BUS114 GLOBAL | My Assignment Tutor

Session 10Research DesignandMethodologyMondays 9.30amonline synchronous classDr Alex [email protected] AND INTERNATIONAL MARKETSBUS114 GLOBAL Experiment= roots in natural science, laboratory-based research= due to the precision required to conduct it mean thatthe ‘experiment’ is often seen as the ‘gold standard’against which the rigour of other strategies is assessed.= a form of research that owes much to the naturalsciences, although it features strongly in psychological andsocial science research.= the purpose: to study the probability of a change in anindependent variable causing a change in another,dependent variable .= uses hypothetical explanations, known as hypotheses,rather than research questions. This is because theresearcher hypothesises whether or not a relationship willexist between the variables.Source: Saunders, Mark N. K., et al. Research Methods for BusinessStudents, Pearson Education Limited, 2019.ExperimentSource: Saunders, Mark N. K., et al. Research Methods for BusinessStudents, Pearson Education Limited, 2019.ExperimentTwo types of (opposing) hypotheses are formulated in astandard experiment:– the null hypothesis (the explanation that there is nodifference or relationship between the variables).An example of a null hypothesis is: User satisfaction of onlinecustomer support is not related to the amount of trainingsupport staff have received. The hypothesis is the explanationthat there is a difference or relationship between thevariables– the hypothesis (also referred to as the alternativehypothesis ).An example of a (directional) hypothesis is: User satisfactionof online customer support is related to the amount oftraining support staff have received.Source: Saunders, Mark N. K., et al. Research Methods for Business Students,Pearson Education Limited, 2019.ExperimentDifferent experimental designs may be used, eachwith different advantages and disadvantages,particularly in relation to control variables andconfounding variables.Experimental designs include:– classical experiments– quasi-experiments– within-subject designs.ExperimentIn a classical experiment, a sample of participants is selected and thenrandomly assigned to either an experimental group or to the control group.In the experimental group, some form of planned intervention ormanipulation will be tested.In the control group , no such intervention is made. Random assignmentmeans each group should be similar in all aspects relevant to the researchother than whether or not they are exposed to the planned intervention ormanipulation. In assigning the members to the control and experimentalgroups at random and using a control group, you try to control (that is,remove) the possible effects of an alternative explanation to the plannedintervention (manipulation) and eliminate threats to internal validity. This isbecause the control group is subject to exactly the same external influencesas the experimental group other than the planned intervention and,consequently, this intervention is the only explanation for any changes tothe dependent variable.Source: Saunders, Mark N. K., et al. Research Methods for Business Students,Pearson Education Limited, 2019. Quasi Experiment= still uses an experimental group(s) and a control group, but the researcherwill not randomly assign participants to each group, perhaps becauseparticipants are only available in pre-formed groups (e.g. existing workgroups).= differences in participants between groups may be minimised by the useof matched pairs. Matched pair analysis leads to a participant in anexperimental group being paired with a participant in the control groupbased on matching factors such as age, gender, occupation, length ofservice, grade etc., to try to minimise the effect of extraneous variables onthe experiment’s outcomes. Those factors relevant to the nature of theexperiment will need to be matched.Source: Saunders, Mark N. K., et al. Research Methods for Business Students,Pearson Education Limited, 2019.Case Study= an in-depth inquiry into a topic or phenomenon within its real-life setting (Yin 2018).The ‘case’ in case study research may refer to:– a person (e.g. a manager),– a group (e.g. a work team),– an organisation (e.g. a business),– an association (e.g. a joint venture),– a change process (e.g. restructuring a company),– an event (e.g. an annual general meeting)– as well as many other types of case subject.Source: Saunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019.Case StudyChoosing the case to be studied and determining the boundaries of the study is a key factor indefining a case study. Once defined, case study research sets out to understand the dynamics ofthe topic being studied within its setting or context= has the capacity to generate insights from intensive and in-depth research into the study of aphenomenon in its real-life context, leading to rich, empirical descriptions and the developmentof theory= ‘the interaction between a phenomenon and its context is best understood through in-depthcase studies’. These can be designed to identify what is happening and why, and perhaps tounderstand the effects of the situation and implications for action. To achieve such insights, casestudy research draws on data, often both qualitative and quantitative, from a range of sourcesto understand fully the dynamics of the case.Source: Saunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019.Case Study= criticised by some as a research strategy because of ‘misunderstandings’ about their abilityto produce generalisable, reliable and theoretical contributions to knowledge. This is largelybased on positivist criticisms of using small samples and more generally about usinginterpretive, qualitative research.The long and widespread use of case studies has resulted in them being designed in differentways and for different purposes.They have been used– by ‘positivist’ as well as ‘interpretivist’ researchers;– deductively as well as inductively; and– for descriptive, exploratory or explanatory purposes.= can also incorporate multiple cases, that is, more than one case. The rationale for usingmultiple cases focuses on whether findings can be replicated across cases. Cases will becarefully chosen on the basis that similar results are predicted to be produced from each one. Ethnography= used to study the culture or social world of a group.= literally means a written account of a people or ethnic group.= it is the earliest qualitative research strategy, with its origins in colonial anthropology.From the 1700s to the early 1900s, ethnography was developed to study cultures in so-called‘primitive’ societies that had been brought under the rule of a colonial power, to facilitateimperialist control and administration. Early anthropologists treated those among whom theylived and conducted their fieldwork as subjects and approached their ethnography in a detachedway, believing that they were using a scientific approach, reminiscent of a positivism, to producemonographs that were meant to be accurate and timeless accounts of different cultures.= involved researchers living among those whom they studied, to observe and talk to them inorder to produce detailed cultural accounts of their shared beliefs, behaviours, interactions,language, rituals and the events that shaped their lives= study people in groups, who interact with one another and share the same space, whether this isat street level, within a work group, in an organisation or within a society. Ethnography= is relevant for modern organisations. For example, in market research ethnography is auseful technique when companies wish to gain an in-depth understanding of their marketsand the experiences of their consumers= relevant to marketing= relevant in other business and management subject areas.Saunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019. Action Research= emergent and iterative process of inquiry that is designed to develop solutions to realorganisational problems through a participative and collaborative approach, which usesdifferent forms of knowledge, and which will have implications for participants and theorganisation beyond the research project= five themes: purpose, process, participation, knowledge and implications.= purpose of the strategy = to promote organisational learning to produce practicaloutcomes through identifying issues, planning action, taking action and evaluating action.Saunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019.Action Research= involves participation and cooperation:1) firstly, organisational members need to cooperate with the researcher to allow theirexisting work practices to be studied.2) Then it requires participation in the form of collaboration through its iterative cycles tofacilitate the improvement of organisational practices.Collaboration means building a democratic approach to communication and decision makingwhen constructing, planning, taking and evaluating each Action Research stage or cycle.The researcher passes on her or his skills and capabilities to participants so that theyeffectively become co-researchers in the Action Research process. Without such participation,this approach simply would not be viable, although creating such participation is likely to bedifficult in practice and to meet with resistance at various levelsSaunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019. Grounded Theory= can be used to refer to a methodology, a method of inquiry and the result of a research process= ‘Grounded theory methodology’ refers to the researcher’s choice of this strategy as a way to conduct research= was developed by Glaser and Strauss (1967) as a response to the ‘extreme positivism’ of much social research atthat time. They disputed the view that social research should use a paradigm based on a premise that theory willreveal a pre-existing reality.= is used to develop theoretical explanations of social interactions and processes in a wide range of contexts,including business and management. As many aspects of business and management are about people’s behaviours,for example consumers’ or employees’, a Grounded Theory strategy can be used to explore a wide range ofbusiness and management issues.= the aim is to ‘discover’ or generate theory grounded in the data produced from the accounts of social actors. Thisinductive, theory- building approach of Grounded Theory illustrates an important difference from the theorytesting approach associated with much previous social research, where hypotheses wereSaunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019. Narrative Inquiry= a story;= a personal account which interprets an event or sequence of events;= as a research strategy, it has a more specific meaning and purpose; there will be research contexts where theresearcher believes that the experiences of her or his participants can best be accessed by collecting and analysingthese as complete stories, rather than collecting them as bits of data that flow from specific interview questions andwhich are then fragmented during data analysis.= seeks to preserve chronological connections and the sequencing of events as told by the narrator (participant) toenrich understanding and aid analysis;= as providing the opportunity to connect events, actions and their consequences over time into a ‘meaningful whole’.Through storytelling the narrator will also provide his or her interpretation of these events, allowing the narrativeresearcher to analyse the meanings which the narrator places on events. Where there is more than one participantproviding a personal account of a given context, the narrative researcher will also be able to compare and totriangulate or contrast these narratives. The narrative provided may be a short story about a specific event; a moreextended story (for example, about a work project, managing or setting up a business, or an organisational changeprogramme); or a complete life history= Other sources of narratives include autobiographies, authored biographies, diaries, documentation and informaldiscussions.Saunders, Mark N. K., et al. Research Methods for Business Students, Pearson Education Limited, 2019.6) Sampling: design and proceduresThere is no hope of making scientificstatements about a population based on theknowledge obtained from a sample, unlesswe are circumspect in choosing a samplingmethod.Naresh Malhotra and David Birks, Marketing Research, 3rd Edition, © Pearson Education Limited 2007Slide 14.26THE MAIN STAGES OF A RESEARCH PROCESS1. Formulate and clarify the research topic2. Critical literature review3. Choose the research strategy (methodology)4. Access and ethical issues6. Data analysis (qualitative and quantitative) 7. Writing up / conclusions / recommendations 5. Data collection through:secondary data / sampling / observation / interviews / questionnairesPopulation – The aggregate of all the elements,sharing some common set of characteristics, thatcomprise the universe for the purpose of themarketing research problem.Census – A complete enumeration of the elementsof a population or study objects.Sample – A subgroup of the elements of thepopulation selected for participation in the study.Sample or censusPopulation and Sample RelationshipsPOPULATION SAMPLEParameters measuresdescribing populationcharacteristicsSample is part ofthe populationStatistics measuresdescribing samplecharacteristicsstatisticsestimate parametersPopulation and SampleXXXXXXXXXXXXXXXXXXXPopulationSample ASample BSampleelementsPopulationelementsWhy Sample???• The concept of sampling involves taking a PORTION of thepopulation, making observations on this smaller group andthen GENERALISING the findings to the large population• Generalisation is a necessary scientific procedure, since rarelyis it possible to study all the members of a defined populationTHE SAMPLING PROCESS• Define population• Choose data collection method• Choose sampling frame• Select sampling method• Determine sample size• Develop/execute operational planBased on McDaniel & Gates Contemporary Marketing ResearchDefine the target populationThe target population is the collection of elements or objectsthat possess the information sought by the researcher andabout which inferences are to be made. The target populationshould be defined in terms of elements, sampling units, extentand time.• An element is the object about which or from which theinformation is desired, for example, the respondent.• A sampling unit is an element, or a unit containing theelement, that is available for selection at some stage of thesampling process.• Extent refers to the geographical boundaries.• Time is the time period under consideration.SAMPLING FRAME• Objective list of ‘the population’ from whichsampling selections can be made.• Issues to consider:• Relevant?• Complete?• Precise?• Up-to-date?• Access for research?TYPES OF SAMPLINGPROBABILITY orRANDOM• Simple Random• Systematic• Stratified• ClusterNON-PROBABILITY orNON-RANDOM• Convenience• Quota• Judgement• SnowballSampling techniques Probability samplingtechniquesNon-probability samplingtechniquesJudgmentalsamplingSnowballsamplingQuotamplings ConveniencesamplingSimplerandomsamplingSystematicsamplingStratifiedsamplingClustersampling RANDOM☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺SYSTEMATIC☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺STRATIFIED☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺CLUSTER☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺CONVENIENCE☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺QUOTA ☺  ☺ ☺  ☺ ☺  ☺☺  ☺  ☺ ☺ ☺  ☺ ☺☺ ☺  ☺ ☺  ☺ ☺  ☺JUDGEMENT☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺SNOWBALL☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺ ☺SAMPLE SIZE• Budget• Gut feeling• Size of sub-groups• Homogeneity of population• Statistical biasProbability Samples• The objective selection of elements allows the objectiveassessment of the reliability of the sample results, somethingnot possible with non-probability samples…• This is not to say that probability samples will always be morerepresentative than non-probability samples.• Nevertheless, probability samples can assess “sampling error”that is likely to occur because a sample rather than a census wasemployed when gathering data…Probability Sampling (Random Sampling)• Simple Random Sampling – the ‘lottery method’ [remember thisfrom Intro to Statistics?] …population elements have an equalchance of being selected!• Systematic Random Sampling – sample obtained by taking every‘nth’ subject or case from a list containing the total population• Stratified Random Sampling – a random sample drawn fromparticular categories of the population being studied. Individualswithin the categories should be highly similar to one another anddifferent from individuals in other strata.Simple random sampling• Each element in the population has a known and equalprobability of selection.• Each possible sample of a given size (n) has a known andequal probability of being the sample actually selected.• This implies that every element is selected independently ofevery other element.Systematic sampling• The sample is chosen by selecting a random starting point andthen picking every ith element in succession from thesampling frame.• The sampling interval, i, is determined by dividing thepopulation size N by the sample size n and rounding to thenearest integer.• When the ordering of the elements is related to thecharacteristic of interest, systematic sampling increases therepresentativeness of the sample.Systematic sampling (Continued)• If the ordering of the elements produces a cyclical pattern,systematic sampling may decrease the representativeness ofthe sample.For example, there are 100,000 elements in the populationand a sample of 1,000 is desired. In this case the samplinginterval, i, is 100. A random number between 1 and 100 isselected. If, for example, this number is 23, the sampleconsists of elements 23, 123, 223, 323, 423, 523 and so on.Tennis’s systematic samplingreturns a smashTennis magazine conducted a mail survey of its subscribers to gain abetter understanding of its market. Systematic sampling wasemployed to select a sample of 1,472 subscribers from thepublication’s domestic circulation list. If we assume that thesubscriber list had 1,472,000 names, the sampling interval would be1,000 (1,472,000/1,472). A number from 1 to 1,000 was drawn atrandom. Beginning with that number, every 1,000th subscriber wasselected.A brand-new dollar bill was included with the questionnaire as anincentive to respondents. An alert post card was mailed one weekbefore the survey. A second, follow-up, questionnaire was sent to thewhole sample ten days after the initial questionnaire. There were 76post office returns, so the net effective mailing was 1,396. Six weeksafter the first mailing, 778 completed questionnaires were returned,yielding a response rate of 56%.Stratified sampling• A two-step process in which the population is partitioned intosubpopulations, or strata.• The strata should be mutually exclusive and collectivelyexhaustive in that every population element should beassigned to one and only one stratum and no populationelements should be omitted.• Next, elements are selected from each stratum by a randomprocedure, usually SRS.• A major objective of stratified sampling is to increase precisionwithout increasing cost.Stratified sampling (Continued)• The elements within a stratum should be as homogeneous aspossible, but the elements in different strata should be asheterogeneous as possible.• The stratification variables should also be closely related tothe characteristic of interest.• Finally, the variables should decrease the cost of thestratification process by being easy to measure and apply.• In proportionate stratified sampling, the size of the sampledrawn from each stratum is proportionate to the relativesize of that stratum in the total population.• In disproportionate stratified sampling, the size of thesample from each stratum is proportionate to the relativesize of that stratum and to the standard deviation of thedistribution of the characteristic of interest among all theelements in that stratum.Stratified sampling (Continued)Cluster sampling• The target population is first divided into mutuallyexclusive and collectively exhaustive subpopulations,or clusters.• Then a random sample of clusters is selected, basedon a probability sampling technique such as SRS.• For each selected cluster, either all the elements areincluded in the sample (one-stage) or a sample ofelements is drawn probabilistically (two-stage).Cluster sampling (Continued)• Elements within a cluster should be as heterogeneous aspossible, but clusters themselves should be ashomogeneous as possible. Ideally, each cluster should be asmall-scale representation of the population.• In probability proportionate to size sampling, the clustersare sampled with probability proportional to size. In thesecond stage, the probability of selecting a sampling unit ina selected cluster varies inversely with the size of thecluster.A graphical illustration ofcluster sampling (2-stage) ABCDE16111621271217223813182349141924510152025 Randomly select 3 clusters, B, Dand E.Within each cluster, randomlyselect oneor two elements. The resultingsampleconsists of population elements7, 18, 20, 21 and 23. Note, noelements are selected fromclusters A and C.Non-Probability Sampling• …involves personal judgement somewhere in theselection process.• The fact the elements are not selectedprobabilistically precludes an assessment of“sampling error”.• Without some knowledge of the error that canbe attributed to sampling procedures, we cannotplace bounds on the precision of our estimates!Non-Probability Sampling(individuals in the population DO NOT have an equal chance of selection)• Convenience Sampling = attempts to obtain a sample of convenientelements. The selection of sampling units is left primarily to the interviewer.• Judgement Sampling = a form of convenience sampling in which thepopulation elements are purposely selected based on the judgement of theresearcher.• Snowball Sampling = an initial group of respondents is selected randomly.Subsequent respondents are selected based on the referrals or informationprovided by the initial respondents.• Quota Sampling = a 2-stage restricted judgmental sampling technique. The1st stage consists of developing control categories. In the 2nd stage sampleelements are selected based on convenience or judgement.Convenience samplingConvenience sampling attempts to obtain a sampleof convenient elements. Often, respondents areselected because they happen to be in the rightplace at the right time.• use of students, and members of social organisations• street interviews without qualifying the respondents• ‘people on the street’ interviewsJudgmental samplingJudgmental sampling is a form of conveniencesampling in which the population elements areselected based on the judgment of the researcher.• test markets• purchase engineers selected in industrial marketingresearch• expert witnesses used in courtSnowball samplingIn snowball sampling, an initial group of respondents isselected, usually at random.• After being interviewed, these respondents are asked toidentify others who belong to the target population ofinterest.• Subsequent respondents are selected based on thereferrals.Quota samplingQuota sampling may be viewed as two-stage restricted judgmental sampling.• The first stage consists of developing control categories, or quotas, ofpopulation elements.• In the second stage, sample elements are selected based on convenienceor judgment. PopulationcompositionSamplecompositionControlCharacteristicPercentagePercentageNumberSexMale4848480Female52____10052____100520____1000 A classification of Internet samplingInternet samplingOnline interceptsamplingRecruited onlinesamplingOther techniquesNon-random Random Panel Non-panelRecruitedpanelsOpt-inpanelsOpt-in listrentals


Leave a Reply

Your email address will not be published. Required fields are marked *