1 Colin Clarke-Hill and Ismo Kuhanen Research Methods Seminar Data Collection Choosing a Strategy Quantitative Data Qualitative Data Getting Started Nex t Research Methods
Mar 26, 2015
1
Colin Clarke-Hill and Ismo Kuhanen
Research Methods SeminarData Collection
Choosing a StrategyQuantitative Data
Qualitative Data
Getting Started Next
Research Methods
2
First
Choosing A Research Strategy
• The general principle is that the research strategy or strategies, and the methods or techniques employed must be appropriate for the questions you answer
• Quantitative / Qualitative
• Qualitative Methods:
– “An array of interpretive techniques which seek to describe, decode, translate and otherwise come to terms with the meaning, not the frequency, of certain more or less naturally occurring phenomena in the social world” (Van Maanen (1983)
Previous Next
3
First3 Traditional Research
Strategies• Experimental
– measuring the effects of manipulating one variable on another variable
• Case Study– development of detailed, intensive
knowledge about a single ‘case’, or of a small number of related ‘cases’
• Survey– collection of information in
standardised form from groups of people
• May also have a Hybrid Strategy or Action Research
Previous Next
4
First
Classification of the Purposes of Enquiry• Exploratory:
– to find out what is happening
– to seek new insights
– to ask questions
– to assess phenomena in a new light
– usually, but not necessarily, qualitative
Previous Next
5
First
• Descriptive:– to portray an accurate profile of persons,
events or situations
– requires extensive previous knowledge of the situation etc. to be researched or described, so that you know appropriate aspects on which to gather information.
– May be qualitative and/or quantitative
• Explanatory:– seeks an explanation of a situation or
problem, usually in the form of causal relationships
– May be qualitative and/or quantitative
Classification of the Purposes of Enquiry
Previous Next
6
First
Purpose & Strategy
• case studies as appropriate for exploratory work
• surveys as appropriate for descriptive studies
• experiments as appropriate for explanatory studies
• NB: These links are not immutable & each strategy can be used for any or all of the three purposes (See Yin, 1981)
Past use of the 3 main strategies has tended to make following links:
Previous Next
7
First
Experiment & Case Study
Survey
How
Why
Who
What
Where
How much / how many
Strategy Research Question
Strategy & Question Previous Next
8
Colin Clarke-Hill and Ismo Kuhanen
Collecting Quantitative Data
SamplesSurveys
Pitfalls etc...
Research Methods
Previous NextGetting Started
9
First
Research Design
•Defining the Problem
•What Data is needed to solve it
Research Question Previous Next
10
First
Typical Stages
• Problem Definition• Review of Secondary Sources• Select Appropriate Approach for the
Collection of New (Primary) Information
• Determine the Details of the Research Design
• Data Collection• Analysis and Interpretation of the
Data• Evaluation and Recommendations
Previous Next
11
First
Data Types
• Secondary Data - Information that that is available from existing published sources– Internal to the Company– External to the Company
• Primary Data - Information has been collected for the first time– Can come from internal sources– External sources - survey data
etc..
NextPreviousResearch Methods
12
First
Data Sources Cont.
• Primary Data is usually collected by the means of:– A Survey– Depth Interviews– Observation of behaviour– Establishing Motivations– etc..
Previous Next
13
Colin Clarke-Hill and Ismo Kuhanen
The Questionnaireand its Design
The questionnaire is usually the common form of collection of
survey data
NextPreviousFirstResearch Methods
14
First
Questions can either:
•Classify - organizations or people
•Describe Behaviour•Discover Attitudes and Perceptions
Previous Next
15
First
Types of Questions
• Open Ended - Some sort of free form and expression on the part of the respondent
• Closed ( Structured)– Yes/No– Multiple Choice– Rankings– Check Lists
Previous Next
16
First
Wording the Questions
• Avoid Ambiguity• Consider the Respondent’s
ability to answer• Consider the Respondent’s
willingness to answer• Avoid Influencing the
Answer
Previous Next
17
FirstQuestion Sequence
• Initial questions to provide motivation
• Logical order - general to the specific. This is known as funnelling.
• Rotating the questions to reduce bias
• Difficult questions - where do you place these in the questionnaire ?
• Routing and Excluding - be careful of the logic of your questionnaire.
Previous Next
18
First
Some Standard Questions
• Have you ever?• Do you ever ?• How Often ?• When did you last ?• Who does it ?• In what way do you do it ?• Which do you do more ?• In the future will you ?
Previous Next
19
FirstSome Points to Consider
• Is your questionnaire too long ?• Do you need cards for your respondents ?• Do you need to aid recall ?• What is going to happen after the
survey ?• Avoid multi punch data - this leads to
coding problems and data analysis problems
• How are you going to administer the survey?
Previous Next
20
First
Measuring Attitudes
• Attitudes influence preferences and are related to behaviour. They usually contain three components: – Beliefs– Emotions and – Behaviour
Previous Next
21
First
Measuring Attitudes
• Two Stages:– Pilot - some form of depth
interview with a small group to ascertain the key characteristics
– Survey to measure the attitudes and perceptions
• Use:– Adjective check lists– Multiple choice questions– Rating scales: Monopolar vs Bipolar
Scales, Likert Scales - 5 point or 7 point scales
Previous Next
22
First
Some Problems with Scales
• Are the chosen adjectives ambiguous ?
• Can meaning differ between respondents ?
• Scale length - short scales may not be sensitive enough, long scales may be unmanageable.
• What about the ‘don’t know’• Respondents often tend to
choose the mid point.
Previous Next
23
FirstHow do you Reach your
Respondent ?
• Mail• Telephone• Personal Interview
All have different advantages and disadvantages for the researcher - Follow up to obtain a better response rate.
Previous Next
24
Colin Clarke-Hill and Ismo Kuhanen
Sampling
Why do we sample ?Types of Sample
Issues about Samples
Sample Size CalculatorMarket Research
Glossary of sampling and quantitative research
NextPreviousFirstResearch Methods
25
First
Basic Premise of Sampling
• The researcher is looking to generate data has is representative of the population from which the sample is drawn.
• Samples are practical as resources are finite
• Samples can give an accurate view of a phenomenon.
NextPrevious
26
First
Sampling Process Decisions
• Who is to be surveyed ? The sampling unit.
• How many to be sampled ? The sample size.
• How can they be selected ? The sampling procedure.
NextPrevious
27
First
Sampling Procedures• Two broad categories:• Probability Samples
– Random - everyone in the population has an equal change of being chosen.
– Stratified Random Sample - to account for a variable in the population
– Sequential Sample– Cluster Sample
NextPrevious
28
First
Sampling Procedures• The opposite of Probability
Samples is the Non Probability Sample:– Quota Samples– Judgmental Samples
NextPrevious
29
FirstSampling Procedures
• In Selecting samples and sample size allow for non response. This can reduce the validity of your results
• Also be aware of bias in your survey. – Bias in non response– Bias by the interviewer– Bias through the questions asked– Sample bias– Bias from the respondents
NextPrevious
30
First
Sample Frame• Adequacy - in coverage of the
population• Completeness - missing units
cannot be selected introducing bias• No duplication - double counting • Accuracy - is the sample frame up
to date ?• Convenience - Can the sample
units be accessed at reasonable cost ?
NextPrevious
31
First
Finally• Editing
– Has each relevant questions been answered ?
– Accuracy - some inaccuracies can be spotted by eye.
– Uniformity - have interviewers interpreted the questions and instructions in a uniform way ?
• Coding - think about your coding for data input
• Tabulating - think about your tables and cross tabulations. This is useful for the next stage on analysis
NextPrevious
32
First
Research Methods
Qualitative Data Techniques
NextPrevious
33
Colin Clarke-Hill and Ismo Kuhanen
Planning Qualitative ResearchQualitative design involves articulating compelling & researchable questions salient to target respondents
NextPreviousResearch Methods
34
First
Qualitative Research• Qualitative inquiry demands an
ever-present curiosity on the one hand and an ever-present suspicion on the other. The curiosity is manifested in the uneasiness with existing answers. This uneasiness generates new questions (Hawes,1975)
• Works at many levels... In the early stages the researcher asks questions about a problem that grows increasingly subtle, pertinent and penetrating.
NextPrevious
35
First
Qualitative Research• In the field questions used as
navigational tools - simple, naive, wise or purposely contradictory ones; asking the same question of a number of people; mutating it to fit different segments, expertise, etc.
NextPrevious
36
First
Qualitative ResearchQualitative researchers develop unique design solutions for every project
• Need for understanding drives the qualitative researcher to take time to understand culture or research segment.
• Not just a language issue, social/cultural taboos as well.
NextPrevious
37
First
Some issues of importance
Uncertain control qualitative researcher can expect to exercise in the field
• Researcher must be able to fit-in with events or people, that operate by their own rules of conduct. Must learn when to watch, when to listen, when to go with the action, when to reflect on pieces of info, and when to intervene tactically (& tactfully). This is totally alien to quantitative research which strives to be in control.
NextPrevious
38
FirstSummary of Qualitative Design
Process
1. Question formulation in myriad forms is the core feature of designing and starting a study.
2. Every scene & situation presents a unique, never-before encountered configuration of features, requiring strategic flexibility on the part of the investigator.
3. The researcher willingly shares control in the research scene in the interest of learning the rules and meanings of social life from the inside.
NextPrevious
39
First
Phenomenology• This kind of inquiry recognises that the
phenomena being studied are sentient.• Ambiguity is anticipated and planned
for.• The best preparation consists of a
sense of purpose, some researchable questions, an understanding of resources available and an idea of the overall features and dynamics of the setting to be entered.
• Link with presentation on Research Approach
NextPrevious
40
First
Depth Interviews (Individual & Group)
Allows access to a different level of reality, ...deeper than in quantitative research.
Interviews are dominant method in qualitative research.
Qualitative Data CollectionNextPreviousResearch Methods
41
FirstDepth Interviews –
individual and group
Totally different from structured, rigid quantitative interviews, qualitative interviews referred to as “conversation with a purpose”! The interview is loose, informal, flexible, interactive.
But interviewing is not just conversation, it is informedby its purpose .... conversation occurs within thisremit.
NextPrevious
42
FirstDepth Interviews –
individual and group
Process is also dynamic ...
“What distinguishes in-depth interviewing is that the answers given continually inform the evolving conversation. Knowledge thus accumulates withmany turns at talk. It collects in stories, asides, hesitations, expressions of feeling and spontaneous associations ... The specific person interviewing the “I” that I am, personally contributes to the creation of the interview’s content because I follow my own perplexities as they arise in our discourse.” (Paget 1983)
NextPrevious
43
First
ObservationGuidelines:
• Best conducted in relationships with respondents.
• Characterised by a difficult and often ambiguous course of study.
• Analyst must be disciplined.• Requires attention to detail.• Invaluable for ethnography and case
study research.• Observation allows the subtleties of
responses to be noted and placed within the overall spectrum of data collected for the study.
NextPreviousResearch Methods
44
First
ObservationParticipant observation preferred ...
“The participant observer gathers data by participating in the daily life of the group or organisation he studies.
He watches the people he is studying to see what situations they ordinarily meet and how they behave in them. He enters into conversation with some or all of the participants in these situations and discovers their interpretations of the events he has observed.” (Backer 1970)
NextPreviousResearch Methods
45
First
Case Study Methodology• Typical features:
– selection of a single case (or a small number of related cases) of a situation, individual or group of interest or concern
– study of the case in its context
– collection of information via a range of data
– collection techniques including observation, interview and documentary analysis
NextPreviousResearch Methods
Resources• Case Study
Method
46
First
Case Study Methodology
• 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 (Robson, 1993)
Previous Next
47
First
Characteristics of Qualitative Research
• DIAGNOSTIC
• DEEPER UNDERSTANDING
• IMPRESSIONISTIC
• PROBING
• OBSERVES & REFLECTS
• SUBJECTIVE
NextPrevious
48
First
Focus of Qualitative Research
• WHAT?
• WHY?
• HOW?
NOT HOW MANY
NextPrevious
49
FirstQualitative Approach Used in
Research Requiring:
• EXPLORATION OF CONSUMER
MOTIVATIONS, ATTITUDES & BEHAVIOUR
• IDENTIFICATION OF DISTINCT BEHAVIOURAL GROUPS
NextPrevious
50
First
Qualitative Research Techniques
• DEPTH INTERVIEWS
• FOCUS GROUPS
• PROJECTIVE TECHNIQUES
NextPrevious
51
FirstCharacteristics of
Depth Interviews
• UNSTRUCTURED QUESTIONNAIRE
• QUESTION AREAS (DERIVED FROM RESEARCH OBJECTIVES), NOT FORMAL QUESTIONS
• RESPONDENT TALKS FREELY WITHIN CONFINES OF BRIEF
NextPreviousResearch Methods
52
First
Characteristics of
Depth Interviews
• INTERVIEWER PROMPTS AS NECESSARY
• DURATION - 1 TO 3 HOURS
NextPrevious
53
First
Characteristics of Focus Group Interviews
• 6 TO 12 PEOPLE• RESPONDENTS UNIFORM TO CREATE COHESIVE GROUP
• FRIENDS OR RELATIVES NOT RECOMMENDED
• OPINIONS ASKED, DISCUSSED AS A GROUP
NextPreviousResearch Methods
54
First
Characteristics of Focus Group Interviews
• INTERVIEWER FACILITATES DISCUSSION
•DURATION 1 TO 3 HOURS
NextPrevious
55
FirstAdvantages of Depth
and Focus Group Interviews
• RICHNESS OF INFORMATION
• INSIGHT & UNDERSTANDING
• EXPLORATION
NextPrevious
56
FirstDisadvantages of
Depth and Focus Group Interviews
NEED SKILLED INTERVIEWS
LACK OF STRUCTURE
DURATION OF INTERVIEW
DIFFICULT TO SET UP
DIFFICULT TO MODERATE
SPECIALISTS FOR DATA ANALYSIS
REQUIRED
NextPrevious
57
First
Qualitative vs Quantitative Research Research
QUALITATIVE QUANTITATIVE
EXPLORATION ASSESSMENT
DIAGNOSIS DESCRIPTION
UNDERSTANDING QUANTIFICATION
INSIGHT ENUMERATION
AGGREGATION
NextPreviousResearch Methods
58
First
QUALITATIVE QUANTITATIVE
UNSTRUCTURED STRUCTURED
FLEXIBLE RIGID
UNSTRUCTURED/ STRUCTUREDSEMI-STRUCTURED
SMALL SAMPLE LARGE SAMPLE
Qualitative vs Quantitative Research Research
NextPrevious
59
First
Qualitative Approach Used
in Research Requiring:
•GREATER UNDERSTANDING &
KNOWLEDGE
•CLARIFICATION OF REAL ISSUES
•GENERATION OF HYPOTHESIS
• IDENTIFICATION OF RANGE OF BEHAVIOUR
NextPrevious
60
First
That’s all folksPrevious