Annotated bibliography
Limwh
bbs200-sessions5-6-jr-for-circulation.pptxBBS200 Sessions 5-6
Dr Juergen Rudolph
June 2018
1
Agenda
Recap of workshops 1-2 (MCQ)
Chapter 6 – sampling (Q&A and Guinness case study)
Chapter 9 – data collection (Data mining case study)
Annotated Bibliography (AB) assignment discussion
Referencing test and recap
Overall online test 1 performance at Kaplan Singapore (mean: 7.13 / 10)
Some MCQs to recap the first two workshops (1)
A useful way to approach the research process is to state the basic dilemma that prompts the research and then…
A. Start finding solutions
B. Develop questions
C. Breaking down the original question into other more specific questions using the management research question hierarchy
D. Breaking down the original question into more specific questions
Some MCQs to recap the first two sessions (1)
A useful way to approach the research process is to state the basic dilemma that prompts the research and then…
A. Start finding solutions
B. Develop questions
C. Breaking down the original question into other more specific questions using the management research question hierarchy
D. Breaking down the original question into more specific questions
ANSWER: C
MCQ (2)
Is research always problem-solving based?
A. No, only applied research is problem solving based
B. Yes research is always problem solving based
C. No, as sometimes research is designed to investigate theory only
D. No, only basic or applied research is problem solving based
MCQ (2)
Is research always problem-solving based?
A. No, only applied research is problem solving based
B. Yes research is always problem solving based
C. No, as sometimes research is designed to investigate theory only
D. No, only basic or applied research is problem solving based
ANSWER: B
MCQ (3)
Good research has the following characteristics
A. Follows the scientific method, has a clearly defined purpose and has high ethical standards applied
B. Has a clearly defined purpose, has high ethical standards applied and sophisticated statistical analysis
C. Justifies conclusions, reveals the limitations, and presents findings ambiguously
D. Follows the scientific method, has a clearly defined purpose and is only published in academic journals
MCQ (3)
Good research has the following characteristics
A. Follows the scientific method, has a clearly defined purpose and has high ethical standards applied
B. Has a clearly defined purpose, has high ethical standards applied and sophisticated statistical analysis
C. Justifies conclusions, reveals the limitations, and presents findings ambiguously
D. Follows the scientific method, has a clearly defined purpose and is only published in academic journals
ANSWER: A
MCQ (4)
Research philosophies include:
A. Realism, Positionism, Intrepretivism
B. Realism, Positivism and Interpretivism
C. Realism, Qualitative and Quantitative
D. Interpretivism, Quantitative and Qualitative
MCQ (4)
Research philosophies include:
A. Realism, Positionism, Intrepretivism
B. Realism, Positivism and Interpretivism
C. Realism, Qualitative and Quantitative
D. Interpretivism, Quantitative and Qualitative
ANSWER: B
MCQ (5)
The aim of a literature review is to
Show the reader what has been done previously
Synthesize and gain a new perspective on a problem
C. Relate theories and ideas to a problem
D. All of the above
MCQ (5)
The aim of a literature review is to
Show the reader what has been done previously
Synthesize and gain a new perspective on a problem
C. Relate theories and ideas to a problem
D. All of the above
ANSWER: D
MCQ (6)
A systematic review includes:
A. Planning the proposal, identifying articles, reading and writing up
B. Planning, conducting review, reporting and dissemination
C. Formulate the problem, find the solution, write up the report
D. Formulate the problem, identifying the articles and reporting
MCQ (6)
A systematic review includes:
A. Planning the proposal, identifying articles, reading and writing up
B. Planning, conducting review, reporting and dissemination
C. Formulate the problem, find the solution, write up the report
D. Formulate the problem, identifying the articles and reporting
ANSWER: B
MCQ (7)
The focus of a literature review can be:
A. Integrative
B. theoretical
C. methodological
D. all of the above
MCQ (7)
The focus of a literature review can be:
A. Integrative
B. theoretical
C. methodological
D. all of the above
ANSWER: D
MCQ (8)
The four criteria used in critical reviews are:
A. Contribution to the field; agreement; methods and analysis; writing
B. Contribution to the field; argumentation; methods and analysis; writing
C. Literature; argumentation; methods and analysis; writing
D. Literature; agreement; methods and analysis; writing
MCQ (8)
The four criteria used in critical reviews are:
A. Contribution to the field; agreement; methods and analysis; writing
B. Contribution to the field; argumentation; methods and analysis; writing
C. Literature; argumentation; methods and analysis; writing
D. Literature; agreement; methods and analysis; writing
ANSWER: B
MCQ (9)
Criteria to assess relevance and value of literature
A. Prominence of article/book/chapter documented by citations or the source
B. Recency of the article or (book) chapter and methodological quality of the article or (book) chapter
C. A and B are both correct
D. None of the above
MCQ (9)
Criteria to assess relevance and value of literature
A. Prominence of article/book/chapter documented by citations or the source
B. Recency of the article or (book) chapter and methodological quality of the article or (book) chapter
C. A and B are both correct
D. None of the above
ANSWER: C
MCQ (10)
What are ethics?
A. Moral principles, norms or standards of behaviour that guide moral choices about our behaviour and our relationships with others.
B. Rules of behaviour based on ideas about what is morally good and bad
C. The set of moral principles that guide a person's behaviour
D. All of the above
MCQ (10)
What are ethics?
A. Moral principles, norms or standards of behaviour that guide moral choices about our behaviour and our relationships with others.
B. Rules of behaviour based on ideas about what is morally good and bad
C. The set of moral principles that guide a person's behaviour
D. All of the above
ANSWER: D
Sampling Introduction to Data Sampling
BBS200
Understanding Business Research: An introductory Approach
24
Questions on sampling
(1) What is a population?
(2) What is a sample?
(3) What is data sampling?
(4) Why do researchers use a sample?
(5) Describe the factors that should be considered to ensure a ‘good’ sample?
(6) What is the difference between a Probability sample and a non-probability sample?
What is a population?
According to the Australian Bureau of Statistics a population is any complete group with at least one characteristic in common.
That is all the employees who work for Myer, all the accountants in Australia, all the surf-lifesavers in Western Australia, or the farmers in the South West, all the economists in the public service…
Populations are not just people. Populations may consist of, but are not limited to, businesses, buildings, computers, farms, objects or events.
A population may be studied using one of two approaches: taking a census, or selecting a sample.
What is a sample?
A sample is a subset of units in a population, selected to represent all units in a population of interest. Information from the sample is used to estimate the characteristics for the entire population of interest.
What is data sampling?
Sampling is selecting a portion of the population, in your research area, which will be a representation of the whole population.
Why do researchers use a sample?
Save time and money…very costly to use the whole population, and very difficult.
Describe the factors that should be considered to ensure a ‘good’ sample?
The ultimate test of a sample design is how well it represents the characteristics of the population you are studying.
In measurement terms, the sample must be valid. Validity of a sample depends on two considerations: accuracy and precision.
What is the difference between a Probability sample and a non-probability sample?
Probability (or random) and non-probability (or non-random) sampling.
Flowchart: Business Research Process
Note: Diamond-shaped boxes indicate stages in the research process in which a choice of one or more techniques must be made. The dotted line indicates an alternative path that skips exploratory research.
Zikmund, W.G., Babin, B.J., Carr, J.C., & Griffin, M. (2013). Business Research Methods. 9th ed. Mason, OH: South Western Cengage Learning.
2
1
3
1
3
4
5
6
32
Sampling Terminology
Census
Population
Sample
sampling
33
Nature of Sampling
1. Relevant
2. Parameters of interest
3. Sampling Frame
4. Type of Sample
5. Sample size
6. Cost
34
Unit of Analysis
Unit of analysis depends on:
What is the research problem, i.e. on what level do you look for answers?
At what level do we need information, what do we measure?
At what level do we want to implement the answers found?
35
Countries
Sectors
Firms
Departments
Teams
Employees
Decisions
Why sample?
Lower costs (budget)
Greater speed (time)
Availability of sample elements
Greater accuracy trade-off between
asking everybody
versus
obtaining better and more data from a representative subgroup
36
What is a Good Sample?
Accuracy
and
Precision
37
Probability
Non-probability
Convenience
Purposive
Sampling Overview
Simple
random
Systematic
Cluster
Stratified
random
Quota
Snowball
38
Probability
Non-probability
Convenience
Purposive
Sampling Overview
Simple
random
Systematic
Cluster
Stratified
random
Quota
Snowball
39
Simple Random Sampling (SRS)
1
2
3
4
5
6
8
7
10
9
11
12
13
14
15
16
18
17
20
19
21
22
23
24
25
26
28
27
30
29
40
Simple random sampling
Each population element has an equal chance of being selected into the sample. Sample drawn using random number table/generator.
You could do it like a lotto draw. That is put all the numbers 1 through to 30 in a hat and then draw numbers randomly out of the hat until you have your sample of 30.
Or you could use technology that randomly choose 10 participants from the numbers 1-30.
Advantages:
Easy to implement if the population is in a database
Disadvantages:
Requires a listing of population elements. Takes more time to implement. Uses larger sample sizes. Produces larger errors. Expensive.
Stratified sampling
Strata
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
42
Stratified sampling
In stratified sampling you want an equal number of people from each group.
(1) So the first thing you have to do is to group a population into their separate groups. Girls/boys; postcode, suburb, job type, level in the organisational hierarchy.
Each one of these groups is called a stratum, for example: the accountants stratum, the finance employees stratum, HR employees stratum. I have stratified them.
(2) I then want four from each stratum. Here I would use simple random sampling again.
I would then pull numbers from a hat or use excel to randomly generate numbers.
Advantages:
Researcher controls sample size in strata. Increased statistical efficiency. Provides data to represent and analyse sub-groups. Enables use of different methods in strata.
Disadvantages:
Increased error will result if sub-groups are selected at different rates. Expensive. Especially expensive if strata of the population have to be created.
Systematic sampling
Start at 3
Interview every 3rd person
So I still have a population of 30…Again I want a sample of 10. This time I am going to use systematic sampling to draw my sample probability sample. I am going to use a system to get my sample of 10 characters.
To start with I am going to choose the third character in my sample. You could think of the rows of characters here as businesses on St. Georges Terrace in the city. As such I would be starting with the No 3 St. Georges Terrace and then choosing every 3rd building.
so systematic sampling uses a random start and then selects a sampling fraction that is every kth element. In my example every 3rd characeter. We determine the what the K will be by dividing the population size by the desired sample size.
Advantages:
Simple to design. Easier to use than the simple random. Easy to determine sampling distribution of mean or proportion. Less expensive than simple random.
Disadvantages:
Periodicity within the population may skew the sample and results. If the population list has a monotonic trend, a biased estimate will result based on the start point.
44
Systematic sampling
To start with, I am going to choose the third person in my sample. So systematic sampling uses a random start and then selects a sampling fraction that is every kth element. In my example every 3rd character. We determine what K will be by dividing the population size by the desired sample size.
Advantages:
Simple to design. Easier to use than the simple random. Easy to determine sampling distribution of mean or proportion. Less expensive than simple random.
Disadvantages:
Periodicity within the population may skew the sample and results. If the population list has a monotonic trend, a biased estimate will result based on the start point.
Cluster sampling
46
Cluster
Cluster sampling should be used only when it is economically justified - when reduced costs can be used to overcome losses in precision. This could be when constructing a complete list of population elements is difficult, costly, or impossible.
For example, it may not be possible to list all of the customers of a chain of pizza shops such as Dominos. However, it would be possible to randomly select a subset of stores (stage 1 of cluster sampling) and then interview a random sample of customers who visit those stores (stage 2 of cluster sampling).
Cluster sampling is classified as a probability sampling technique because of either the random selection of clusters or the random selection of elements within each cluster.
So in essence in cluster sampling the population is divided into internally heterogeneous (that is mixed) sub-groups and then some are randomly selected for further study.
Advantages:
Provides an unbiased estimate of population parameters if properly done.
Disadvantages:
Often lower statistical efficiency (more error) due to sub-groups being homogeneous rather than heterogeneous.
Probability
Non-probability
Convenience
Purposive
Sampling Overview
Simple
random
Systematic
Cluster
Stratified
random
Quota
Snowball
48
Quota Sampling
Various sub groups represented on the important characteristic
Examples of characteristics that might be used
Female and Male leaders
Full-time and part-time students
Stay at home and working mothers
Stratified sampling, a probability sampling procedure also has this objective and is often confused with quota sampling. In quota sampling, the researchers has a quota to achieve. For example, the interviewer may be assigned 100 interviews with leaders 85 with male leaders and 15 with female leaders. The interviewer is responsible for finding enough people to meet the quota.
49
Quota
Quota sampling is a non-probability sampling procedure that ensures that various subgroups of a population will be represented on the important characteristics that the researcher is interested in.
Stratified sampling, a probability sampling procedure also has this objective and is often confused with quota sampling.
In quota sampling, the researcher has a quota to achieve. For example, the interviewer may be assigned 100 interviews with leaders 85 with male leaders and 15 with female leaders. The interviewer is responsible for finding enough people to meet the quota.
Purposive Sampling
Participants chosen on the basis of judgment:
Typical case
Critical case
Heterogeneous case
Theoretical case
51
Purposive
Purposive sampling is where participants are chosen on the basis of judgment.
Purposive samples are the most frequently used form of non-probability sampling in qualitative research (Miles and Huberman, 1994). These techniques require our judgment in choosing cases that will best enable us to answer our research h question and meet our aim. They are normally used to choose relatively small number of participants, such as those that are particularly informative.
This might be:
typical case, to illustrate the profile of a typical consumer of a product;
critical case, to highlight important aspects or make a dramatic point;
heterogeneous cases – to reveal key themes about diverse characteristics to provide maximum variation possible in the data collected
theoretical case – to inform emerging theory, data collection based on concepts that appear to be relevant to an evolving theory and derived from the data.
Snowball Sampling
53
Snowball
Just as a snowball rolling down a mountain gathers snow and gets larger as it goes, snowball sampling is the same.
Snowball sampling is an approach for locating information-rich key informants. Using this approach, a few potential respondents are contacted and asked whether they know of anybody with the characteristics that you are looking for in your research.
In snowball sampling, participants are volunteered. This is commonly used when it may be difficult to identify members of the desired population. Such as those people who work for cash, but claim unemployment benefits. Such as: I interview Mary who is working for cash, but claiming unemployment benefits she informs me about Hussein who is also working for cash and claiming unemployment benefits and so on….
Snowball sampling may be the only possibility for finding participants.
Convenience Sampling
Sampling by obtaining people (or units) that are convenient.
Students in a class – a captive sample
People on the street
Shoppers in a shopping centre
55
Convenience
Convenience sampling is a sampling procedure of obtaining those people or units which are most conveniently available. A researcher may decide that the most convenient and economical method is to set up an interviewing booth at a shopping centre to intercept shoppers.
Just as news reporters do during elections – they might set themselves up on the main street in the city and ask people on the street their views on …who they might vote for, or what they think of a new government policy, or a decision handed down by the court.
This type of sample is convenient but perhaps not so representative.
Guinness is good for you, or is that only in Ireland?
https://www.youtube.com/watch?v=wSwSmls9ACI
Questions
(1) What were the strengths and weaknesses of the approach adopted by this study’s researchers?
(2) What other possible interpretations could be drawn from the findings of this research?
(3) In what other ways could researchers discover how good the taste of Guinness is in Ireland, compared to the rest of the world?
(4) What other food and drink could you test to determine whether they are better in a particular place, rather than others?
(1) What were the strengths and weaknesses of the approach adopted by this study’s researchers?
Strengths
It utilised an international team of researchers, which is more objective than a group drawn from one country.
It developed a systematic way of measuring the quality of the Guinness, the technical expertise of the pint pourer and of the ambience of the pub in which the pint was poured. There were 26 factors in all being measured.
It had a large amount of tastings, 103, in 14 different countries.
It involved tasting the product directly rather than talking about the product being tasted.
(1) ctd.
Weaknesses
The team of researchers could have been bigger and so they might have had a greater range of views on their measurement factors.
The research team could also have included women, who may view Guinness differently from men.
Only two continents were included in the sample, Europe and North America. A future study might also include Africa and Asia as these are also large markets for Guinness.
Mainstream draught Guinness was used in this present research. Future studies might also taste bottled Guinness to see if the effect found in the study is more for the place in which, and the person with whom, the Guinness is served than for the Guinness itself.
(2) What other possible interpretations could be drawn from the findings of this research?
Irish pubs: what the researchers may also have found was that they enjoyed Irish pub ambience more than foreign pub ambience.
Guinness pub specialisation: Since Guinness is so popular in Ireland they may be better at tending it and pouring it and so if other countries took as much care and attention their scores for Guinness may have been better.
Expectancy effect: The researchers may have expected to enjoy their Guinness more in Ireland and so this may have influenced their perception of the product in Ireland. It may thus have been better to employ researchers who were unaware of the theory being tested.
(3) In what other ways could researchers discover how good the taste of Guinness is in Ireland compared to the rest of the world?
A survey: ask those who have drunk Guinness to rate it and to say what they thought of it in the different places they drank it.
A focus group: give focus groups in pubs around the world samples of Guinness and see how they rate it. This could establish average levels for satisfaction for the product before you bring those focus groups to Ireland, and vice versa, to see if they change their rating once they sample Guinness in Ireland (if that isn’t their home country).
A tasting panel: get a panel of expert testers to sample Guinness in different locations around the world. They could even test whether Guinness is better in a particular part of Ireland, perhaps Dublin as that’s where it’s brewed in Ireland.
Bottles: use bottled Guinness instead of draught Guinness so the Guinness is exactly the same, which may not be the case in different parts of the world as it is brewed in a number of different locations.
What other food and drink could you test for whether they are better in a particular place rather than others?
French wine: do the French keep the best wine for themselves and export the rest? The same with the Italians and the Spanish? Or is it that the climate and the “je ne sais quoi” of drinking French wine in France make it better? The same may be true of Cognac or Armagnac brandies. Another worthwhile research study.
Dutch Edam: one of the great world cheeses. Is it as good outside the Netherlands or does it have to be eaten beside a canal in Amsterdam, having a pint of Amstel or Heineken as you do so?
Smorgasbord: the famous Scandinavian offering of a buffet including hot and cold meats, salads and hors d'oeuvres may be much more enjoyable when in Oslo, Helsinki, Stockholm or Copenhagen than when sampled anywhere else.
Teh halia, roti prata, Hainanese chicken rice, chili crab…
Data Collection Overview
BBS200
Understanding Business Research: An introductory Approach
64
What is data?
Facts
Observations
Published information
65
Data Collection
Primary
Data
Secondary
Data
Interviews*
Questionnaires*
Experiments
Focus groups
Case Studies
Observations
Academic Journals
Business Info. Systems
Government Reports
Aust. Bureau Statistics (ABS)
Organisational Reports
66
Data Collection
Primary
Data
Secondary
Data
Research Tools
Interviews*
Questionnaires*
Focus groups
Observations
Case Studies (Topic 9)
Experiments (Topic 9)
Academic Journals
Business Info. Systems
Government Reports
Aust. Bureau Statistics (ABS)
67
http://www.slideshare.net/lucypark/introduction-to-data-mining-for-newbies/5
68
Advantages: Secondary data
Saves time & money
High quality and
easily accessible
Analysis can start
69
Disadvantages: secondary data
70
Video on Data Mining https://www.youtube.com/watch?v=f2Kji24833Y
Data mining case study – intro (1)
This case concerns the use of data mining and microtargeting techniques in the 2004 American presidential election. Karl Rove, President George W Bush’s campaign manager, used these techniques to identify voters in 18 key swing states who had voted for the Republican Party previously and who would be likely to do so again.
Rove bought consumer data, from the likes of credit card companies, and put together profiles of the type of consumers that, based on their consumer habits, were more likely to vote Republican in the Presidential election.
Data mining case study – intro (2)
Rove and Bush were successful in the 2004 election, by the narrowest margin ever for any incumbent President. Bush couldn’t run in 2008 as he had served two terms and so was barred for standing for presidential election.
The Republicans lost that election to Barrack Obama, whose campaign was the first to make widespread use of social networking communications and attempted to attract disenfranchised and new voter constituencies. Nevertheless, the use of data mining is alive and well, as can be seen from the existence of the Republican Voter Vault voter database and the Democrat’s Catalist counterpart.
Data mining case study: questions
(1) What advantages and disadvantages were there in Karl Rove’s use of data mining to identify Republican voters in the 2004 US Presidential election?
(2) What other techniques could Karl Rove have used to identify Republican voters?
(3) In which other areas could data mining be used?
(4) What are the ethical dilemmas involved in using data mining?
(1) What advantages and disadvantages were there in Karl Rove’s use of data mining to identify Republican voters in the 2004 US Presidential election? (1)
Advantages:
Once a large, representative and clean database is acquired then most of the analysis can be completed by a small team who are well versed in certain statistical techniques.
It can be very precise in targeting voters.
It may be quicker and easier than traditional research techniques, such as surveys and focus groups
It takes advantage of the increasingly available commercial databases from large retailers and credit card companies
(1) What advantages and disadvantages were there in Karl Rove’s use of data mining to identify Republican voters in the 2004 US Presidential election? (2)
Disadvantages:
It is not a mainstream technique and some might say that it hasn’t been used enough to say it is definitively successful in electoral or retail settings. Success in one major election is no proof it works. Its opponents might say it is a useful adjunct to traditional polling techniques.
Just because someone conforms to a particular consumer profile doesn’t mean that they’ll vote a particular way. What proportion of the electorate is like this?
The electorate might be annoyed if they found out that data gleaned from their credit or loyalty cards is being used and this might rebound on the party that used it.
(2) What other techniques could Karl Rove have used to identify Republican voters?
There are a number of more traditional survey techniques he could have used:
Surveys: Phone, personal or online surveys could be used to identify those supporting the party. Each has subtle advantages in terms of the group surveyed, with phone surveys perhaps better for an older demographic and internet surveys being better for younger voters. They have the advantage that people can be asked about their opinions in more depth and questioned about their past voting history also.
Focus groups: These may be more expensive than data mining but they are very useful in testing opinions in detail. They could be used to test if a certain consumer profile actually does support a certain party and what policies resonate most with them.
(3) In which other areas could data mining be used?
Health: to identify patterns of disease and the spread of same amongst certain areas of the population, as gleaned from hospital data sets.
Urban planning: to decide where to build houses and new road networks, according to public authority data sets.
New store planning for retailers, according to residential and spend data from consumer purchase records.
New product development, according to the type of goods already purchased- from a producer’s inventory data set.
(4) What are the ethical dilemmas involved in using data mining?
Should credit card companies have the right to sell data about their customers, even if anonymized? Don’t we have ownership in some respect of information about ourselves?
Should decisions be made by governmental authorities on the basis of data mined findings from large data sets? How representative are they? Do they have an empirical basis and how safe would be the decisions made on them?
Is it acceptable to reduce consumers or voters to particular categories and to treat them accordingly? Do people act according to a particular category they find themselves in, without knowing it, or is that too simplistic and so not worth taking seriously?
Assessments
Annotated bibliography (1)
Annotated bibliography (2)
Cover sheet for AB
Suggestion: keep the cover sheet very simple (your full name, student ID and class (e.g. BBS200A).
And:
Word count (e.g. 1,099 words).
Urkund score: (e.g. 20%).
Annotated Bibliography: LMS materials
Details required in the annotation (1)
(1) Full Reference Details (Chicago)
(2) What was the aim of the research? The aim (or purpose) of the research was to…
(3) How did the researchers/authors conduct the research? Or: What did the authors do? Was the research Qualitative or Quantitative or Mixed Methods? What research instrument/s did the researchers use: Interviews, face-to-face survey, observation, online survey…?
(4) Who were the participants in the research? Who were the participants? Were they students, academics (a.k.a. lecturers; professors), business managers, employees, employers of graduates? How many participants were there?
(5) Was a sample used? Did the researchers use a sample? How did the researchers/authors obtain their sample? How many potential participants did they identify? How many participants actually took part? What was the response rate?
Details required in the annotation (2)
(6) What were the main findings or results of the research study? Main finding 1:… Main finding 2:… Main finding 3:… Main finding 4:…
(7) What are the strengths and weaknesses of the paper (Critique)? Consider research methods, number of participants, type of participants, geography, participants...
Strengths of the research are:… Weaknesses of the research are:…
(8) What did the researchers conclude?
(9) Does this article help answer the research question: Are business students work ready? If so, how does it help answer the question?
Example provided by Anne Clear