observation has a high level of predetermined structure. The researcher is concerned with quantifying behaviour. Therefore it is usually used as only a part of data collection because its purpose is to tell the researcher how often things happen rather than providing a reason for why things happen.
When using this type of research method the researcher needs to determine whether he will use an ‘off-the-shelf ‘coding schedule (with this method you predetermined recording sheet) or designs his own.
Validity and reliability
There are a number of issues when using structured observation: observer error, informant error, time error and observer effects.
Chapter 10 – Collecting primary data using interviews
10.1 Research interviews
A research interview is a conversation between two or more people, requiring the interviewer to make a rapport, to ask concise (beknopte) and unambiguous questions, to which the interviewee is willing to respond, and to listen attentively. This chapter provides an overview of a few types on interview: semi-structured, in-depth and group interviews.
10.2 Interviews and their function in research and research strategy
Types of interview
When categorising types of interviews one could distinguish between interviews that are highly formalised and structured, those who lack such a predetermined structure and a types that fit in between:
- Structured interviews use questionnaires based on predetermined and standardised sets of questions and are referred to as interviewer-administered questionnaires. They are used for the collection of quantifiable data. Often referred to as quantitative research interviews.
- Semi structured interviews are non-standardised and are often referred to as qualitative research interviews. The researcher has a list of themes and sometimes some key questions, but he does not use them in a structured way. The researcher may use some questions in some interviews and other questions in others.
- Unstructured interviews are informal. This is used to explore in depth a particular area in which the researcher is interested. Therefore these interviews are termed in-depth interviews. An informant interview is one in which the interviewee is free to talk about events, behaviour and beliefs regarding the subject area and it is the interviewee who guides the conduct of the interview. A focused interview is one in which the interviewer has greater control over the direction of the interview while in the meantime allowing the interviewee’s opinions to emerge.
Interviews may be conducted on a one-to-one basis, between a researcher and a participant, but they may also be conducted on a one-to-may basis, between a researcher and a group of people (see figure 10.1 on page 375).
10.3 Semi-structured and in-depth interviews
Below is a list of the reasons why the use of semi-structured and in-depth interviews may be advantageous:
The purpose of the research – when undertaking an exploratory or explanatory research it is likely that one includes in-depth or semi-structured research interviews in his design. Interpretivists are also likely to use these types of interviews since it allows them to let the interviewees explain their responses.
The significance of establishing personal contact – sometimes participants wish to reflect on events without needing to write anything down (as with filling out a questionnaire).
The nature of questions – the use of semi-structured and in depth interviews is most advantageous when there are a large number of questions to be answered, the questions are either complex or open-ended and where the order and logic of questioning may need be varied.
Length of time required and the process’ completeness – some negotiation is always possible and the interview can take place at a time that pleases the interviewee.
10.4 Data quality issues
Data qualitative issues with semi-structured and in-depth interviews could be related to reliability, forms of bias, generalizability and validity. The lack of standardisation in these kinds of interview may lead to reliability, because other researchers may not reveal similar information. It is not feasible to try to overcome this issue, since it will always be present. In-depth and semi-structured interviews will always be flexible and because different questions were used with different participants it would be difficult to repeat this.
The next data quality issue is bias. There are three types of bias to consider:
- Interviewer bias – this is when comments, tone or non-verbal behaviour of the interviewer may lead to bias in the way interviewees respond to the questions
- Interviewee/response bias – this may be caused by perceptions about the interviewer.
- Participation bias – this is bias that results from the nature of the individuals or organisational participants who agreed to be interviewed. Because the interview may be time-consuming, participants might become tired and less willing to talk.
The issue of generalizability refers to the extent to which the results of a research are applicable to other settings. This is an issue with semi-structured and in-depth interview based researches because they are often based on sample sized samples.
Validity may also be an issue because it refers to the extent to which the researcher succeeded in gaining access to a participant’s knowledge and experience, and has been able to infer meanings that the participant intended.
10.5 Preparing for semi-structured and in-depth interviews
When preparing for in-depth or semi-structured interviews it is useful to remember the ‘five P’s ‘: prior planning prevents poor performance. To be able to avoid data quality issues when conducting in-depth or semi-structured interviews one should consider some key measures to prepare for the interviews:
- Level of knowledge: the researcher should always be familiar with the research topic and the organisational or situational context in which the interview will take place. Interviewing participants from different cultures requires the interviewer to gain some knowledge about those cultures in order to successfully undertake the interviews and to avoid misinterpretations.
- Developing interview themes and providing the interviewee with information before the interview: this provides the interviewees the opportunity to prepare themselves for the interview.
- The appropriateness of the intended interview location: it is possible that the location where one conducts interviews will influence the data he collects. The location should be convenient for the participants, it should be a place in which they feel comfortable and where the interview is unlikely to be disturbed.
10.6 Conducting in-depth or semi-structured interviews
This section will discuss the different aspects of conducting in-depth or semi-structured interviews and are intended to avoid forms of bias that could affect the reliability and validity of the data produced.
- Appropriateness of the researcher’s appearance: an interviewer’s appearance may affect the perception of the interviewee, it may have an effect on his/her credibility or result in a failure to gain their confidence.
- Nature of the first comments when the interview starts: especially when the interviewee has never met the interviewer, the first few minutes of conversation will have a large impact on the outcome of the interview – this is also related to the issues of credibility and the interviewee’s confidence.
- Approach to questioning: the questions need to be phrased clearly so that the interviewee understands them and the interviewer should have a neutral tone of voice. The use of open questions should help to avoid bias. Questions that are seeking to lead the interviewee should be avoided. One approach to questioning is the ‘critical incident technique’ in which participants are asked to give a detailed description of a critical incident relevant to the research question.
- Appropriate use of different types of questions: when conducting in-depth or semi-structured interviews one should consider formulating appropriate. There are various types of questions:
- Open questions – these allow interviewee’s to define and describe a situation or event.
- Specific/Closed questions – these may be used as opening questions when one commences questioning about a particular interview theme.
- Probing questions – these types of questions can be used to provide for further exploration of an interviewee’s response. They are used to dive deeper into the research topic.
- Nature and impact of the interviewer’s behaviour during the interview: this also relates to the issues of credibility and the interviewee’s confidence.
- Demonstrating attentive listening skills: when conducting in-depth or semi-structured interviews it is important for the interviewer to listen attentively to the interviewee’s answers in order to build understanding.
- Scope to summarise and testing understanding: the interviewee may test his understanding by summarising an explanation given by the interviewee.
- Dealing with difficult participants: always remain polite with difficult participants and do not show irritation.
- Approach to recording data: it is beneficial to audio-record an interview and make notes as it progresses. In addition to notes an interviewer should always note down the following contextual data: location of the interview, date and time, setting (noisy, interrupted), background information about participant and the interviewers immediate impression of how well the interview went.
10.7 Managing logistical and resource issues
When conducting interviews one should always consider issues regarding time spent on the interview, cost and resources and logistics of scheduling interviews. Therefore one should always manage the time, schedule the interview appropriately and to manage the interview well.
10.8 Group interviews and focus groups
Group interviews are in-depth or semi-structured interviews conducted with two or more people. When conducting this type interview be aware of the threat of the group effect, where a few interviewee’s may dominate the discussion while others may feel inhibited. Focus groups are group interviews where the topic is defined clearly and precisely. The purpose running the focus group is named the ‘moderator’ or ‘facilitator’, because he needs to keep the group within the boundaries of the topic being discussed and generate interest in the topic and encourage discussion.
When using group interviews one should consider the fact that once a sample has been selected, participants should be grouped so as not to inhibit people’s contributions.
It is helpful to make groups in which the participants have a similar status and similar work experiences. Also, a researcher should seek to reduce the contributions of people who tend to dominate the discussion. When conducting many group interviews and the fourth or fifth group interview is no longer providing the researcher with new information, he has reached saturation.
10.9 Telephone, Internet-mediated interviews and intranet-mediated interviews
Interviews conducted through the telephone may have the advantage of lower cost and easier access as well as speed. Sometimes telephone interviews may be the only possible option to interview an interviewee because of ‘long-distance’.
However, conducting interviews by telephone may lead to issues of reduced reliability, where participants are less willing to engage themselves in exploratory discussion. Also the interviewer is not allowed to witness the non-verbal behaviour of his interviewee, which may affect the way he interprets the interviewee’s answers.
Internet-and intranet-mediated interviews
Electronic interviews are those held both in real time(synchronous) as well as those that are offline (asynchronous). Examples of synchronous interviews may be chat room interviews or Voice-over Internet Protocol (such as Skype) interviews. Asynchronous interviews could be held with the use of emails or internet forums.
Chapter 11 – Collecting primary data with questionnaires
Questionnaires refer to all methods of data collection in which each person is asked to respond to the same set of questions in a predetermined order. The design of a questionnaire will affect the response rate and reliability and validity of the collected data. These can be maximised by:
- Carefully designing individual questions
- Design a clear and pleasing layout of the questionnaire
- Clear explanation of the goal of the questionnaire
- Pilot testing
- Carefully planned/executed delivery and return of completed questionnaires
11.2 Overview of questionnaires
Questionnaires tend to be used for descriptive or explanatory research since they work best with standardised questions which all respondents interpret the same way. There are different kinds of questionnaires:
- Self-completed questionnaires are completed by the respondents and are sent
- electronically via the internet: Internet-mediated/web-based questionnaires
- via intranet: intranet-mediated questionnaires
- posted via mail to respondents who post them back: postal/mail questionnaires
- or delivered by hand to each respondent after which they are collected later: delivery and collection questionnaires
- Interviewer-completed questionnaires are recorded by the interviewer himself on the basis of each respondent’s answers.
- Telephone questionnaires
- Structured interviews: those questionnaires where the interviewer physically meets respondents and asks the questions face-to-face.
Sometimes it might occur that the respondent’s answers reduce your data’s reliability simply because they have insufficient knowledge or experience and may purposefully guess at the answer. This is known as ‘uninformed response’. Respondents to self-completed questionnaires are sometimes likely to discuss their answers with others, and thereby contaminating their response.
11.3 Collecting the data
Questionnaires offer you only one chance to return collect data, because it is often difficult to identify respondents or to return to them to collect additional information. This is the reason why one needs spend time planning precisely what data one needs to collect.
Types of variables
There are three types of data variables that can be collected with a questionnaire:
- Opinion variables: these variables record how respondents feel about something or what they believe is true or false.
- Behaviour variables: these data include what people did in the past, do now or will do in the future.
- Attributes variables: these contain data about the respondent’s characteristics. Attributes are thing respondents possess.
To be sure that the data collected will lead to answers on to the research questions and achievement of the objectives it could be helpful to create a data requirements table (see table 11.2 on page 425). Investigative questions are those that need to be answered in order to successfully address to each research question and to meet each objective.
11.4 Designing questionnaires
Internal validity relating questionnaires refers to the ability of the questionnaire to measure what the researchers intended to measure. Another type of validity that can be assessed is content validity. This refers to the extent to which the measurement device (measurement questions) provides adequate coverage of the investigative questions. Criterion validity or predictive validity concerns with the ability of the measures, the questions in this case, to make accurate predictions. Finally, construct validity is the extent to which measurement questions actually measure the presence of those constructs the researcher intended to measure.
Mitchell (1996) outlined three approaches to asses reliability:
- Test re-test – this is done by comparing data collected with those from the same questionnaire under as near equivalent conditions possible. In other words, conducting the same questionnaire interview twice and test if the results are similar.
- Internal consistency – this involves correlating the responses to question in the questionnaire with each other. Thus measuring whether the responses are consistent across a subgroup or all of the questions of a questionnaire. ‘Cronbach’s alfa’ is often used a statistic to measure the responses using a particular scale
- Alternative form – this is comparing the responses to alternative forms of the same questions or groups of questions.
To design individual questions for a questionnaire, researcher could adopt or adapt questions used in other questionnaires or they could develop their own questions. The first two options are helpful when the researcher wishes to compare his findings with another study. This method ensures the reliability of the questions since they have already been tested.
Just like in other types of surveys, questionnaires may use open, closed or forced-choice questions (when the respondent has to choose from a given set of answers). Other kinds of questions are
- List question - where the respondent is given a list of responses, any of which he could choose (by ticking the box for example)
- Category questions - these only allow the respondent to choose one category only
- Ranking questions - ask the respondent to place things in rank order (for example when a respondent has to rank certain issues on degree of importance to him).
- Rating questions – these are used to collect opinion data. They use the Likert-style rating in which a respondent has to express how strongly he or she agrees or disagrees with a statement. A variation of this is the semantic differential rating question, where the respondent has to rate something on a series of bipolar scales (fast – slow; good-bad etc.).
- Quantity/self-coded questions– these are questions of which the answers are numbers which demonstrate the amount of a characteristic.
- Matrix questions – questions that enable one to record the responses to two or more similar questions at the same time.
Translating questions into other languages
When translating questions of a questionnaire into another language one should pay attention to the lexical meaning (precise meaning of a word), idiomatic meaning (meaning of a group of words), experiential meaning (meanings of words and sentences for people in their daily experiences) and grammar/syntax (the correct use of language).
Explaining the goal of the questionnaire
In a covering letter or email the researcher can explain the purpose of the survey to the respondent. This is important, because the messages contained particularly in self-completed questionnaire’s covering letter will affect the response rate.
At the beginning of the questionnaire the researcher needs to explain clearly and concisely why he wants the respondent to complete the questionnaire. This should be done on the first page of the questionnaire as well as in the covering letter. According to Dillman (2009) comments should include a clear unbiased banner or title, a subtitle and a neutral graphic illustration or logo to add interest.
Pilot testing and validity assessing
Prior to using the questionnaire for data collection it should be pilot tested. The goal of a pilot test is to refine the questionnaire to avoid problems for the respondents in answering the questions and to ensure that there will be no problems in recording the data.
Sometimes a researcher does not have much time and finds that it is better to pilot test the questionnaire using friend or family than not at all. This will at least provide his research with face validity: whether the questionnaire appears to make sense (because even friends and family are able to determine this).
Chapter 12 – Analysis of quantitative data
12.1 Quantitative data
Raw quantitative data, that haven’t been processed or analysed, convey very little meaning to most people. In order for these data into useful information they need to be processed. Quantitative data refer to all numerical primary and secondary data and can help the researcher to answer research questions and meet objectives.
12.2 Preparing, checking and inputting data
Types of data
Quantitative data can be divided in two groups: categorical data and numerical data. Categorical data are those whose values cannot be measured numerically but can be classified into sets/categories according to the characteristics that describe or identify the variable or they could be placed in rank order. There are two types of data:
- Descriptive/nominal data – these data can simply count the number of occurrences in each category of a variable. When a variable is divided into two categories (female/male for example) than the data are known as dichotomous data.
- Ranked/ordinal data – these are data that are a more precise form than categorical data. An example of ranked data may be answers to rating or scale questions.
Alternatively, numerical data are those whose values are numerically measured or counted as quantities (Berman 2008). Numerical data are therefore more precise than categorical ones because one can assign each data value a position on a numerical scale. Numerical data can be subdivided in two ways: based on interval and ratio data: or based on continuous or discrete data. Interval data can state the difference (interval) between any two data values of a certain variable, whereas ratio data can calculate the relative difference (ratio) between any two data values of a certain variable. Continuous data are those whose values can take any value (given that you measure them accurately) while discrete data can be measured precisely (often whole numbers/integers).
After determining the types of data that are to be collected the researcher can start to enter the data into data computer data processing software (RSS/EXCELL). To do this the data need to be coded using numerical codes. This enables the researcher to enter the data quickly with fewer errors. When this is done the data should be checked for errors.
12.3 Exploring and presenting the data
Turkey’s (1977) exploratory data analysis (EDA) is a useful approach to start the analysis of quantitative data. This approach focuses on the use of diagrams to explore and understand the data. Sometimes it might be possible that this approach enables you to look at other relationships in data, which your research was not designed to test.
When looking at the collected data it is best to explore specific values, highest and lowest values, trends over time, proportions and distributions. Once these have been explored one can start to compare them and look for (causal) relationships between variables)
The easiest way of summarising the data is by using tables. However, tables do not demonstrate visual significance to highest or lowest values so it may be that diagrams are a better option for summarising the data. Another way to present data is by using a bar chart, where the height or length of each bar represents the frequency of occurrence. Bar charts are similar to histograms, another type of data presenting, where the area of each bar represents the frequency of occurrence and where the continuous nature of the data is emphasised by the absence of gaps between bars. Finally, a pictogram, also similar to a bar chart, shows a series of pictures chosen to represent the data. Other kind of data presentation are:
- Line graph – this is a suitable approach when trying to explore a trend.
- Pie chart – this is a diagram that is divided into proportional segments according to the share each has of the total value.
Shapes of diagrams
If a diagram shows a bunching to the left and a long tail to the right (figure 12.3 on page 291) then the data are ‘positively skewed’. If this is the other way around then the data are ‘negatively skewed’. When the data are equally distributed on each side of the highest frequency they are ‘symmetrically skewed’.
A bell-shaped curve is called a normal distribution. With the indicator ‘kurtosis’ one can compare a diagrams pointedness or flatness with that of the normal distribution. When a distribution is flatter then it is called platykurtic and the kurtosis value is negative . When the distribution is more peaked, than it is leptokurtic and the kurtosis value is positive.
Contingency tables or cross tabulation are approaches one could use examine the interdependence between variables. Other approaches are:
- Multiple bar charts - to explore highest and lowest values.
- Percentage component bar chart – this is used to compare proportions between variables.
- Multiple line graph – this Is used to compare trends and conjunctions.
- Stacked bar chart – used to compare totals between variables.
- Comparative proportional pie chart – this is used to compare proportions of each category or value as well as the totals between variables.
- Scatter graphs or scatter plots – this diagram is often used to explore the possible relationships between ranked and numerical data variables by plotting one variable against another
12.4 Describing data with use of statistics
Turkey’s exploratory data analysis approach is a good approach to understand the data using diagrams. Descriptive statistics, on the other hand, enable one to describe the variables numerically. They describe a variable focus on the central tendency and the dispersion. Central tendency is measured by general impressions of values that could be seen as common, middling or average. These measures are determined by:
- The mode – the value that is visible most often
- The median – the middle value or mid-point after the data have been ranked
- The mean – also known as the average
The dispersion (how data are distributed around the central tendency) could be described by:
- Inter-quartile range – the difference within the middle 50 per cent of values
- Standard deviation – extent to which the value differs from the mean
- Range – the difference between the lowest and the highest values
- Coefficient of variation – this is to compare the relative spread of data between distributions of different magnitudes, for example hundreds of tons with billions of tons (calculated by dividing the standard deviation by the mean and multiply the answer by 100)
12.5 Explore relationships, differences and trends using statistics
In a research one often wishes to find the relationship between variables. This is called hypothesis testing, where one is actually comparing the collected data with what he expected to happen. There are two general groups of statistical significance tests: the non-parametric tests (used when the data are not normally distributed) and the parametric tests( these are used with numerical data).
Testing for normal distribution
A way to test for normality is to use statistics to determine whether the distribution for a variable differs significantly from a comparable normal distribution. This could be done using statistical software that use the Kolmogorov-Smirnov test and the Shapiro-Wilk test. A probability of 0.05 means that there is a 5 per cent chance that the data distribution differs from a comparable normal distribution. Thus if the probability is lower than 0,05, the data are not normally distributed.
Testing for significance
If a there is a relationship between variables than the researcher will reject the null hypothesis and accept the alternative hypothesis. It is difficult to obtain a significant test statistic with a small sample, by increasing the sample size more relationships found will be significant. This is because the sample size resembles that of the population from which it was selected.
Type 1 and 2 errors
A Type 1 error occurs when the null hypothesis has been wrongly rejected and the alternative hypothesis should not have been accepted. In other words, the researcher states that two variables are related when they are actually not. Statististical significance is the same as determining the probability of making a Type 1 error. A Type 2 error is when a researcher does not reject the null hypothesis when he should. Thus he states that two variable are not related when they actually are.
When descriptive or numerical data are summarised as a two-way contingency table it is helpful to use a chi square test. A chi square test makes it possible to determine how likely it is that two variables are associated. In order to do this test two assumptions should be met:
- The categories of the contingency table are mutually exclusive. Each observation falls into one category only
- Not more than 25 per cent of the cells can have expected values of less than 5. When the table consists of two rows and two columns, no expected values can be less than 10.
Exploring the strength of a relationship
There are two kinds of relationships:
- Correlations: this is when a change in one variable leads to a change in another variable, but it is not clear which variable has caused the other to change
- Cause-and-effect relationship: when a change in one or more variables cause a change in another variable
The correlation coefficient quantifies the strength of a linear relationship between two ranked or numerical variables between a number of +1 and -1. A value of +1 means positive correlation, which means that the two variables are exactly related and when one increases, the other one will increase as well. A value of -1 demonstrates a negative correlation, where the two variables are precisely related, but when one increases the other one decreases.
Chapter 13 – Qualitative data analysis
13.2 Qualitative data
Often, qualitative data is associated with an interpretive philosophy because researchers tend to explore the subjective and socially constructed meanings of the participants of the research. Social constructionism means that meanings are dependent on people’s interpretations of events that occur around them. Because qualitative data depend on interpretation they are usually more complex than quantitative data.
13.3 Approach to analysis
Researches begin from using either an inductive or a deductive approach. When using a deductive approach one seeks to use existing theory. On the other hand, an inductive approach requires one to build up a theory that is adequately grounded in the data.
13.4 Preparing for data analysis
Qualitative research interviews are normally audio-recorded and then transcribed, which means that they are reproduced as written (usually with use of a word processor) account using the actual words. Along with the actual words, a researcher also has to note down the tone in which they were said as well as the non-verbal communication of the interviewee. The researcher should also make sure that the transcription is accurate by correcting for errors, this is known as data cleaning.
13.5 Tips for the analysis
There are many ways of recording information to supplement notes and transcripts:
- Interim summaries – these can be made during the analysis and outline what you have found so far, whether you trust your findings and what you need to do to improve the quality.
- Transcript summaries – these compress long statements into shorter ones in which the key element of what was said is rephrased in a couple of words.
- Document summary – describes the goal of a document and lists a few key points.
- Self-memos – these record ideas that occur to the researcher about any aspect of the research
- Research notebook – this is alternative to record ideas about the research
- Reflective diary – in this the researcher writes his reflections about his experiences of undertaking the research, what he has learnt, and how he will seek to apply his learning as the research progresses.
13.6 Approaches to analysis
In this section the generic approach to analysing qualitative data will be provided. This approach consists of five aspects:
- Categorising data: creating categories into which the data will be divided
- Unitising data: the units of data are attached to the appropriate categories that you have devised
- Examining relationships and creating categories: analysing the rearranged data.
- Developing testable propositions: the existence of relationships need to be tested if one is able to conclude that there is a relationship, by developing testable propositions. This could be done by seeking alternative explanations and negative examples.
- Drawing conclusions: interpreting and analysing the data
According to Miles and Huberman (1994) the process of analysis consist of three concurrent processes: data reduction, data display, and drawing conclusions. With data reduction they mean summarising and simplifying the collected data and/or selectively focusing on specific parts of these data. Data display focusses on organising and assembling the data into summary diagrams or other visual displays. All data that is not summarised or reduced is called ‘extended text’. Data displays allow the researcher to make comparisons between aspects of the data and to explore relationships, key themes, patterns and trends.
13.7 Inductive procedures
Grounded Theory Method
The goal of grounded theory methods is to use an inductive approach to develop a grounded theory around the key category that emerges from the data. Different kinds of coding could be used to analyse the data:
- Initial or open coding - the disaggregation of data into smaller units
- Focused coding - reanalysing the data in order to test which of the initial codes may be used to categorise larger units of data
- Axial coding - the process of discovering relationships between categories
- Labelled selective coding - the integration of categories around a core category in order to generate a theory.
When conducting grounded theory research it could be helpful to use theoretical sampling to find a sample. Theoretical sampling is choosing samples following analysis of initial data to further develop analytical categories and concepts. In order to this one it is important to constantly compare the collected data with the categories and concepts being used. Theoretical sampling is used until theoretical saturation is reached.
A template is a list of the codes or categories for the themes discovered from the collected data. This type of analysis uses both a deductive and an inductive approach in order to analyse the codes. Other than in Grounded Theory, Template Analysis permits prior specification of codes to analyse data, while Grounded Theory tries to hold everything as purely inductive as possible. Also, Grounded Theory is more structured than Template Analysis. Just as in Grounded Theory data are both coded and analysed to discover themes, patterns and relationships. The template approach enables the researcher to display the codes and categories hierarchically.
This is the process of collecting and analysing of strategically selected cases in order to empirically establish the causes of a particular phenomenon. An explanation is developed by extensively examining the process being explored. This is done through repeated cycles of developing and testing propositions. Unlike Grounded Theory, this approach focuses more on existing knowledge and theory than on participant’s data.
Using this approach, a researcher collects data through narratives such as experiences of the participants. Narratives cannot be easily fragmented since the essence of the story might be lost in the process. Instead, the narratives need to be either left intact, or they need to be ‘re-storied’, into new narratives in a more coherent form.
This is a general term that covers a very wide variety of approaches to the analysis of language. It also explores how and why individuals’ language is used by individuals in particular social contexts. Put differently, this approach explores how language (discourse) in the form of speech an text reproduces/changes the social world (Phiillips 2002) . Researchers who use this approach are often subjective ontologists.
13.8 Deductive approaches
This approach is concerned with predicting a pattern of results based on theoretical propositions in order to explain what the researcher expects to find from analysing the data. In order to do this the researcher will need to develop a conceptual or analytical framework, using existing theory, and test the adequacy of this framework as a method to explain the findings.
This approach involves an attempt to build an explanation by collecting data and analysing them, rather than testing a predicted explanation. This approach is similar to Grounded Theory but is designed to test a theoretical proposition while Grounded Theory is designed to construct a theory inductively.
Add this content to my World Supporter Magazine
Summaries & Study Note of formerworldsupporters
Access level of this page
- WorldSupporters only
- JoHo members