1.0 Introduction – objectives
2.0 Why do we need to understand how theory is developed?
3.0 The great debate – a choice of philosophies
4.0 The three different research strategies
5.0 Making your study credible
6.0 The use of hypotheses and objectives
8.0 References / Further
The learning objectives of section a:
a) To provide with you an overview of the concepts of theory development.
b) To look at different methods for developing theory.
c) To identify what hypothesis and objectives are, and when and how they should be used.
2.0 Why do we need to understand how theory is developed?
In undertaking research you have to make choices about how to develop theory. The choice of an approach to theory development will guide you as to what type of study you will follow e.g. exploratory or descriptive, the type data you will seek and how it will be collected. Finally, the approach to theory development provides a conceptual framework about how to use each element of the research and thus create new or amend existing theory. Even within the scope of a MA dissertation/project you will add to the existing body of knowledge of a subject and therefore create theory. However, there are rules to be followed about what makes for good or bad theory development. This section looks at these rules and places them in the broader context of how they influence the design of research strategy.
According to Easterby-Smith et al (2008) there are three main reasons why you should understand how to develop theory:
I) It helps to clarify research design – the way data is collected and analysed. It is the overall configuration of a piece of research: what kind of evidence is gathered, from where, and how such evidence is interpreted to provide answers to the research questions.
II) Knowledge of research philosophy can help the researcher recognise which designs will work and which will not. This should enable the researcher to avoid going up too many blind alleys and indicate the limitations of particular approaches.
III) Knowledge of research philosophy can help the researcher identify, and even create, designs that may be outside his or her past experience. It may also help the researcher to adapt research designs according to the constraints of different subject or knowledge structures.
3.0 The great debate – a choice of philosophies
How theory should be generated has been an area of great debate for a very long time. Indeed some of the first forms of ‘scientific approach’ to theory development date back to the Ancient Greeks. More recently we have seen the creation of two distinct schools of thought about how theory can be developed. The two schools are classified as either “Positivism” or “Phenomenology”. In very general terms each of the schools provide standpoints about how research should be conducted, what theory can be generated and what cannot be considered to be good theory.
Theory – a statement that summarises and organises knowledge by proposing general relationships between events.
Positivism (scientific method) - an approach that starts with a theory of how something works and tests via experimentation whether it is true.
Phenomenology – an approach that seeks to create theory from the identification of what exists and how things work without using preconceived ideas.
3.1 The Positivist approach
The positivist approach is often described as scientific research in so much that a stated proposition is tested e.g. a stated proposition is “water gets hotter when it is heated”. Using what is known the research conducted would be to heat some water and use a thermometer to track changes in temperature. The scientific approach also allows us to control the experiment by altering some of the variables. For example, the temperature at which pure water boils is 100 degrees centigrade, if salt were added to the water, the point at which it boils may alter. As a research strategy this is attractive in that by controlling the variables any changes should be identifiable, as are the reasons why a change has occurred.
Robson (2002) states that the scientific approach tends to have five sequential steps:
I) Deducing a hypothesis (a testable proposition about the relationship between two or more events or concepts) from the theory.
II) Expressing the hypothesis in operational terms (i.e. ones indicating exactly how the variables are to be measured) which proposes a relationship between two specific variables.
III) Testing this operational hypothesis. This will involve an experiment or some other form of empirical enquiry.
IV) Examining the specific outcome of the enquiry. It will either tend to confirm the theory or indicate the need for its modification.
V) If necessary, modify the theory in light of the findings. An attempt is then made to verify the revised theory by going back to the first step and repeating the whole cycle.
Easterby-Smith et, al. (2008) add further insight into the way positivism functions by categorising the research philosophy as being based on two assumptions:
1 “Reality is external and objective”
2 “knowledge is only significant if it is based on observations of this external reality”.
Therefore the theory developed is based on deductive reasoning (defined as logical reasoning that something must be true because it is a particular case of a general law that is known to be true). Based on this they amended a typology of features that describe positivism formulated by Comte (1853).
II) Value freedom. The choice of what to study, and how to study it, can be determined by objective criteria rather than human beliefs and interests.
III) Causality. The aim of the social sciences should be to identify causal explanations and fundamental laws that explain regularities in human social behaviour
IV) Hypothetico-deductive. Science proceeds through a process of hypothesising fundamental laws and then deducing what kind of observations will demonstrate the truth or falsity of these hypotheses.
V) Operationalisation. Concepts need to be operationalsed in a way which enables facts to be measured quantitatively.
VI) Reductionism. Problems as a whole are better understood if they are reduced to the simplest elements.
VII) Generalisation. In order to be able to generalise about regularities in human social behaviour, it is necessary to select samples of sufficient size.
VIII) Cross-sectional analysis. Such regularities can most easily be identified by making comparisons of variations across samples.
Both lists present a logical process to conducting research that starts with the assumption that the researcher has a high level of understanding of how things work. This allows the researcher to control variables and thus test a stated hypothesis. The presumption is that you have a relatively high level of knowledge of how a specific relationship exists and can produce a testable hypothesis.
In management oriented studies it is more likely that the focus of the research will not be to test a hypothesis but to generate one. So the research is conducted to identify variables that may then be capable of being tested using an experimentation procedure. NB This may be a more appropriate way for you to consider using the scientific method in examining some aspect of management theory.
3.2 The Phenomenology approach (sometimes referred to as “naturalistic enquiry”)
Phenomenologists tend to reject the scientific approach when applied to management studies. Indeed the very notion that the focus of study can be pre-defined without looking at data is perceived by them to be a major flaw in scientific method. Specifically there is a belief that the notion that all aspects of a study are controllable so that a hypothesis is testable is seen to ignore the fact that social science research is to do with the individual.
These individuals are unique, operating in different social environments that influence their behaviour and deal with situations based on a context that in itself is highly variable. This tends to lead to the observation that a phenomenology research based approach is about the identification of how things exist within the context of the often messy real world. Certainly management is a subject that exists as a business activity and functions in the competitive environment complicated by the aspirations and activities of organisations.
Thus it is the combination of many factors that influence how management works, all of which may be relevant to a research project, but may not be controllable or sufficiently understood to set and test a hypothesis.
Underlying the concept of phenomenology is that the researcher is part of the process and helps to interpret what is perceived based on their experiences. This makes the researcher part of the study (internal rather than external), further that findings are acknowledged to be subjective because they are based on inductive reasoning. (defined as a general law that exists because particular cases that seem to be examples of it exist)
Robson (2002) adapted the work of Lincoln and Guba (1985 pp. 39 – 45) to illustrate the characteristics of a naturalistic enquiry. Take some time to review each of the 14 characteristics and compare them to the characteristics of a positivistic approach put forward by Easterby-Smith et, al. (2008)
I) Natural setting – research is carried out in the natural setting or context of the entity studied.
II) Human instrument – the enquirer(s), and other humans are the primary data-gathering instruments.
III) Use of tacit knowledge – tacit (intuitive, felt) knowledge is a legitimate addition to other types of knowledge.
IV) Qualitative methods – qualitative rather that quantitative methods tend to be used (though not exclusively) because of their sensitivity, flexibility and adaptability.
V) Purposive sampling – is likely to be preferred over representative or random sampling, as it increases the scope or range of data exposed and is more adaptable.
VI) Inductive data analysis – is preferred over deductive as it makes it easier to give fuller description of the setting and brings out interactions between enquirer and respondents.
VII) Grounded theory – preference for theory to emerge from (be grounded in) the data.
VIII) Emergent designing – research designs emerges (unfolds) from the interaction with the study.
IX) Negotiated outcomes – preference for negotiating meaning and interpretations with respondents.
X) Case study reporting mode – preferred because of its adaptability and flexibility.
XI) Idiographic interpretation – tendency to interpret data idiographically (in terms of the particulars of the case) rather than nomothetically (in terms of law-like generalisations).
XII) Tentative application – need for tentative (hesitancy) in making broad applications (generalisations) of the data.
XIII) Focus-determined boundaries – boundaries are set on the basis of the emergent focus of the enquiry.
XIV) Special criteria for trustworthiness – (equivalent to reliability, validity and objectivity) to be devised which are appropriate to the form of the enquiry.
3.3. The advantages and disadvantages of the two approaches.
It is perhaps not a safe assumption that you will stick to any one-research approach when designing your own study. Many researchers tend to think first about the information needs of the study and then the methods, which often blend different elements of positivist, and phenomenological design concepts together. There is nothing inherently wrong with this, indeed you may find that it is the only way to actually collect the data that most suits the needs of your study. It is also important to realise that the two approaches, although presented as opposites that cannot meet, often do. More of how to combine research methods to gain data is discussed later on.
Saunders et, al. (2007) summarise the key advantages and disadvantages of each research design approach:
v Economical collection of large amounts of data.
v Clear theoretical focus for the research at the outset.
v Greater opportunity for the researcher to retain control of the research process.
v Easily comparable data.
v Facilitates understanding of how and why.
v Enables researcher to be alive to changes as they occur.
v Good at understanding social process.
v Inflexible – direction often cannot be changed once data collection has started.
v Weak at understanding social processes.
v Often doesn’t discover the meaning people attach to social phenomena.
v Data collection can be time consuming.
v Data analysis is difficult.
v Researcher has to live with the uncertainty that clear patterns may not emerge.
v Generally perceived as less credible by non-researchers.
3.4 Relevance of research methods to you.
So far the discussion of research methods has been more orientated to the theoretical than to the practical. You may be asking yourself why is this relevant to me, after all I am producing a MALIC dissertation or project and not a PhD. Well the simple truth is that whether you are producing an MA dissertation/project or a PhD, you are still concerned with seeking to identify the ‘truth’ of something via your secondary and primary research that is transparent to others who will read and use your work.
At a very practical level the issue of whether you are fundamentally a positivist or a phenemenologist is about how you devise your research strategy. Easterby-Smith et, al. (2008) provide a useful summary of what the two approaches to research design mean in terms of guiding choice for research strategy:
Key features of positivist and phenomenological paradigms.
(Easterby-Smith et, al. 2008)
v The world is external and objective.
v Observer is independent.
v Science is value-free.
v The world is socially constructed and subjective.
v Observer is part of what is observed.
v Science is driven by human interests.
v Focus of facts.
v Look for causality and fundamental laws.
v Reduce phenomena to simplest elements
v Formulate hypotheses and then test them.
v Focus on meaning.
v Try to understand what is happening.
v Look at the totality of each situation.
v Develop ideas through induction from data.
Preferred methods include:
v Opertionalising concepts so that they can be measured.
v Taking large samples.
v Using multiple methods to establish different views of phenomena.
v Small samples investigated in-depth or over time.
The comparison of key features is very useful in that it allows you to identify how the research paradigms help formulate overall strategy for a study and what the options are once the method is chosen i.e. the tactics of the research project.
The issues to be considered are:
Ø Do you want to conduct experiments, a survey, build a case study, or use of combination of these techniques?
Ø How do you want to use hypotheses, objectives and research questions to guide you in data collection and analysis?
Ø What approach to data collection is likely to yield results that best fit the study’s information need? – quantitative or qualitative, or a mixture of the two?
Try the quiz to see how well you have understood the outlined issues. If you score less than 8 (one point per correct answer, answers at the back of the book), review section 2 & 3. Some questions have more than one correct answer
Activity one Quiz Number 1
1) Which of the following can be considered to be roles of theory development?
a) To impress your supervisor?
b) To identify which research designs will work?
c) To understand whether your research will be credible?
2) What is a theory?
a) Something that you have made up?
b) A statement that describes in very general terms how something might work
c) A general statement that summarises and organises knowledge by proposing general relations between events.
3) Which of the following describes a positivist approach to theory development?
a) It is when you evaluate meaning from the data.
b) It is when you test a hypothesises.
c) It is when you use qualitative methods to collect data.
4) The role of the researcher in a positivist approach is:
a) To be independent of what is studied.
b) To be part of what is studied.
c) To get others to conduct the study.
5) The phenomenological approach to theory development can be described as:
a) Led by hypothesis setting and testing.
b) Based on inductive reasoning.
c) Existing only if the data can be generated via an experiment.
6) The advantages of a Phenomenological approach are:
a) It identifies the how and why of things.
b) It is very controllable as a theory generating method.
c) It is adaptive to the realities of the study.
7) The disadvantages of the Phenomenological approach are:
a) Data analysis is difficult.
b) There is a high level of certainty of what will be found out.
c) It can be vary time consuming to conduct.
8) The advantages of a positivist approach are:
a) It is an economical way of collecting a large amount of data.
b) Data from it is capable of being easily compared.
c) The data provides a good understanding of how things work.
9) The disadvantages of a positivist approach are:
a) The outcome of the data collection is uncertain.
b) It does not always identify the meaning of a phenomenon.
c) It is weak at understanding social processes.
10) In using a positivist approach the researcher should:
a) Focus on facts
b) Focus on the meaning of things.
c) Reduce phenomena to its simplest elements.
11) In using a the phenomenological approach the researcher should:
a) Develop idea through induction from the data.
b) Focus on meaning.
c) Look at the totality of each situation.
12) Do you think that the two approaches are mutually exclusive in their use?
4.0 The three different research strategies
Section 3.4. considered the link between the overall concepts of theory development and the practical issues of how to operationalise a study. This section looks at a number of techniques for collecting data. According to Robson (2002) there are three traditional research strategies:
Ø Experiment: measuring the effects of manipulating one variable on another.
Ø Survey: collection of information in a standardised form from groups of people.
Ø Case study: development of detailed, intensive knowledge about a single ‘case’ or of a small number of related ‘cases’.
Before looking at each of these forms of data collecting, it is worth just considering what the purpose of a study is. The purpose can be defined as “the reason why you are collecting data to answer a specific type of question”. This purpose should be defined within either the hypothesis to be tested and the research objectives set to be addressed. The research is opertionalised via creating specific research questions that provide the constructs for the data collection instrument.
Classification of the purpose of a study (Robson, 2002)
Ø To find out what is happening.
Ø To seek new insights.
Ø To ask questions.
Ø To assess phenomena in a new light.
Ø Usually, but not necessarily, qualitative.
Ø To portray an accurate profile of persons, events or situations.
Ø Requires extensive previous knowledge of the situation etc. to be researched or described, so that you know appropriate aspects on which to gather information.
Ø May be qualitative and/or quantitative.
Ø Seeks an explanation of a situation or problem, usually in the form of causal relationships.
Ø May be qualitative and/or quantitative.
The purpose of your study ultimately depends on what the starting point of it is. Most MA dissertations/projects tend to fall into the category of exploratory research. This is because of the timescale for completion of the study, the resources available for data collection, and perhaps most importantly the students’ relatively weak subject knowledge base. This does not mean that your study has to be purely exploratory, indeed you may wish to test an existing model. Thus, depending on the overall focus, a study may be either descriptive or explanatory.
4.1. Experiment: measuring the effects of manipulating one variable on another.
The concept of experimentation is the one that most students are familiar with, having been taught the key principles in science lessons whilst at school. However, the key difference from what we learnt at school and experimentation for management research is that the objects of the study are human rather than chemicals. The basic principles still hold true. According to Leary (1991) a well-designed experiment has three essential properties. These are:
I) The researcher must vary at least one independent variable to assess its effects on subjects’ behaviour.
II) The researcher must have the power to assign subjects to the various experimental conditions.
III) The researcher must control extraneous variables that may influence subjects’ behaviour. (Leary, 1991, p. 122)
This can be further refined into the following elements: Saunders et, al. (2007)
I) the definition of a theoretical hypothesis;
II) the selection of samples of individuals from known populations;
III) allocation of samples to different experimental conditions;
IV) introduction of planned change on one or more of the variables;
V) measurement on a small number of the variables;
VI) control of other variables.
EXAMPLE OF EXPERIMENTAL DESIGN.
A study into different forms of advertising medium for recruitment– the Internet versus radio advertising.
Exploratory (low level of knowledge of Internet advertising for recruitment)
A radio advert is more likely to be remembered than an Internet advert.
Three groups of 50 people.
The groups represent the socio demographics of the
An advert is designed that conveys the same message for delivery by the radio in verbal format and by the Internet in pictorial and written format.
One group of 50 is exposed to both the radio and the Internet advert (comparison of medium).
One group of 50 is exposed only to the radio advert (to focus on specific attributes of medium).
One group of 50 is exposed only to the internet advert (to focus on specific attributes of medium).
Data collection tool:
Self completion questionnaire
Number of times exposed to advert
Introduction of a distraction
Adding a verbal element into the internet advert
The extent to which each group recalls the advert
The elements of each advert that can be recalled
The elements of the advert that could not be recalled
Take a few minuets to try activity two
Look at the example of an experiment and consider the following questions.
1) Which parts of the experiment can be considered to be constants?
2) Which parts of the experiment can be manipulated by the researcher?
3) How do you think the experiment can be improved?
Before you go down the experiment route as a way to collect data you need to be aware of its limitations. These will help to guide you as to whether to use this type of method.
Ø Can you define the elements of the research to the level that a hypothesis can be set?
Ø How adequately can you control aspects of the experiment?
Ø How much information do you lose by not having a real world exposure to the research?
Ø Will an experiment provide the opportunity to identify meaning as well as causality?
Ø If the actual subject focus exists outside of a laboratory how meaningful are laboratory-generated results?
4.2 Survey: collection of information in a standardised format from groups of people.
Most of us tend to have an understanding of the survey approach to data collection, in that we have been targeted by market researchers on the street or through the post and asked to complete a questionnaire. The very word survey is synonymous with the idea of dipping into a sample at a particular point in time and gathering standardised data in the form of a questionnaire. Also we tend to describe a survey as quantitative data (though qualitative surveys also exist).
“Survey research … is the method of collecting information by asking a set of pre-formulated questions in a predetermined sequence in a structured questionnaire to a sample of individuals drawn so as to be representative of a defined population.” (Hutton, 1990: p. 8)
Ø You have sufficient grasp of the subject issues to formulate standardised questions.
Ø There is a large definable population from which to select potential respondents (see topic 4).
Ø The research need is to generate generalizable data that relates to large sets of respondents rather than identifying the activities of a single case.
Ø You intend to collect a relatively small amount of data from an individual (compared to qualitative method) to verify a variable or test the intensity of a response.
Ø You intend to use statistical methods to help in the analysis of the data.
Ø You do not need to manipulate the way the data is collected in the same way as you would in an experiment.
Ø You intend to replicate the survey, therefore standard questions and an identified sample are vital.
B) Ease of use of a survey method.
The survey method is perceived as a relatively easy way of collecting a large amount of data in a relatively short space of time. It is also seen to be a straightforward way to operationalise a study given that it has very definable sequential steps.
Ø Conduct a literature search to identify possible issues to be translated into specific questions.
Ø Design a questionnaire.
Ø Identify potential respondents and sampling technique.
Ø Pilot questionnaire / make amendments.
Ø Send out survey.
Ø Follow up to increase response rates.
Ø Conduct analysis either manually or by use of an analysis package.
Ø Report findings.
Each of these steps takes considerable thought and effort to get right.
The impact of culture change on staff motivation
Descriptive (builds on work already conducted in the area with the research need to verify an existing model)
To identify the extent to which culture change has a positive/negative impact on staff motivation
A census of staff in YSJ
The first questionnaire is sent to all elements of the sample.
The second follow up questionnaire sent to those who responded to the first.
All questionnaires are coded to enable them to be matched to the first questionnaire.
The questionnaires are coded and loaded into SPSS.
Frequencies are calculated to ensure quality of the data.
The data set was then subjected to a number of different statistical techniques
a) Design of questionnaire 1 = two weeks.
b) Design of questionnaire 2 = two weeks.
c) Pilot testing for each tool was 1 weeks.
d) Getting the sample together – 2 weeks.
e) Sending out questionnaire one and getting responses 3 weeks.
f) Sending out of questionnaire 3 weeks.
Activity three: Take a few minutes to identify answers to the following questions
1) Looking at the example of a survey method to data collection, what additional sample details do you think would be useful?
2) What type of disadvantages do you think exist in the survey method described?
4.4. Case study: development of detailed, intensive knowledge about a single ‘case’ or of a small number of related ‘cases’.
According to Robson (2002) a case study is defined as:
“Case study is a strategy for doing research, which involves an empirical investigation of a particular contemporary phenomenon within its real life context using multiple sources of evidence”
A) Conditions of use for a case study.
Robson’s definition suggests that a case study should incorporate some if not all of the following elements:
Ø A specific intent to describe the whole focus of the research, not just disjointed component parts.
Ø To ensure that the overall data collection activity is in line with a overall strategy about how the data will be used.
Ø That there should be a use of a number of different data collection methods to expose totality of the phenomena under investigation.
Ø That the study has to be conducted within the context of the phenomena’s existence, and that it is written up to reflect the context.
Ø That the objects of the study have a high level of interdependence.
Ø That the study must indicate how generalisable the theory from the case is to other research.
Ø That the study is only to be concerned with a very small sample or cases - perhaps just one.
B) Ease of use of a case study tool
The question of how easy it is to use case studies as an approach to collect and describe a researched activity is not easy to answer. Good case study research at masters level tends to be hard to achieve for the following reasons:
Ø You need a high level of access to an organisation to apply different types of data collecting methods.
Ø The time needed to collect data in this way is often much greater than for less complex methods.
Ø The ability to write up case studies is partly about your skills to collect data, but also has a great deal to do with your writing skill.
Ø Ordering and comparing all data collected by multiple sources is highly complex, very subjective and time consuming.
4.5 Multiple methods
The previous sections tend to suggest that a research project falls into one of two types of theory paradigms (positivist or phenomenological), and that you have three types of data collection methods (experimentation, surveys & case studies). The reality of designing a research study is that the information need, the knowledge base of the researcher, the ability to gain access to the desired sample, and resource issues all help determine which theory approach and data collection methods are most suitable. It is rare for good studies to stick to one form of approach or method, instead the study is divided into stages that require different ways of gaining and interrogating the data. You may want to consider the use of different methods for the following reasons:
Ø Your knowledge base as a student on any given topic at the start of a study is relatively low. Therefore to be able to complete your dissertation you will need a strategy that helps to build specific subject knowledge, before any form of testing is undertaken.
Ø Research process is about producing building blocks. The process starts with an idea, which is developed by collecting and thinking about theory described in the literature. This is perhaps refined by conducting a number of face to face interviews with subject experts to develop an outline theory. Finally, a questionnaire is sent to a large sample to verify the constructs of the theory.
Ø Different approaches can be used as a way to assess the honesty of the study. It can be used to ask and answer the question of whether each of the data collection methods provides similar types of data. The use of multiple methods is sometimes refereed to as triangulation.
What governs your choice of approach or method is really dependent on what you think is the best way to get the data that answers the research need. For the type of study you should be contemplating, it would be wrong to prescribe a “best” way. There is no perfect method; all have distinct advantages and disadvantages which vary depending on the situation of use. You will be the best person to ultimately make the judgement about how to blend different approaches and methods together. This is something that should be discussed at length with your supervisor.
5.0 MAKING YOUR STUDY CREDIBLE
In assessing your dissertation the examiners will be asking whether you have followed the accepted practices that make for good research. In essence this means establishing credibility of your data set. This includes the creation of objectives or hypothesis, collecting and discussing literature, selecting and using a theory development approach, and completing and disseminating primary data via the choice of appropriate methods.
This is no easy task and it safe to say that students who master the research process are the ones who can give an honest appraisal of how far they have succeeded in creating credibility in their work. The overall concept of credibility can be translated into three areas:
QUESTIONS OF RELAIBILITY, VALIDITY AND GENERALISABILTY (Easterby-Smith et al., 2008)
Does an instrument measure what it is supposed to measure?
Has the researcher gained full access to the knowledge and meanings of informants?
Will the measure yield the same results on different occasions (assuming no real change in what is to be measured)?
Will similar observations be made by different researchers on different occasions?
What is the probability that patterns observed in a sample will also be present in the wider population from which the sample is drawn?
How likely is it that ideas and theories generated in one setting will also apply in other settings?
The chances of undertaking research that can be considered to be valid and reliable are much improved if these are considered to be design issues rather than things that occur by chance. How you build validity and reliability into your study will depend on how much time you spend planning each stage of the activity, also it is about having an awareness of the factors that compromise credibility. These are the threats to credibility.
5.1. Threats to Credibility
The most obvious threat to credibility in research is lack of effort made by the researcher. You have a responsibility to embrace each stage of your study with due regard to putting quality into it and making sure that the rules which define credibility are followed. The difference between a good thesis and a mediocre one comes down to the amount of effort and regard to design issues made. In addition it is worth looking at other threats that may or may not be within your control. Reliability and validity have different threats and therefore it is wise to consider each of these in relation to the factors that compromise their existence. (for a more extensive discussion see Robson, 2002)
I) SUBJECT ERROR – For example, a study examining undergraduate class participation levels. You measured a group’s participation in a Monday 10.15am. class. The exercise is repeated with the same students on a Friday Morning at 10.15am. The results are compared and show a high level of variance. Question is why? Possible explanations are a) it is the end of the week and the students are tired, b) it was students night at a local club and most of the class were partying till the early morning. Knowledge that this phenomena exists would suggest that the study should perhaps try and collect data say on a Tuesday and Thursday to get a more realistic picture of the true level of participation in class by students.
II) SUBJECT BIAS – For example, you conduct a study into employee satisfaction. The respondents might feel that the study results will be directly attributable to them and used by their boss. Therefore instead of giving you the truth they say what they think their boss would want to hear.
B) Validity - Threats to internal validity.
I) HISTORY – Things that have changed in the participants’ environment other than those forming a direct part of the enquiry e.g. occurrence of a major air disaster during study of effectiveness of a desensitisation programme on persons with a fear of air travel.
II) TESTING – Changes occurring as a result of practice and experience gained by participants on any pre-tests, exposure to the research tool alters their views.
III) INSTRUMENTATION – Some aspect(s) of the way participants were measured changed between pre-test and post-test e.g. significant changes made to the data collection tool resulting from a pre-test, that are not subjected to a further pre-test.
IV) REGRESSION – If participants are chosen because they are unusual or atypical, later testing will tend to give less unusual scores. This is relevant if matched groups are used who are exposed to on-going remedial training to change skill levels; therefore the measured variance over time may well disappear.
V) MORTALITY – Participants drop out of the study. This is especially relevant if you intend to conduct follow up research with a matched sample.
VI) MATURATION – Growth, change or development in participants unrelated to the treatment of the enquiry. For example you may be looking at the effects of “pester power” on parents with 3-year-old children. Over even a short period of time the child’s development in their ability to be influenced by pester power advertising may change significantly.
VII) AMBIGUITY ABOUT CAUSAL DIRECTION – Does A cause B, or B causes A? How certain are you that you have identified and described how the cause of an event actually occurs? So does new product development fail due to poor market research, or is it that the research is not used appropriately by the design team?
Adapted typology originally created by Cook and Campbell, 1979, pp 51 – 55 (in Robson 2002)
C) Validity – Threats to external validity
The concept of generalisability can be considered to be a test of external validity. Generalising your findings means applying what you have found to different contexts. For example, you may undertake a study looking at how the Internet alters the way new product are developed and launched. You undertake the study in financial services and identify a distinct set of issues that relate solely to the use of the Internet. Can the issue you have identified be used to describe the NPD process in say the leisure industry or a manufacturing sector? If they can then it is safe to say the findings are generalisable and have a wider use than to be applied only to your original sample. If not, then the reasons why it cannot be used in a generalised sense should be explained. According to LeCompte & Goetz, (1992) there are reasons why studies may not be generalisable. These are:
I) Selection – findings being specific to the group being studied.
II) Setting – findings being specific to, or dependent on, the particular context in which the study took place.
III) History – specific and unique historical experiences may determine or effect the findings.
IV) Construct effects - the particular constructs studied may be specific to the group studied.
From this section you should be able to distinguish what makes credible research and what does not. You should be familiar with the ideas of ‘reliability, validity and generalisability’ at a conceptual level and why you should use these concepts to plan your study.
Finally, and by far the most important learning point, is that credibility in research is something that you have to plan for from the start. You cannot rely on it occurring by chance.
Activity four Attempt Quiz Number 2. Some question will have more than one answer. The answers are in appendix one. If you find it hard to answer the question take some time to review the section.
1) Exploratory research is concerned with:
a) Describing what has occurred.
b) Finding out what is happening.
c) Asking questions.
2) Descriptive research is concerned with:
a) Assessing phenomena in a new light
b) Portraying an accurate profile of persons, events or situations.
c) Seeking to identify causalisation.
3) The qualities of a good experiment are:
a) The researcher must vary at least one independent variable to assess its effects on subjects’ behaviour.
b) All of the variables must be viewed as uncontrollable to allow greater insights into meaning.
c) An experiment usually is hypothesis driven.
4) A qualities of a good survey are:
a) That there is a sample that can be surveyed.
b) That the questions sought to be asked can be standardised.
c) That the knowledge base of the research be has to be low.
5) The qualities of a good case study are:
a) That a study has been conducted within the context of the phenomena’s existence.
b) That the study has sampled many different organisations.
c) That the objects of the study have a high level of interdependence.
6) Triangulation is best described as:
a) Using three references in the literature review to support a point being made.
b) Using a mixture of primary and secondary sources to identify if an appropriate meaning of something has been identified.
c) It is asking a set of varied questions to respondents that cover the same issues.
7) Credibility in research is best described as:
a) Targeting individuals who you know well to gain primary data.
b) Adopting a process to the research that involves some form of triangulation.
c) Having a high level of transparent honesty in the study.
8) Validity in research occurs when:
a) The test instrument measures what it purports to measure.
b) The basic constructs of a test instrument are developed using some form of triangulation technique.
c) The researcher gains full knowledge and meaning of informants.
9) Reliability in research occurs when:
a) The test instrument will give the same results if the study was replicated under the same conditions as the original study.
b) Different researchers will identify the same meaning from the data set.
c) The study is conducted without an understanding of any factors that may alter response intensities over the period of data collection.
10) Credibility can be improved if:
a) The design of the study considers the issues of validity from its outset.
b) The researcher understands that specific approaches to theory development require different measures of credibility.
c) A rigorous process of theory building is employed based on the use of triangulation.
6.0 THE USE OF HYPOTHESES AND OBJECTIVES
So far you have looked at the issues that define your approach to a study, and you have considered how to ensure there is credibility in what you do. The final part of this topic is considering how to make your study operational. This is about choosing whether to be predominately hypothesis driven or objective driven. This section explores what is meant by these terms, and when to use them.
6.1 What is a Hypothesis?
A Hypothesis is defined as “an unproven statement or proposition about a factor or phenomenon that is of interest to the research.” Malhotra (2007)
A study can be hypothesis driven. The discussion on the scientific approach suggests that this type of research lends itself to the testing of hypothesis. This means that you set a hypothesis (which is a theory of how something works) and then evaluate whether the data generated from the research actually confirms or rejects the stated assumption.
How and when you generate a hypothesis is a key design issue. Hypothesis tend to relate to the positivist approach of theory generation. In setting hypothesis you are making judgements about the nature of the research variables and whether they can be manipulated so as to be tested. Thus to be able to set testable hypothesis you must have a good, if not a high level of knowledge of the research subject.
You also will need to be able to design a study that will generate data that can be used to prove or disprove the hypotheses. A question for you to answer before you design a research instrument to test a hypothesis is whether you be able to control the stated variables in a meaningful way i.e. operationalise the study?
6.2 What are research objectives?
Research objectives are a succinct summary of what you hope to achieve in your study. They give a focus to the study and should emphasise what will be examined.
Setting objectives is a very important task because they define the exact nature of the study you are going to undertake. Your objectives should provide some or all of the following information:
Ø The subject you are interested in e.g. product elimination.
Ø The relationship that you want to consider e.g. how product elimination is conducted.
Ø The industry or sector that will provide the context for the research e.g. product elimination in the financial services sector.
Ø Whether it is a stand alone study or a comparison study e.g. a comparison of product elimination from the physical goods and financial services sectors.
Ø What you want to gain from the study e.g. new insights into how product elimination is conducted.
As you can see, the type of guidance that objectives provide tend to relate to the broad constructs of phenomenological research e.g. looking for meaning from what is observable rather than testing an assumption of how something works. For most MSc students the research they conduct falls into this approach. The type of studies conducted tend to be very exploratory in nature and can be classified as one of the following types:
Ø Research into new concepts of management that have little previous research exposure.
Ø Research that takes theory and tests its applicability in a different business context.
Ø Research that focuses on identifying specific sector practices to identify an element of management theory. (PROJECT)
6.3. Objective setting
It is likely that you will have to set objectives for your study. What are the rules? Well the rules for objective setting are relatively straightforward.
Ø Start of with very general objectives. These should be sufficient to give you an overall focus for the study but not to tie your hands until you have done more background reading.
Ø Be prepared to refine your objectives over a period of time by creating a number of iterations of them.
Ø Write objectives that actually link the theory you are interested in to the practical issues of what you want to find out.
Ø Ask yourself key questions:
Ø Are the objectives practical?
Ø Are there too many objectives to achieve with available resources?
Ø Are there too few so that you have not given a focus to your study?
Ø Are the objectives too wide?
Ø Are the objectives too narrow?
Ø Agree you objectives with your supervisor as early as possible.
A very rough rule of thumb says that you should have at least two objectives and no more than about four. If you have ten objectives then you need to refine your research parameters to reduce the total number.
FINALLY ONCE YOU AGREE YOUR OBJECTIVES KEEP THEM TO THE FOREFRONT OF YOUR MIND. ALL ASPECTS OF YOUR THESIS RELATE BACK TO THE OBJECTIVES!
6.4 Are research objectives and hypothesis mutually exclusive?
A common question asked by research students is “should I have a hypothesis and an objective in my study? You should by now have an idea as to answer this. The rules governing the formulating of research parameters (after all that is what setting objectives or posing a hypothesis does) do not specifically exclude using both within a study. But what an understanding of research methods should suggest is that objectives and hypothesis relate to different data collecting and theory generating processes. Therefore, you can have research objectives that are used to generate a hypothesis, but you probably should not use hypotheses to generate the overall research objectives.
For example. You could formulate an overall objective that says you wish to examine the use of the Internet as an advertising medium. Within this you could test for a specific hypothesis e.g. the level of browser used by individuals on their PCs determines their receptivness to Internet advertising messages.
6.5 Link between research objectives or hypothesis and research questions
Whether your study is objective or hypothesis driven, there is still a need to translate these constructs into instructions for collecting data. These instructions are usually termed research questions. One way of looking at it would be say that objectives and hypotheses give guidance into the areas of the study, whereas the research question spell out specifically what you intend to find out.
The link between objectives and research questions is very strong but not direct in the sense that once you create objectives you cannot sit down and write out your research questions. Instead you have to think about the process of generating knowledge. Objectives establish the broad parameters of the study; they guide your choice of literature, and influence the overall approach to the study. Research questions, if the process has been followed, should fall out of the literature review that you have conducted. Thus it is through the literature that research questions link back to the objectives. The research questions should also have a transparent relationship to what is tested for in the primary data collecting activity. The research questions should be capable of being seen in the questionnaire or in the topic guides used for interviews.
Thinking about your own study.
a) Do you think you will follow a positivist or a phenomenological approach to theory development?
b) What is your rationale for this choice?
c) If you are going to use a positivist approach try and formulate the hypothesis for the study.
d) If you are going to use a phenomenological approach try and set research objectives
This topic set out to provide grounding into the concepts of theory development and how such an understanding can be used to guide you in your choice of research strategy.
You should now be able to distinguish between a positivist and a phenomenological approach to theory development. You should be able to identify what types of data collection methods are appropriate to either approach.
You should be aware of how the nature of a study – exploratory, descriptive and explanatory, impact on the choice of approach and the type of data collecting activities that can be conducted.
The final part of the topic dealt with the use of objectives and hypotheses and their link to research questions. You should have an appreciation of what type of studies should be objective driven or hypothesis driven.
9.0 REFERNCES / FURTHER
Blaxter L, Hughes C, & Tight M. (1996) How to Research. OU Press
Comte A. (1853) The positive philosophy of Auguste Comte, [translated: H Martineau],
Cryer P. (1996) The Research Student’s Guide to Success. OU Press
Easterby-Smith M, Thorpe R, & Lowe A. (2008) Management Research, an introduction. Sage
LeCompte M. D. & Goetz J. P. (1982). Problems of reliability and validity in ethnographic research. Review of Educational Research, 52, 31 – 60.
Malhotra N, (2007) Marketing Research: An applied orientation. 3Ed. Prentice-Hall,
Robson C. (2002) Real World Research, Blackwell
Sunders M & Thornhill A. (2007) Research Methods for Business Students. Pitman Publications
1 b & c
7 a & c
9 a & b
10 a & c
11 a & b & c
1 b & c
3 a & c
4 a & b
5 a & c
6 b & c
7 b & c
8 a & b & c
10 a & b & c
1. Introduction & objectives
2. The role of a research proposal
3. Setting objectives
4. Rationale for the study
5. Mini literature review
6. Sample Frame / sampling procedure
7. Tools for collecting primary data
8. Analysis techniques
9. Timescale, resource statement
11. Format of proposal
Producing a research proposal is a vital stage in moving forward with your study. The proposal is where you pull together various strands of information into a coherent document that states why and how you are going to conduct your research. This topic takes a practical stance in outlining the role of producing a proposal so that you can provide a blue print for your study, and more importantly a proposal that actually gets you started with your dissertation.
1.1. The learning objectives of the section B are:
a) To outline the role of each of the sections of a proposal and indicate how they help plan a study.
b) To provide you with a proposal format that encourages you to think through each of the stages of undertaking a research study.
2.0 The role of a research proposal
The role of the research proposal is often said to be that of providing a blue print for the study to be undertaken. As a document it serves a number of purposes:
a) It allows you to state the research objectives that will guide all other design decisions of your study.
b) It allows you to rationalise those objectives by summarising why you think the study is worth investigation at Masters level.
c) It allows you to build a case for the research by writing a mini literature review.
d) It allows you to outline your sample frame and sampling procedure.
e) It allows you to identify your approach to theory development and specify the tools you intend to use to collect primary data.
f) It allows you to specify the analysis techniques you will use to interpret primary data and report the findings.
g) Finally, it allows you to place all of the above in the context of the timescales for the study, resources required and potential problems that may arise.
Each of these roles of a proposal need to be considered both separately and also collectively in terms of how they link to each other. This will be dealt with in the rest of this topic. There is another role of a proposal and that is how it gets you to think about the research process.
2.1 The thought process behind producing a proposal.
Going back to the concept that the proposal is the blue print for a study, which should have been really well thought through by you. If not you may end up changing parts of the study, this may alter your ability to make progress, collect primary data or even complete a dissertation/project that cannot be passed by your examiners.
A good proposal should give you a solid foundation to complete your dissertation. This requires a great deal of fore thought and hard graft on your part So trying to write a proposal the day before it is submitted is likely to result in you failing the module and underachieving with your dissertation.
3.0 Setting Objectives
In formulating the objectives for your study you should bear the following points in mind:
a) The objectives should identify the area of theory you are interested in.
b) The objectives should identify what you want to evaluate within that theory.
c) The objectives should identify the specific context of the evaluation e.g. the industrial sector or a specific type of consumers.
d) Finally, the objectives should identify any sub-element to your study i.e. an evaluation that can only be done once you have looked at something else.
3.1 Are objectives set in stone?
Objectives outlined in the proposal are not set in stone, they can and often do change as an individuals understanding of a subject increases. Although the more up front work you conduct the less likely you are to have to re-write your objectives, you do have to be prepared to make changes. These often result from:
a) Not being able to get to your proposed sample.
b) Identifying new factors from the literature.
c) Additional guidance from your supervisor.
d) Finding out that someone else has done the identical work you proposed to do.
e) Finding something more interesting to do.
Most of these rationales for changing your objectives can be removed by careful planning and reading around the subject before you write your proposal. As for the last one, a word of caution, what sounds more interesting in a research context often turns out to be as interesting or uninteresting as your original idea. Most researchers will admit that there are always things that look much more interesting than what you have committed yourself to do!
Write down the objectives for your study.
Objective One _______________________________________________________________________________________________________________________________________________________________________________________________________________________
Objective Two ________________________________________________________________________________________________________________________________________________________________________________________________________________________
Objective Three _______________________________________________________________________________________________________________________________________________________________________________________________________________________
Objective Four (only if absolutely essential) ______________________________________________________________________________________________________________________________________________________________________________________________________________________
4.0 Rationale for the Study
The rationale for the study is where you summarise answers to the following questions:
a) What the study is about?
b) Why it is worth conducting?
c) What do you hope to learn from the study?
The rationale provides a link between the objectives you have set and the rest of the proposal that looks at the literature pertaining to the subject and your ideas of how to collect primary data.
5. Mini literature review
A key part of producing a proposal is satisfying those who will read it that you have done the up front work to validate the study. This means that you should be able to answer yes to the following questions:
a) Can you identify most of the areas of theory that will form part of the literature review?
b) Do you know who are the key writers in the area?
c) Do you know the extent to which literature exists on the subject you are interested in?
Being able to answer yes to these questions is a good start. The next stage is to construct a set of arguments that reinforce the objectives for the study and prove that the rationale you have provided is justified. In the course work you have a word limit of ****. This means that your mini literature review should be approximately **** words in length. This is sufficient for you to provide a detailed literature review that will cite (depending upon the subject) 20 – 30 sources. The principle reason for students failing the course work for this module is that they do not provide a literature review that support their objectives or proves that they have read around the subject. Unless you are aware of the arguments and issues of the research subject area you are not in a position to create a proposal that can be used to blue print your study (if in doubt about the role of literature in the research process review week 3’s workbook).
Try and provide answers to the following questions:
a) How long has the subject you are interested in been established?
b) Who are the key writers in the subject area?
c) Do you think there are any gaps in the existing theory that you can address with your own study? If so what are they?
d) What areas of management/leadership theory do you think you will need to include in your proposal? (spider topic two)
e) What are the key arguments that you intend to discuss in your literature review?
6.0 Sample Frame / Sampling Procedure
In producing a sample frame and outlining how you intend to select your sample you again are proving that you have thought through the underlying issues in operationalising your study. Certainly, your supervisor when he or she reads your proposal will be looking to see if your research can be undertaken within the timescales & resources available to you. Who you intend to access and how you will access them has a big bearing on whether your timesacle are realistic or not.
Within your general discussion of whom and how to access the sample you should state what your approach to theory development will be and why. The nature of research you will conduct e.g. exploratory, influences the type of sample required to answer the studies needs and the type of data collection tools that can be employed. Also by acknowledging the importance of theory development methodology you will provide a justification for the approach you intend to take.
The sample frame should be outlined with a full rationale of why you think the individuals or organisations are relevant to your study. You should also state how easy it will be to get them to participate in the research. If you are going to send out questionnaires how many will you need to get back to be able to conduct a specific type of analysis? Or if you intend to conduct interviews, will the person you want to interview grant you access? There is a need to provide some indication of a back up plan to cover the chance that your primary data collection method fails.
a) What is the nature of your research?
b) Who do you think you will need to target to answer your research questions?
c) What is the rationale for choosing these individuals to be your sample?
d) What contingency can you make if your target sample do not respond to your research tool?
7.0 Tools for Collecting Primary Data
A key stage in any study is committing yourself to how you will collect primary data. The decision should be influenced by the type of study you intend to undertake - positivist or phenomenological, and the nature of the study – exploratory, descriptive or explanatory.
The choice of how to collect data should be an informed one. You should think about the type of data needed to answer the specific questions generated from the objectives and the literature search. This often means a combination of quantitative and qualitative techniques. So for example you may conduct 3 interviews with product managers so that you can use the data derived from this exercise to create a questionnaire to send out to a much larger sample. Within your proposal it is important to make explicit your rationale for the choice of data collection tool, why it is the most suitable tool for the study, and what limitations exist on its use. Again, your supervisor will be looking to see that you have thought through all of the issues that relate to formulating a data collection approach and how you intend to make the study operational. This does not mean to say that your overall approach, the mixture and type of tools, can’t be modified as knowledge and experience of the study area increases.
The key questions to ask before you produce a choose a data collection tool are:
a) If I intend to replicate an existing study, can I use the same data collection tools and get a similar level of access to the target respondents?
b) If I intend to use specialised data collection tools do I have enough experience / expertise to make them work?
c) Do the timescales I have available for the study enable me to use the data collection tools in the way that I want to?
d) What type of data do I need to generate to satisfy the objectives of the study?
These questions should provide you with a guide as to which data collection tools are most appropriate to a study.
8.0 Analysis Technique
There is a link between the type of data collection tool you intend to use and how the data it generates can be analysed. For example, if you use in-depth interviews you will need to consider how to get usable data from the narrative of the conversation with the respondent. The simplest way to do this is to use some form of data reduction technique e.g. content analysis. For a few interviews this can be undertaken in a relatively short space of time. If you are going to interview more than ten people you may find that the time it takes you to manually reduce the data and then interpret is extensive. Therefore, you may wish to use an IT programme to help you. This type of argument is as relevant for quantitatively generated data as it is for qualitatively generated data. The relevance of this for your proposal is to make you aware of the following issues:
a) Do I have the knowledge base to analyse the data in a way that will generate the type of information needed to satisfy the objectives of the study?
b) How long will it take me to gain the knowledge and expertise in using specific software to analyse the data?
c) Is the software available to me to enable an analysis to be conducted using it?
d) How long will it take to analyse the data? Do I have that amount of time available to me?
In formulating answers to these questions you are really thinking about whether the data collection methods you want to use are appropriate within the constraints of the study. It is a sad truth that most of us start off seeing our research as a vehicle to test different methods and ideas of how to collect data, and end up having to make compromises because of time, access and availability of resources. For example, if you think that you are going to use a specific type of software to analyse your data, ask is it available to me? If I have to buy it how much will it cost? And, will I ever use it again? Will there be any support in the Business school to help me use it?
9.0 Timescale, resources requirement statement
The final part of the research proposal is to provide a break down of the timescales for completing each stage of the study. This is actually much harder to do than it may sound. You are given the start date and the finish date, but the rest of the timescales are only partially in your control. Certainly collecting primary data often takes much longer than most of us would like or even hope. You also have to aware that writing up your study into a thesis may be a much harder task than you had envisaged.
It is possible to think about creating a time plan based on rough estimates of how long each task will take. What is provided are approximates of the time it would take the average full-time researcher to complete each task.
a) Identifying the parameters of the study 4 weeks
b) Collecting and producing a literature review 4 weeks
c) Creating a sample frame 2 weeks
d) Creating and testing the data collection tool(s) 3 weeks
e) Collecting primary data 5 weeks
f) Analysing primary data 2 weeks
g) Writing up 4 weeks
Total 24 weeks
Although in producing a breakdown of the rough time it takes to complete each activity it suggests they are sequential, in reality they can be simultaneous activities. Indeed whilst you are collecting primary data you should also be thinking about the selection and design of the primary data collection activity. You should also be writing as you go along, so that the final write up is more of a polishing exercise then having to produce a whole thesis in one go.
9.1 Resource statement
It may be appropriate to put a resource statement into your proposal. This would be especially relevant if you need to conduct a study linked to an outside sponsor that involves a large sample for a postal questionnaire, or the purchase of a piece of software.
Producing a fully rationalised proposal is the first concrete stage within the research process. In essence it is the blue print for how you will conduct the study, and ultimately translate each component of the research into a document that can be examined.
The chances of you producing the perfect proposal in isolation are low. The actual process needs exposure to at least one member of the department if not a number. It is essential that you take time to discuss your ideas for each of the areas covered in this module and what you think will go into your proposal before finalising your coursework. If you spend the time developing a good proposal you will find that the actual task of undertaking your study will be made simpler. In addition the effort you put into the proposal now will save you time later when you should be completing the stages of the study rather than planning how to do them.
The final point is that the proposal should be considered to be the “best plan” that can be produced at this point in time with your current level of knowledge and experience. Research by its very nature is iterative - you learn from the process, as you learn your ideas are refined. Conducting research is journey of exploration, not every path that you take can be predicated at the outset. A well researched proposal should at the very least get you started on the correct path and make the journey a little easier than it might otherwise be.
11.0 Format for proposal
1) Contents page
3) Rationale for study
4) Literature review
5) Sample Frame & Procedure
6) Data collection tool(s)
7) Analysis framework
8) Timescale & resource statement
All pages should be numbered.