RBM

• Logistics
• Costing
• Suppliers
• Contract Opportunities
• Promotion
• Staff Performance
• Environmental Performance
• Quality Assurance

• Research Proposal/Project
• Literature Review
• Critical Writing Methods
• Research Philosophies/Methodologies
• Sampling Techniques

Qualitative 

• Action Research
• Archival Research
• Case Studies
• Interviews
• Focus Groups
• Participant Observation
• Coding/Decoding data

Quantitative

• Using excel for Research
• Understanding Descriptive Statistics
• Implementing Descriptive Statistics
• Understanding Tests & Significane
• Implementing Tests & Significance

• Primary Sources
• Secondary Sources
• Ensure it is balanced, relevant, reliable, ethical

Systematic, Comprehensive, Thorough -  In order, covers everything that is relevant or choose an area of study to cover

Seeks to discover the truth - Finds questions that needs to be answered & answers them

Results should be trustworthy and meaningful - Meaningful = Value

Trustworthy = sensible approach to research that is repeatable

Builds on existing knowledege & understanding - Adds Layers

Considers what has happened (reflection) - Helps make more informed choices, improves efficency in relation to time and money

Considers what needs to happen (action) - No point if not going to be used or nothing changes

Donald Rumsfield - NATO 2002

• Research on theory informs practise
• Research on practise infroms theory

• Philosophy adopted depends on the way we think about developing and creating knowledge (Saunders 2000)
• Considers the practial considerations required for undertaking research
• Certain philosophies are better suited to certain research problems

Positivism - considered a science, controlled by rules and laws, measuring world around us, emotionally detached, researcher is independant of the research, does not affect and is not affected by it

Interpretivism - science misses the rich details, too complex to be controlled by rules and laws, tries to understand the world around us, empathy, part of the research (sensing)

Realism - what we sense is reality. Direct Realism - operates on a single level in an unchanging world. Critical Realism - operates on multiple level in constantly changing world (processing and sensing)

How we view reality?

• Positivism - Objective, independant of social actors
• Interpretivisim - Subjective, socially constructed upon social actors 
• Realism - Objective, independant of social actors but interpretated through our social conditioning 

What we deem as acceptable knowledge?

• Positivism - observable phenomena to provide credible data and facts
• Interpretivism - social pheenoma to provide subjective meaning 
• Realism - focuses on explaining social world around us using observable phenomena (facts and data)

• Testing Hypotheses to find relationships
• Collecting/analysing data (testing theory)
• Quant
• Researcher is independent of research
• Characteristics are measurable 
• Results are generalisable so large sample is required

• Exploring nature/context of the problem
• Developing theory from data (creating theory)
• Views and opinions
• Qual
• Researcher is part of the research
• Generalisation not so important so small sample is accepted

• Moves from theory to data and data to theory
• Starts with observation and develops contestable theory
• Quan & Qual

Consider socio-economic groups (varying levels of disposable income)

• Higher professionals/managers
• Administrative/supervisor
• Skilled/Semi-skilled/unskilled
• Retired, Students, Unemployed

Understanding Generationns

• Silent Generation - 65+ followed rules, very principled
• Baby Boomers - 55-64 worked hard, golden times, high disposable income
• Generation X - 35-55 IT aware, low disposable income
• Generation Y - 20-35 grown up with tech, individualistic, value their own time
• Generation Z - 0-19 mobile based apps, instant answers, saturated social media

Balancing Data

Religious Views, Political Views, Maritial Status, Home Owner vs Renter, Qualification & Education, Status, Gender

• Continous Data - can take any value  (e.g time or temp) 
• Discrete Data - can only take values in a range (e.g integers: no of people)
• Categorical Data - can be grouped but not measured
• Numerical data - can be measured
• Ranked (ordinal) Data - Data that can be grouped but also ranked (e.g agree to disagree, xs-xl)
• Descriptive (nominal) Data - data can be grouped but not ranked (e.g types of vehicle)

• A census collects data from everyone in a population
• In reality this is not rational 
• Extremely expensive, time consuming and often not accessible
• Use a sample as it is representative of whole population as it considers a range of cases in a meaningful way (Saunders 2012)

non probability - 

• characteristics often known therefore better for quantitative
• moving from theory to data (testing theory)
• generalisation possible
• always facts/numbers

probabibility -

•  characteristics often not known better for qualitative
• moving from data to theory (creating theory)
• generalisation possible but not on statistical grounds
• views/opinions

• Random Sampling - sampling at random across the data
• Systematic Random Sampling - sampling at regular intervals throughout the data using formula
• Stratifed Random Sampling - identifying layers in the data and then using random/systematic random in each layers
• Cluster Sampling - indentifying groups in the data and then using random/systematic in each group
• Multi-Stage Sampling - too many groups so first random/systematic random sampling but be used to decide which to include

Saunders et al 2012

• Larger sample size = lower levels of error
• Probability Sampling is a compromise between acurracy and time/cost

Sample size therefore considers:

• level of confidence in results
• Tolerable level of error
• Size of total population
• Number of categories to be included in analysis 

• 95% confidence = repeated 100 times 95 times would be fully representative of whole population
• Margin of error usually = -/+3% or -/+5%
• Smaller the margin of error required = Increased sample size

Sampling Fraction = Actual Sample Size / Total Population

e.g. 1/4 is every 4 

n = p x q (z/e)2

n = sample size

p = % belonging to a category 

q = % not belonging to a category

z = confidence in results e.g 1.96 = 95%

e = margin of error

If population is less than 10,000 it is possible to adjust minimum sample size without compromising accuracy

n ' = n / 1 + (n/N)

n ' = adjusted sample size

n = sample size 

N = total population

• Quota
• Haphazard - choosing those that are easiest to access for convenience
• Volunteer - *********** works well when population is hidden/hard to reach. Self- Selection allows individuals to take part on own accord
• Purposive - designed for specific goal - homogeneous, heterogenous, critical, theorectical, extreme

Ethics refers to good practise undertaken in reserach to consider:

• the use of humans 
• the use of animals
• the use of intellectual property (ideas, themes, works)

Should be:

• Honest
• Integrity
• Open to scrutiny from others
• Respectual to participants
• Not judge responses

Physical Access - often granted or rejected by management (e.g. database)

Cognitive Access - recieving support from others involved in the research maybe gatekeepers. Winning hearts & minds

Continuing Access - although inital access has been granted may be an incremental process that requires additional access from other people at different levels

Virtual Access - use of chatrooms, blogs? do you trust them?

Internal vs External Researcher

internal

• knows the people and practises
• low risk to organisation
• may have developed previous relationships/bias perspectives
• likely to gain required access

external

• fresh eyes
• unkown = increased risk to org
• increased reliability & validity
• objective

• do not monitor emails/calls
• do not record without permission
• do not photogragph without permission especially children
• do not judge participant responses
• no bribing/incentives
• no punishment for non-compliance
• adhere to data protection regulations whe storing personal data
• opportunity for participants to withdraw at anytime
• offensive/sensitive/ gender specific language
• do not cause embarrassment/harm or death
• Anonymity

Quant

• meaning derived from facts/number
• collect results in the form of standardised numerical data
• analysis in the form of diagrams/statistics

Qual

• meaning derived from views/opinions
• collect results through non standardised data requiring categorisation
• analysis conducted through conceptualisation

1. use categories based on previous works and theories 

2. use categories based on terms used by participants

3. use categories that emerge from data analysis

1. seek relationships and patterns in the data

2. develop testable prepositions

3. test prepositions to ensure relationship exists

An interview is a conversation between the interviewer asking questions and the interviewee answering them. 

Structured = QUANT

Unstructured = QUAL

Structured 

• Limited social interaction
• Standardised questions
• Pre coded - answers
• Quantitative data 
• Interviewer asks questions 
• Interviewee answers 

• Extended social interaction
• Non - standardised
• Themes to cover 
• Qualitative
• Questions may vary
• Not everyone has same questions
• Additional questions may be asked

• Inital social interaction
• Agreed theme for interview
• Interviewer allows intereviewees to speak (non directed)
• Qualitiative 

• opinion based questions - e.g agree to strongly disagree
• belief based e.g always to never
• behaviour based e.g once a week to everyday
• fact based e.g age 0-65+

open question

• normally about situation/event
• encouraged developmental answer
• reveals facts/attitudes
• usually start with what, why or how?

probing questions

• similar to open questions (what, why or how?)
• but with a particular focus

closed questions

data gathering

introductory

ADV

• Allows interviewer to focus on what is being said?
• Ensures accuracy in interpretation as can be played multiple times
• Enables quotes to be extracted
• Permanent copy for other researchers to access

DISADV

• May adversely effect participants if they are aware they are being recorded
• Should not solely rely on recording due to technical faults
• May stop responses from being so open
• Unsettling

generalisable

• larger and more representable the sample = more generalisable the findings
• ACTION ; careful sample selection

bias 

• tone/non verbal behaviour
• sensitive issues 
• ACTION: plan & record

reliability

• lack of standardisation = lack of variability 
• ACTION: seek trends/issues not details

validity - ability to interpret correctly

• ACTION: seek clarification reguarly

Should you use PM?

1 significant characteristic

• 1 manager uses + 1 manager does't use = 2 interviews

1 significant characteristic improved

• 3 managers us + 3 don't use = 6 interviews

2 significant characteristics

• 3 managers use female + 3 managers don't use female 
• 3 managers use male + 3 managers don't use male      = 12 interviews

3 significant characteristic

• 3 F junior use + 3 F junior don't use + 3 F senior use + 3 F senior don't use + 3 M junior use + 3 M junior don't use + 3 M senior use + 3 M senior don't use = 24 interviews

• neutral setting to avoid interuptions/distractions
• circular seating plan to ensure everyone is equal
• organised by one person and recorded for data accuracy
• usually 4-12 people - more complex issue less participants required
• horizontal slicing - people with similar experiences 
• vertical slicing - people with different experiences

• wide breadth of points
• questions asked but also challenge the opinions of others
• ends when reach an agreed concensus 

• Narrow/clearly defined single topic
• Tolerant environment where nobody's opinion is questioned
• repeated with different participants to identify strong trends
• end when saturation point is reached no more new points are made

co-ordinator of group interview is FACILITATOR 

• finding the balance between encouraging discussion and allowing freedom
• ensures equal opportunity 
• ensures structure whether that be highly structured (unlikely) to unstructured (likely)

co-ordinator of focus group is MODERATOR

• dual roles between strictly ensuring participants do not deviate from topic and generating interest in the topic

• unpick data to see trends/patterns
• allows for better understandings of issues/problems
• ability to make better reccomendations for future action

large amounts of data use NVIVO 

manageable amounts of data use recursive abstraction

1. Highlight important text

2. Extract important text

3. Paraphrase

4. Group into themes

5. Replace with codes

6. Check meaning, look for trends/patterns

(Walsh & Ungson)

• individuals
• cultures
• transformations
• structure
• ecology 
• archives

• (Laura Schmidt)
• Special Collections
• Historical Archives
• Museum Archives
• Corporate Archives
• Government Archives
• Religious Archives
• College/Uni Archives

1. Discovery

2. Access

3. Assessment

4. Sifting

5. Cross-checking

• HR records
• Financial records
• Payroll records
• Meeting notes
• Intellectual property ownership

worldcat (Laura Schmidt)

database that lists global archives

General Data Protection Regulation

• refers to any data that belongs to an identifiable living person

Actions inclusive of:

• Manipulating
• Erasure
• Recording
• Retrival

Lewin (1946)

• social process facilitating change
• addresses practical problems
• solves real issues
• collaborative apporoach (participant involved)
• Implications for wider organisation

Saunders 2012 - Action Research Cycle

1. Diagnose Problem

2. Plan action

3. Take Action

4. Evaluate Action 

5. Repeat

• Participants must co-operate and become part of the research
• Descisions are joint
• Selecting the right participants is essential

BENEFITS

• Participant part of generating outcome means more likely for it to be adopted
• Increase success = desired impact
• More likely to be embedded and change behaviours
• Once noted that behaviour needs to change it is hard to ignore
• Bottom up approach = positive impact = "fit for purpose"

(Saunders 2012) Focused on understanding the social world around us through first hand field study

involves looking into people, cultures, habits and mutual differences

• study groups and relationships that occur within groups
• output is in the form of cultural account
• uses language they use to describe themselves
• detached approach treating participants as subjects
• researchers often live amongst those being studie e.g. street gangs
• occurs in their own environment

1. Realist 

• objective, fact finding
• meaning through processes, practises, artefacts based on observation
• 3rd person
• supported by quotations

2. Interpetive

• subjective
• collects impressions to find multiple meanings
• 1st person
• supported by quotations and present perspectives

3. Critical

• investiages the impact of power, privelige and authority on those that are subjected to it
• may also adopt an advocate role (support)

case is a real life situation/event. case study is analysis of this situation/event (Van der Ham 2016)

empirical investigation into particular phenomena in its real life setting (Saunders at al 2012)

one thing in its entirety (Thomas 2011)

• don't consider details instead consider whole situation
• consider a large amount of inter-related features oppose to a survey that looks at a small amount
• consider time, people and contet
• specify your case to study (is it a thing or situation?)
• Define parameters (physical/emotional)

Two distinct sections:

the case itself the area of research

the object of research that requires illuminating

• single case -  often critical/extreme cases
• sometimes typical case
• novel investigation or one that very few have covered

• mutliple case - more than one case covered
• Are all finding replicated in all studies?
• similar charecteristics = literal replication
• differing charecteristics = theoretical replication

• holistic case - focuses on a single department/org

• embedded case -  focuses on multiple departments or multiple organisations

Description - gathering key facts and establishing context

• what is being said/done?
• do others agree? does it contribute to the problem?
• who is saying/doing it?
• are they stakeholders?/what is their role?/what is their relationship to others?
• why is it being said/done?
• is there a lack of leadership?
• where is it being said/done?
• what environment?/does the environment have constriants?/what products/services affected?
• when is it being said/done?
• chronologically?/is there a noticeable before/after/is it incremental over time?
• how is it being said/done?
• threatening behaviour?/tone?/

Analysis, to find how the facts fit together

Van der Ham 2016 - structuring your data

• identify underlying causes of issue? what contributes, why?
• develop action plan - who is involved, steps needed, how to monitor success?
• brainstorm solutions - for org/individual
• evaluate solutions - due to opportunities of context
• identify best practise - what does the literature say about best practise?
• best fit solution - to individuals/org, justify using literature

saunders 2012

oberservation where researcher becomes member of subjects grouo/org/community participating in life/actvities

welch 2000

• qualitative technique
• overt observation - awrare of researchers role/identity
• covert observation - concealed, discovers meaning social actors attach to actions, raises ethical issues

participant as observer 

• takes part in activity
• identity revealed
• e.g joins strikes/shares emotions

observer as participant

• observes activity
• identity is revealed
• purpose is known to those studied
• e.g. observing team building exercise

complete participant

• researcher takes part - fully immerses
• identity concealed

complete observer - oberserves, identity not known, purpose not known

descriptive phase - concentrate on observing describing physical setting, activities & emotions

narrative phase - write up findings & develop framework of theory that helps to explain it

focused phase - concentration on particualar people/events 

contextual data - places research in its setting key roles, org structure etc

experimental data - own views and opinions on what you have witnessed?

primary data - things you have seen e.g. diary noting what/when

secondary data - info from others on thiing you didn't see. Their own interpretation

• predetermined structure to measure observed behaviour
• quantitative - how often things happen but not why
• easy to repeat allowing comparision across days and other events
• researcher is detached focus on observation

e.g. how many times someone accesses a website as a marketing strategy

recording sheet

• coded responses
• check everyone understands the code and is using it the same way 
• make sure it doesn't overlap

• ecological validity is HIGH due to social actors being observed in natural setting
• until observer fully understands research setting may misunderstand
• but once over familiar may lead to:
• observer error - lack of understanding or over familiarity with setting
• oberserver bias - so comfortable with setting no longer objective becomes subjective
• oberserver effect - change in behaviour as observants know they are being watched

specifically relating to STRUCTURED:

• informant error - data may be unreliable if observing when under staffed or staff are ill as behaviour may change
• time error - time data is recorded must be typical period not peaks and troughs in activity

• data is all around us
• nearly everything we do leaves a data imprint. (e.g. making decisions, performing effectively, competitive edge)
• reliable
• objective
• reduces complex problems into fewer variables
• tests hypotheses
• assumes representative of population

• adapt to type of respondent (business, consumer, children)
• consider method of data collection (postal, face to face, online)
• keep length to a minimum (max 20 questions)
• conduct pilot study on small representative sample 
• don't leave highly important questions till last
• be as specific as possible
• don't ask two questions at once

• dichotomous - only 2 answers (yes or no)
• multiple choice - 3 + options 
• numeric - asks for single number (e.g. distance to work)
• rank order - asks participants to rank in order (e.g. preference)
• likert scale - asks participants how much they agree with statement
• personal question - e.g. salary or age. best to use range

• screens - identifies early on whether participant should undertake questionnaire
• ice breaker/warm up - easy to answer, generates interest in questionnaire
• transition question - allows questions to flow and link easily
• skips - e.g. if yes answer Q3
• difficult - should be positioned in the middle when participant is in the flow (online = progress bar)
• categorical - often personal questions so should be left till last to avoid participant feeling uncomfortable

ADV

• practical
• large amounts of info can be collected in short amount of time at low cost
• can be carried out by researcher or any number of people without infulencing validity/reliability
• results quickly/easily quantifitable by researcher or software
• allows for comparison

DISADV

• inadequate to understand some forms of info
• no way to tell how truthful respondent is being
• no way to tell how much time they have put into their answer and if they understood context

Validity

• questions understood/misunterpretated?
• undertake pilot study to check they understand questions in the same way
• construct validity is it measuring what you think it is measuring?
• internal validity does it cover key points & balanced viewpoint?

Reliability

• does data collection techniques provide consistent findings?
• transparent reporting?
• repeated by someone else same results?
• repeated by yourself again on another day same results?

• variables - characteristic of sample that is of interest to research. attribute describing person, place or thing. varies between entities
• data - actual value of variables, needs to be processed to be extracted and understood

continuous - falls on continuum can have fractions or decimals

discrete - finite number of possible values

• nominal - categorised but no intrinsic order. No measure of distance between values e.g. eye colour
• ordinal - categories can be ranked and ordered. Distances between values may be different e.g. education
• interval - orders values, ability to quantify difference between each one. e.g. temp
• ratio - order, interval values, calculate ratios due to their being a true & neutral 0. 

• dependant - measured in the study. If A influences B then B is independant variable?
• independant - changed/controlled to test dependant variable?
• control - held constant, not of primary interest
• moderator - qual/quant affects direction/strength of relationship between independant/dependant variable