Post on 05-Feb-2022
Completing Research
• Develop the idea to understand focus or research objective –observing people, events
• What’s already known? –company data, literature, past research
• Develop the research questions for your investigation
• Prepare an appropriate research strategy • Options - philosophy, practical, develop previous work
• Choice of method and administration• Consideration of approach to analysis• Development of questions• Sampling
• Collection of data – evaluate and review through the collection process • Analysis and review
• Writing up
Developing the Research ProcessLiterature:
Problem, Focus and Objective Alignment
The Problem
Frequent and long delays in the airline business are sometimes inevitable and translate into poor customer satisfaction and significant cost to the business. The problem concerns the impact of delays on customer satisfaction on services and the organisation
Research
Question
Research
Objectives
Research problem
The focus of this study is twofold:
1) To identify the factors that influence the passengers’ waiting experience
2) To investigate the possible impact of waiting on customer satisfaction and service evaluations
The objectives
1. What are the factors that affect the perceived waiting experience of airline passengers and to what extent do these factors affect the perception of waiting times?
2. What are the affective consequences of waiting and how does
3. affect mediate the relationship between waiting and service evaluations?
4. How do situational variables (such as filled time), influence customerreaction to the waiting experience?
Source: based on Sekaran & Bougie (2013)
Questions to ask:
• Time Frame• Cross-Sectional• Longitudinal
• Unit of Analysis• Industry, organisation, team, department, individual
• Sample(s)• Who, what, single sample, multiple samples
• How will they be selected
Data Collection Approaches
• Quantitative methods• Surveys• Fixed Interviews
• Qualitative methods• Interviews (Individual or Group)• Observation • Diaries• Other qualitative methods e.g. ethnography, auto-ethnography,
action research
• Multi-strategy methods• Mixed Methods• Case Studies• Secondary data analysis• Combined methods
Sampling
Rules of thumb:• For survey research, present a minimum of 25 responses (aim for 50), for comparing groups aim for minimum
• 8 responses per group)
• For in-depth interview research, 12-15 one-hour-long interviews would be typical
• For group interviews 18-25 (3-5 groups of 4-7 people)
• For case study research, 3-5 cases
• All these depend on the particular topic and research objectives
Sampling
Probability Non-probability
Simple random
Systematic
Stratified random
Cluster
Convenience
Purposive
Quota
Self-selection
Snowball
Steps for Sampling
• Who are the population• relationship to the investigation context and question
• What sampling frame• numbers, important characteristics
• Sampling approach• Probability or non-probability
• What is an appropriate sample size• Method(s), population
• Select actual sample(s)
• Do the research
Exploring Your Idea With Your Tutor
You should give a short ‘presentation’ to your partner (or trio) about the topic particularly on the research question (focus) and objectives and any specific ‘evidence’ (theory or practice) that you have already ‘gathered’ to support your focus. You might also consider the broad approach to collecting data –for example your likely sampling, type of data and likely methods.
If you are listening to the presentation please ask questions and focus on ensuring you have a clear understanding of:
• What the research question and or objectives are
• What the background or context is and why the research is important
• How the investigation might be done and with whom(if your partner has considered the fieldwork)
• Make suggestions
Mapping your Project
Leadership ProblemImportance
Search keywords
Practicalities
Current thinking
Limitations
Skills
Investigation Strategy
Key theory/theories
Key literature
Background
Management problem
Research problem
Data collectionData
analysis
Sampling
Unit of analysis
Access
Resources/time
Working Title
Key terms/ concepts
Research design
Stakeholders
Sponsor
Quantitative Research
If you can’t explain it simply, you don’t understand it well enough.
- Albert Einstein
Principles of QuantitativeResearch Recap
• Generally large numbers in the sample
• Concerned with mainly structured, numerical data and statistical analysis
• What is being measured and how (scales/questions) set in advance
• Sequential flow of activities for collection and analysis
• Looking to make generalisations
• Researcher ‘independent’
Variables Defined
• Variables are:• Observable and measurable characteristics of objects which may take on two or more values
• May ‘stand for’ concepts which are abstract or generic ideas not directly observable
• Types of variables• Dependent variables are research outcomes or consequences; they are the variables we seek to explain,
predict, and/or understand
• Independent variables are useful for explaining or predicting dependent variables• Variables can be dependent in one context and independent in another.
• Variables can be• Qualitative/non-numeric: numbers as labels or names and no right order
• Quantitative/numeric: can put in order from low to high
• Continuous: may take any value in a range and arise from measurement e.g. time
• Discrete: take integer values only and arise from counts e.g. how many jobs
Types of data and measures (scales)
• Nominal• Labels with no quantitative value and not ordered – Male/Female, Analysis through counts, percentages,
mode, chi-square• Ordinal
• Data that can be ordered or ranked in some way - Attitudes, preferences• Analysis through Percentiles, Median, Rank Order Correlation
• Interval• Data that can be ordered by the differences between the different ranks are the same – Attitudes, opinions,
temperature (C or F)• Analysis through mean, standard deviation, range, product moment correlation
• Ratio • The same as interval data but data that has a true zero –Age; costs; income; sales; units sold; number of
purchasers; probability of purchase; weight, temperature - Kelvin• Analysis through ratio, or coefficient of variation
Steps in ConductingSurvey Research
Your questionnaire
Introduction to the StudyInvitation, details of the study and
administration
1. Opening Questions1. Establish rapport
2. Screening questions
2. Research Topic QuestionsOpinions, attitudes, beliefs, etc.
Organise in logical sections and use a funnel approach
3. Classification QuestionsBranching questions, selective questions
4. Consent Questions(May include statements in the introduction)
Engage and Explain: First Critical Tasks
• Invitation and explanation of why and what
• The purpose of the investigation
• Why you are undertaking the investigation
• Approximate time to complete (average of your pilot sample)
• How the data will be used (include how you are going to ensure confidentiality)
• How the participants can access the findings (if you are going to offer them)
• Disposal of the data at end of project
• Administration
• Instructions on how to complete the questionnaire
• Instructions on how to return the questionnaire
Question development
• Closed questions easy to answer and code for analysis but give much information
• Open questions give you more information but may put people off and are harder code and analyse and may lead to misinterpretation
• Scaled and ranking questions can open closed questions enough to get more useful information and are fairly easy to code
• Put the most important questions early (in case of abandonment)
• Have a logical flow and work in themes
• Avoid leading questions a statement then ‘what do you think’
• Avoid questions that might confuse or upset double barrelled questions, jargon, double negatives,, are culturally and context aware
• General to particular
• Always set questions that are simple and easily understood
Quantitative Data Analysis
Steps in Statistical Analysis
Data Entry
Check Data Integrity
Transform Data
Descriptive Analysis
Inferential Analysis
Steps in Analysis
• Enter Data• To create the dataset to be analysed and set up the spreadsheet• Variables (questions) as columns, • Allocate simple labels for each column as first row• Cases (respondents) as rows; one case per row• Allocate each case (respondent) a unique numerical ID• Numerically code categorical data e.g. Male = 0; female = 1
• Create a master plus backup copy• Work from a copy• Record any changes made to the dataset
• Check Data Integrity• To remove errors and inaccuracies in the dataset• Examining the data for:• Data entry errors• Missing values• Outliers
• Transform the data• Reverse code negatively-worded questions• Calculate summated scales• Reduce number of categories (e.g. age, residence, etc.)
Exploring Numerical Data
• Get a ‘feel’ for the data• Test assumptions (e.g. distribution and mean)• Describe the sample
• Use descriptive statistics:• For nominal (categorical) data
• Frequency counts• Pie charts
• For interval/ratio data• Measures of central tendency (mean, mode and • Measures of dispersion• Shape of the distribution
Distributions
• A distribution is the set of values of a set of data possibly grouped into classes
• Distributions may be summarised by measures of central tendency and variability as well as shape
• The normal distribution is a very important type of distribution that is used widely in inferential statistics
The Normal Distribution
About 68% of the observations fall within 1 standard deviation of the mean,
that is between -σand +σ
All normal density curves satisfy the following property:
About 95% of the observations fall within 2 standard deviations of the mean,
that is between -2σand +2σ
About 99.7% of the observations fall within 3 standard deviations of the mean,
that is between -3σand +3σ
Descriptive and Inferential Analysis
• Descriptive Statistics• Help us to summarise and describe the sample(s) and the
measures . Tests include measures of central tendency –mean, mode, median, measures of spread – standard deviation, variance, quartiles
• Inferential Statistics• Allow us to infer or make predictions (hypotheses) about
relationships (not causality). Tests include Correlation, regression, ANOVA, t-test
• Chi Square
• Allows us to test relationships for small or non normally distributed (nominal) data
Values
How long an employee Count % of total
1-3 years 14 22.22%
Less than 1 year 26 41.27%
More than 3 years 23 36.51%
Grand Total 63 100.00%
Summarising and graphing non-numeric data
• Frequency distribution: • A tabular summary of the data showing the number or
frequency of items in several non-overlapping classes: relative frequency or proportion of a class is its frequency divided by number of observations and the percent frequency is the relative frequency multiplied by 100
• Pie chart: • Conveys impression of relative magnitude and useful for
displaying a limited number of categories
• Bar chart: • Better than a pie chart for conveying frequencies and typically
drawn vertically and bars should be of fixed width and separated – different types e.g. stacked or side-by-side bar charts
Job level Frequency Percent
Frequency
Cumulative
Percent
Senior
Manager
Manager
Other
231
498
616
17.2
37.0
45.8
17.2
54.2
100.0
Total 1345 100.0
Displaying non-numeric data
Frequency Distribution for Job Levels in Study Pie Charts
Job level Frequency Percent
Frequency
Cumulative
Percent
Senior
Manager
Manager
Other
231
498
616
17.2
37.0
45.8
17.2
54.2
100.0
Total 1345 100.0
Bar charts
Excel’s Descriptive Statistics Function
• Use the Descriptive Statistics option
in the Data Analysis Tool Pak
• Include the histogram option – visual
representation
x1
Mean 4.25
Standard Error 0.12
Median 5
Mode 5
Standard Deviation 0.92
Sample Variance 0.84
Kurtosis -1.2
Skewness -0.7
Range 3
Minimum 2
Maximum 5
Sum 268
Count 63
Displaying and using numeric data
• Histogram• Displays distribution of single continuous variable summarised in
a frequency, relative frequency or percent frequency distribution with bars representing class intervals
• A graph of a cumulative distribution is known as an ‘ogive’
• NB unlike bar charts there is no separation between bars and choice of class limits will affect shape of distribution
• Boxplot
• Length of box is interquartile range (50% of cases), line in box is median, whiskers extend to smallest and largest value within 1 interquartile range, and outliers are shown
• Line graph• Shows mean score of a continuous variable across a number of
different values of a categorical variable or time series
Histograms
Displaying and usinginferential statistics
• Different types of correlation to test relationships
• Correlation (Pearson product moment correlation for normally distributed data and Spearman rank correlation for non-normally distributed data)
• Regression
• T-test
• ANOVA
• Chi-square – especially for non- normally distributed data
Scatterplot
• Correlation- tests if there is a relationship between 2 continuous variablesPerfect correlation coefficient = +- 1No correlation coefficient = 0You can use scatterplot or calculate the correlation statistic in Excel
Regression
Regression – predicts the strength of relations between an independent an a dependent variable
Generate scatter plots using Excel graph function.
Add trendline (Excel offers the option of displaying the formula for the line, as here).
y = 0.43x + 3.8374
1
2
3
4
5
6
7
1 2 3 4 5 6 7
The regression line(known as the least squares line) is a plot ofthe expected value of
the dependent variable for all values of the independent variable.
t-test
• Independent samples t-test –tests whether the mean of 2 independent samples is different by chance
• Paired t-test – tests two different related observations – for example the spending of the sample is greater in December than January – in effect the difference in spending for each respondent is calculated, then the average is calculated to see if that average is statistically significantly greater than zero
• NB if the p value is low the (null) hypothesis must go i.e. the difference is not by chance
Observed valuesFr
equ
ency
difference between group means
variability of the groups
test statistic =
ANOVA (analysis of variance)
• ANOVA, like t-test, also compares means of variables in this case two or more variables
• In calculating the values if the f value is greater than the f statistic the null hypothesis is rejected or
• If the p value is lower than the confidence level the null hypothesis is rejected
Chi-square
Chi-square statistic 0.000699103
Testing at <0.05 as the chi-square statistic is
Under this level and therefore it is significant i.e. the difference between observed and expected is unlikely to be by chance
Validity and Reliability inQuantitative Research
• Are the measures going to give similar findings on different occasions?
• Do the results reflect reality/practice?
• Can the results be applied to the wider population or context?
n.b. in SPSS you can test for reliability using Cronbach Alpha
Describing Your Data
• Describe the sample using appropriate numerical and graphical displays
• Sample size
• Demographics
• What are the implications for your research questions?
• Summarise key variables
• Report appropriate descriptive statistics• For the whole sample
• For relevant sub-groups
• Use tabular/graphical displays where appropriate
• What are the implications for your research questions?
• Try to structure your presentation in a ‘thematic’ way rather than just question-by-question
Writing up
• Most relevant results-• not every question• Group data into sections or themes
• Limited demographic data
• Use some charts and diagrams to illustrate and avoid just telling us what the chart says – it is what it means in the context of the analysis and the investigation focus that is important – you do the analysis not the chart or statistics
• Each chart should • show the questions • be properly labelled including the heading• show response rate• Be big enough to read
• Raw data and survey go in the appendices (not in the text)
• Don’t mislead (logical bins)
Qualitative Research
“Not everything that counts can be counted, and not everything that can be counted counts.”
“We cannot solve our problems with the same thinking we used when we created them.”
- Albert Einstein
Principles of Qualitative Research Recap
• Aim is depth of understanding
• Always use semi-structure or open-ended questions
• Process, design and actions flexible
• Less specific specification of what is to be investigated (often themes are used)
• Small sample sizes likely
• Generates words or observed data
• Can be qualitative or mixed
• Process is often iterative both collecting data and analysing data
• The researcher is closely involved in the process
• Pre-planning• Time available and topics to
be addressed• Location and venue
formal/informal ‘yours’/’theirs’/’neutral’
• Data capture• Notes and recording• Transcription?
• Informed consent• Interview guide or
‘schedule’• Briefing• Securing informed consent• Opening question(s)/ice
breaker• Main questions• Probing questions• Concluding questions/thanks
Preparing for Data Collection
Use good questions andquestioning techniques
Depth depends on the depth to which you go withyour questions• Opening questions• Follow-up questions to check understanding• Be led by respondent – listening important but keep control
of the direction• Probing questions
• Critical incident technique• Encourage expansion- How did they feel/think/ experience;
why did they think that; what happened next• Silent probe – wait for them to fill silence• Echo probe – repeat their answer and ask them to continue
(you may want to highlight key words in their response to create focus
• Funnel Approach
Avoid leading questions and themmisinterpreting you or you them
What might be transcribedand used in analysis
• Recordings and ‘field’ notes of interviews, focus groups, consultations
• Secondary data - e.g. company or industry data and reports, minutes of meetings, research data
• Audio or film data
• Your research diary
You need to make decisions about what to include and not include especially if you are not doing verbatim transcripts.
Bryman’s Four Steps forAnalysing Qualitative Data
• Getting a sense of the data• Read text as a whole- make notes at the end. What are the major
themes? Unusual events/issues? Group cases into types/categories. How does it all relate to the literature?
• Preparing the data• Re-read, marking text up – highlight/underline/write in margins etc.
Identify labels for codes, highlight key words/phrases etc. Note any analytical ideas that are suggested by the data
• Coding the data• Ascribe codes or labels to the data systematically. Review codes and
eliminate or combine any replications, consider any logical groupings especially in relation to the research questions (you might use NVivo for coding which you can download via the main university website)
• Analysis• What can you see in terms of patterns of experiences, views, perceptions
etc. What is significant from the participants’ perspectives?Where are the links and inter-connections in relationto the interviewees AND the research questionand literature
In Qualitative analysis we are looking for patterns, themes, similarities and contradictions
• These can be found in• (Thematic or template analysis)
• Repetitions from the data (written or verbal)
• Metaphors
• Common actions if you are observing
• Similarities and differences
• Linguistic connections – words or phrases(conversation analysis or discourse analysis)
Analysing Qualitative Data
Adapted from: Miles and Huberman (1994)
Data Reduction
‘Coding’
Data Display
‘Visualisation’Data Collection
Conclusions:
Drawing/verifying
The Qualitative Jigsaw
Preparing qualitative data for analysis is like sorting out a jumble of jigsaw pieces without knowing what picture you are aiming to make (Saunders et al)
Thematic Coding as a Selective Filter
• How Big or small should the chunks be
• How fine or broad
• How deep or high level
• (Thick or Thin analysis)
Code A Code ECode DCode CCode B Code F
Raw Data
First of all we face the scepticism. I think when I took the job on, nine months ago now, the reaction of my peers was probably "that's great, sounds very exciting -they'll never let you do it". I think we've overcome that to a large extent. Then there is resistance, not because people don't like the idea, but because inevitably it means we need resources, we need skills, we need co-operation from other business units and it’s not necessarily at the moment been proved to them that by giving us those things it helps them towards their objectives at all. So we've had to effectively work with the top teams to make sure that the way the
The Coding Process
RES
C-OBJ
Data copied and pasted into table
SCEP
Code Abbreviation
Definition Interview 1
Scepticism SCEP Doubt the credibility of the programme
First of all we face the scepticism. I think when I took the job on, nine months ago now, the reaction of my peers was probably “that’s great, sounds very exciting – they’ll never let you do it”(1:1-3)
Conflicting objectives
C-OBJ Conflicts between explicit or implicit objectives
It’s not necessarily at the moment been provden to them that by giving us those things it helps them towards their objectives at all (1:7-9)
Resource Res Conflicts over resource allocation/availability
Then there is resistance, not because people don’t like the idea, but because inevitably it means we need resources, we need skills, we need co-operation from other business units (1:4-7)
Themes and their codes
Extract of data with page and line number
Definition of theme/code
Abbreviation for coding
Thematic Coding
• Thematic analysis• Identifying themes relevant to your research question
• Coding
• attaching labels (codes) to them so they are indexed and can be ‘recovered’ later
• Template analysis• Creating a template of codes to organise and structure your
thematic analysis
Data Reduction using Template Analysis
• ‘The essence of template analysis is that the researcher produces a list of codes (‘template’) representing themes identified in their textual data’ (King, 2004: 256)
• Themes may be identified in advance (a priori themes) but will be modified and added to as the analysis proceeds
• ‘The template is organised in a way which represents the relationships between themes, as defined by the researcher, most commonly involving a hierarchical structure’ (King, 2004: 256)
Example Template
Extract of a template from a study of clinical supervision (King and Horrocks, 2010)
1. Group dynamics 1.1. Atmosphere/climate
1.1.1. Formal/informal 1.1.2. Tense/relaxed 1.1.3. Focused/unfocused
1.2. Cohesiveness 1.2.1. Group as a whole 1.2.2. Sub-groups
1.2.2.1. Nurse/doctor 1.2.2.2. Practice-based/practice-attached 1.2.2.3.
2. Roles in the group supervision process 2.1. Supervisee’s role
2.1.1. Issue brought (what and why) 2.1.2. Comfort with the role 2.1.3. Helpfulness (or not) of the group members’ contributions
2.2. Facilitator’s role 2.2.1. Comfort with the role 2.2.2. Style adopted (incl. adherence to model) 2.2.3. Clarity of role…..
The Template Analysis Process
• Identify a priori themes, if appropriate
• Familiarise yourself with your data
• Start coding your data (a subset may be enough)
• Construct your initial template• Hierarchical coding • Parallel coding• Descriptive and interpretive themes
• Apply template to your full data set
• Develop your template • Insert, delete, change scope of a theme or change the higher-
order classification
• Use ‘final’ template to interpret and analyse your findings
See 10 part YouTube video of Nigel King taking a step by step approach (https://www.youtube.com/watch?v=EEIt5pag4Z8 )or the QDA website
Coding and Tabulating Data in Excel
Tips:
1. Create a column for each relevant demographic characteristic (e.g. gender) and use Excel’s sort function to group respondents by demographic characteristic to help with pattern analysis
2. Use cut or copy and ‘paste special - transpose ’ in Excel to transpose rows and columns
Themes/codes (e.g. from your
template)Interviewees
Data cut-and-pasted from interview
notes/transcripts (inclpage: line no.)
Data Display using Matrix Analysis
Interviewee 1
Interviewee 2
Theme/ code A
Category W Category X Category Y Category Z
Group 1
Group 2
Level 1: Cross tabulating themes/codes by interviewee or case
Level 2: Cross-tabulating groups and broader categories or categories with other categories
Cells contain data summaries and/or quotations
Cells contain data summaries and/or quotations
Theme/ code A Theme/ code A Theme/ code A
Qualitative Data AnalysisWord Cloud for Martin Luther King’s‘I have a dream’ speech
http://www.mymarketresearchmethods.com/quantitative-vs-qualitative-research-whats-the-difference/
Post-It Notes Identifying Patterns or Themes
Source Justin Wilcox, Relentless Entrepreneur blog
Validity and Reliability in Qualitative Research
Validity• Transparency in both the method(s) of collection and the method(s) of analysis (audit trail)
• Be clear about why you are making the specific interpretation
• Interrogate your interpretation– do rival explanations and organisation of the data result in different findings
• Are the results credible and authentic? - Has the researcher remained faithful to the participants’ experiences and perspectives (Hammersley 1992)
Reliability• Draw on multiple perspectives, sources of data - or even methods
• Consistency within the data and across the data
• Consider the transferability of your results and findings to other contexts either within the same setting or other settings
• Consider the issue of general relevance (Myers, 2013)
Developed from Cassells
Questions to ask of the dataLet the data ‘speak’!
• Patterns and themes of similarity or difference – why are they different or the same – context factors – role, age, experience, time, opportunity, word repetition or differences
• Does the data suggest groupings or clusters?
• Content analysis /counting – be careful not to use similar approaches to descriptive statistics use this approach sparingly
• Keep asking why/what /how/ who/when
Presenting Quotes
Tips:1. Layout of the quote and identification (anonymised) of the speaker2. Quotations are introduced, presented and then the ‘so whats’ are drawn out3. The notion of being ‘equity sensitive’ is used to link these ideas together conceptually
Dick (2006)
Data Analysis Exercise
As part of a study trying to determine:
What drives students decisions to do a MBA?
• Develop a short interview schedule to address the following:
Why did you decide to study an MBA?Why did you choose to study?Now that you are nearly at the end, what have been the main benefits?
• Use your interview schedule to help you interview a partner and they will interview you.
• When you have completed the interviews you should meet with the rest of your syndicate group to code and present the data
• code the data and develop a code list. • Prepare definitions of the codes you use• Give examples of coded data• What are your preliminary answers to the research question?
• What are the strengths and weaknesses of this way of researching the topic?
Mixed and Combined Methods
Mixed Methods
..a procedure for collecting, analysing and mixing both quantitative and qualitative methods in a single study or series of studies to understand a research problem
Cresswell and Plano Clark 2011
+
Combined Methods
There is no ‘one size fits all’ solution but a key requirement is a strong sense of reflexive engagement throughout the research process.
Pritchard in Cassells and Symons 2011)
+
Qual-Qual Examples of Commonalities and Challenges
• Commonalities• Social constructionism
• Methodologically broad
• Emphasis on reflexivity
• Interest in materiality and hence offers deeper understanding of phenomenon and context
• Challenges• Person centred or text centred (e.g. ethnomethodology /
interviews/secondary data)
• Outside looking in (eg interviews) or inside looking out (e.g. participant observation)
• How to deconstruct and make sense of the data (e.g. discourse analysis) or describe data (eg ethnomethodology)
Adapted from Pritchard and Symons
Mixed Methods – Rationale(Greene et al 1989)
• Triangulation – establishing convergence, corroboration and or correspondence of results from different methods
• Complementarity – seeks elaboration enhancement, clarification, illustration from one method to another
• Development – seeks to use results from one method to inform or develop the other
• Initiation – using discovery of new perspectives, paradoxes etc from one method to recast questions or results of another method
• Expansion – to extend the breadth or rangeof one method
Research Philosophy
Designingthe Mix
Convergent – simultaneous collection – consideration of similarities and dis-similarities from the data. Can offer better generalisations but issues include how to merge the data and how to make sense of the diverging results
Sequential – one method undertaken and analysed and then the results inform the next method – clarity of where results come from but issues to consider include practicalities including time and what results are most important to extend and investigate
Relationships between the Mix
Quant
QUANT
QUAL
Qual
QUAL=
QUAL
Bryman suggests 9 permutations
Pure Qualitative equal
QUANT=
QUAL
Qualitative dominant
Equal status
Quantitative dominates
Issues for Mixed Methods
“Mixed method research is simply more that reporting two distinct ‘strands 'of quantitative and qualitative research, these studies must also link or connect these ‘strands’ some way……to provide a fuller understanding of the phenomenon under study” Cresswell and Tashakkon 2007
Connections important and raise questions about
• How to analyse – using mix to inform and integrate analysis
• How to report the findings and make sense of the analysis
• How to ensure a mixed method approach is justified
Summary