QRM

Quantitative research

measuring things that can be counted using predetermined categories that can be treated as interval or ordinal data —> used for statisical analysis

essential for the overview = using numbers we can describe trends

Qualitative research

focuses on peoples experiences and the meaning they place on events, processes and structures of their normal social setting

gives detail = emotional, non rational, biases

positivism

assumes reality is objectively observable

interpretivism

assumes that reality can only be made accessible through social constructions

Ontology

the nature of the social world and what can be known about it

3 positions:

Realsim = a would exists outside of us, Reality is real, but our views

may differ

Materialism = Only physical things, if its not physical, its not really real

Idealism = Reality only exists through our minds

Epistemology

the study of knowledge

2 ways to obtain knowledge

Induction = Evidence is used to establish a conclusion (specific idea

→ general conclusion)

Deduction: Evidence is used to support a conclusion (general idea

→ specific conclusion)

Functions of qualitative research

1. Contextual (concept): to describe the form or nature of something that exists

2. Explanatory (explanations); to examine the reasons for or associations between something that exists

3. Evaluative (does it work): assessing the effectiveness of something that exists

4. Generative (new ideas): supporting the development of theories, settings, or actions

Naturally ocurring data

data that exists in the real world, not created by the researcher

Generated data

data created through interaction between researcher and participants

Snapshot (single research episode) studies

studies that involve only one episode of field work

Longitudanal qualitative research studies

more than one episode of data collection.

2 broad forms:

1. Panel studies: same people are interviewed more than once over time —> micro-level change

2. Repeat sross sectionals studies: same questions, same type of survey, run again and again over time — but DIFFERENT people each time. —> macro level change

Population

the set of elements about which a researcher wants to make statements

probablitity sampling

sample elements are chosen randomly

non-probability sampling

sample elements are deliberately chosen so they represent certain characteristics

statistical representation

choosing sample elements that are representative of the population

symbolic representation

a sample represents and symbolizes the relevant characteristics

4 ways Non probability sampling can be conducted

purposive sampling

theoretical sampling

convienience sampling

opportunistic sampling

purposive sampling

selecting a criteria based on a particular criterion

2 goals:

Make sure every relevant criterion is covered

Make sure there is diversity so the impact of one characteristic can be examined

A variety of approaches:

● Homogeneous sample: cases have the same characteristics → get a detailed picture of a particular

phenomenon

● Heterogeneous sample: cases vary greatly from one another → identify central themes that apply to

different cases

● Extreme case sample: cases are unusual or special → learn about a phenomenon by looking for exceptions

● Intensity sample: ases are a good representation of the phenomenon of interest → learn about a

phenomenon by looking at extreme cases of the phenomenon

● Typical case sample: cases have characteristics that fall into the “average” or “normal” category → detailed

profiling

● Stratified purposive sample: groups are selected that are homogenous to each other but themselves are

heterogeneous → compare subgroups

● Critical case sample: cases demonstrate a phenomenon dramatically or are critical in the delivery of a process

→ critical to the understanding of the phenomenon

theoretical sampling

a specific type of purposive sampling in which a researcher selects sample elements based on their potential contribution to the development of theoretical constructs

The process is iterative: the researcher selects an initials ample, analyses the data, and selects a subsequent sample to expand their findings

This process is repeated until the researcher reaches theoretical saturation: no new insights can be gained from a new sample, nothing more can be added to the theory

opportunistic sampling

involves the researcher taking advantage of opportunities that arise during the fieldwork The researcher takes a flexible approach and forms the sample based on the context of the field work

Convenience sampling

involves the researcher choosing the sample based on ease of access

● Sample elements that are within easy reach are chosen

● Does not include a clear sampling strategy

Issues to consider when choosing sample size

1. Population heterogeneity

2. The number of selection criteria

3. The extent to which criteria must be nested

○ Nesting: controlling the representation of one criterion with another

4. Special groups to be intensively studied

5. Multiple samples in one study

6. Type of data collection method

7. Budget and resources

parent population

the population from which the sample is to be drawn

multistage design

the study population is located within a collective organizational unit

two types of sampling frame

1. existing sources

• Administrative records

• published lists

• survey samples

2. Generate a new sample

• Household screen = approach households and interview briefly to find out if it contains an individual who is part of the target group

• Use organizations that provide services to the population or represent the population

• Snowball sampling = you find a few people who fit what you need, and ask THEM to refer you to more people like them. Those new people refer you to even more. It grows like a snowball rolling downhill

• Flow populations: individuals approach at a specific location

unstructured data collection

involves a broad agenda which maps the issues to be explored across the sample, but the order, wording, and way in which they are followed up will vary considerably between interviews

semi-structured data collection

where the interviewer asks key questions in the same way each time and does some probing for further information, but this probing is more limited than in unstructured data collection

Interview guide(aide-memoire)

a tool to improve the consistency of data collection

techniques to help participants express themselves:

1. Enabling techniques = use visuals or descriptions to help people talk

• Great for making complex or hidden feelings come out

• Good for when people don’t know how to explain what they feel

2. Projective techniques = use a stimulus, and the person projects their feeling onto it

• best for sensitive or unconscious topics

Enabling and projective techniques are used to overcome 4 barriers

1. Knowledge barrier: person doesn’t know how to explain something (it’s unconscious)

2. Rationality barrier: person gives a logical answer instead of an emotional one

3. Inaccessibility barrier: person doesn’t want to say something

4. Politeness barrier: the respondent gives a socially desirable answer

Four enabling techniques

1. Vignette: brief description of a particular circumstance, person, or event → paints a picture of the moment or person → basis for discussion

2. Card sorting: respondent is asked to sort through a number of written or visual examples of issues

3. Giving information or showing written material: when the discussion is not going well or when respondents don’t know much about the topic

4. Mapping emergent issues: noticing the emergent issues and showing them to the group

5 projective techniques

1. Association techniques (word association and brand personality)

2. Completion procedures (sentence completion and brand mapping)

3. Construction techniques (indirect questions and fill in speech bubbles)

4. Expressive methods (drawing and role playing)

5. Choice-ordering (q-sort)

Benefits of projective techniques:

● Get past defense mechanisms

● Gain access to information

● Entice respondents to make thoughtless, unrationalized responses → elicitation technique

Disadvantages of projective techniques

● Takes time

● Interrupt the discussion

● Can easily be misunderstood

● Group members may have literacy or vision problems

structured interview

pre formulated questions very few open questions, minimal role of the interviewer

advantage: consistency across interviews

Disadvantages:

interviewer cant talk freely

New, unforseen aspects aare overlooked

Unstructured interviews

open ended questions, very few pre formulated questions

Advantages:

Interviewee can talk freely

new unforeseen aspects can emerge

Disadvantages:

Interviewee not talkative = too little info

Interviewee very talkative = too much info

Hard to maintain consistency

Semi structured interviews

predetermined open ended questions, some pre formulated questions, some control over direction and content of discussion

Advantages:

Best of structured/unstructured interviews

minimize risks, cost efficient, easy(ish)

Disadvantages:

limited ability of participant to recall

Limited ability of researchers to ask right questions

Interviewees role

to give fulsome answers, to provide more depth when probing questions are asked, to reflect and to think, to raise issues they see as relevant, but which are not directly asked about

Interviewers role

That of a facilitator, to enable the interviewee to talk about their feelings, views, and experiences

2 Interviewer roles

Task oriented skills: organize the interview to reach your research goal (reliable data for research)

Relation oriented skills: create an atmosphere where the interviewee is willing o provide you with valid/relevant data

task oriented interviewer advantages/disadvantages

Advantages:

Keeps the interview structured and on-topic

Makes data comparable across interviews

Disadvantages:

Risk of feeling like an interrogation if overdone

Can shut down spontaneous, unplanned insights

relation oriented interviewer advantages/disadvantages

Advantage	Disadvantage
Builds trust → interviewee opens up, gives richer/more honest answers	Risk of losing structure — conversation drifts off-topic
Surfaces unexpected, valuable information you didn't plan to ask about	Harder to keep data consistent/comparable across interviews

2 steps to write a good interview guide

1. Topic based guide: consists of a list of areas and issues the interviewer wants to hear about

2. Question based guide: outlines expected content of the interview in terms of a series of questions the interviewer intends to ask

Interview Guide Format

Interview guide: A. Introduction

The guide must include an introduction:

● Background information such as:

○ Info about yourself

○ Objective of the interview, why the interview

○ Why the participant was selected and how you were referred to them

○ The institution / organization responsible for the interview

○ Recording permission and plans to use GenAI → does the interviewer agree

● Results:

○ How the interview will be processed into a final report & who are the report’s readers

● Factual information about the interview

○ Type of information you want

○ Amount of time to conduct the interview

○ What will be done with the data & the recordings

○ Ethical concerns → informed consent, confidentiality, anonymity

Interview guide: B. Topics and subtopics

★ Research questions ≠ interview questions

● You select 3 topics → for each topic:

○ give an opening question (content mapping questions)

○ identify 2 - 4 subtopics (content mining questions)Interview guide:

C. Closure

● Introduce termination of the interview

● Express your gratitude

● Repeat possible arrangements for a follow-up (eg. sending transcript, copy of your final report)

● Ask for the interviewee’s impression of the interview (optional)

Content mapping questions

questions that capture the breadth of the topic

types of content mapping questions

1. Ground mapping questions: first questions asked to open a topic —> broad questions

2. Dimension mapping questions: get respondent to focus more narrowly —> structure and guide

3. Perspective widening questions: allow interviewer to broaden respondents’ perspective —> encourage further thought

Content mining questions

designed to discover details within each dimension

types of content mining questions

● Amplification probes: encourage respondents to go deeper into something

● Exploratory probes: aimed at discovering underlying feelings and thoughts for descriptions, behaviors, events, or experiences → ask about consequences and effects of behavior

● Explanatory probe: go deeper into the “why” behind things

● Clarificatory probes: search for more clarity by:

○ Ask for clarification of terms, language, details, consequences

○ Testing a POV

○ Continue to ask questions when there are inconsistencies

probes

follow up questions designed to reveal more information or explanation

Open question

A question that allows a full, detailed answer, often starts with what, where, how, who, or why.

Closed question

A question that can be answered with only yes or no

choice question

A question that asks the respondent to choose between two or more options

Neutral question

A question that does not suggest any answer or opinion, it invites an unbiased response

Leading/suggestive question

A question that pushes the respondent toward a certain answer, often showing the interviewers bias

Rhetorical question

A question that does not expect an answer because the answer is obvious or implied

Advantages and disadvantages of question types

	Intended effect	Unintended effect
Open	Detailed / thoughtful responses → deeper insights / exploration of opinions & experiences	Too broad or vague → off-topic answers / confusion about how to respond.
Closed	Quick & direct way to confirm facts / clarify a specific point.	Limits depth & does not encourage elaboration, → miss important insights.
Choice	Focus the response by narrowing down options → easier to compare perspectives	Restrictive if provided choices don’t capture interviewee’s true opinion / response falls out of given options
Neutral	Unbiased and open responses without directing the participant toward a particular answer.	If not clear, the respondent may still interpret an implicit bias or struggle to articulate their answer.
Leading / suggestive	Subtly influences the respondent toward a certain response (useful if trying to confirm hypothesis)	Can introduce bias → answer less authentic → potentially skewing research results.
Rhetorical	Reinforces a point or assumption rather than genuinely seeking information.	Can be dismissive → respondent feel like their answer is already predetermined or irrelevant.

4 criteria for evaluating an interview answer

1. Valid: is what the interviewee thinks and says the same

2. Complete: checking if the answer contains all the information there could possibly be

3. Relevant: the answer provides the information the interviewer would like to elicit

4. Clear: are you sure you know what the interviewee means?

if not met = probe

focus group

researcher led group discussion as a form of data collecting

Advantages: Direct access to real opinions, attitudes, beliefs

Disadvantages:

Time consuming and expensive

Som participants too dominant

Point of attention: interviewer moderating style

Delphi method

The Delphi Method is a structured way to gather expert opinions. It involves a step-by-step process:

1. The researcher sends a question to a group of experts.

2. Experts give their individual responses.

3. The researcher summarizes all the answers and sends that summary back to the group.

4. Experts then react to the summary, possibly changing their views.

5. This cycle repeats until there is consensus or no new ideas appear (saturation)

Group processes and stages of a focus group

1. Forming: where the group comes together and there is uncertainty and chaos

2. Storming: where the roles are divided, and the goal is clear

3. Norming: where the standard is set, and the roles are clear

4. Performing: where the task is performed

5. Adjourning / mourning: where the group breaks up

Observation

the act of noticing a phenomenon and recording it for scientific purpose

Types of observational research

1. Complete observer: researcher is completely detached from the setting

○ Not visible and cannot be noticed

2. Observer-as-participant: researcher observes for short periods of time

○ Researcher is known and recognized but perceived as a researcher

3. Participant-as-observer: researcher is fully integrated into the setting/part of the process

○ Respondent and neutral researcher, activities are recognized by other respondents, participates and observes

4. Complete participant: researcher completely disappears into the setting and is fully engaged with the respondents and their activities (going native)

○ Research is not / less concerned with his research agenda

Types of membership

● Peripheral membership: researcher observes and interacts directly with the respondents

○ Does not participate in the activities

● Active membership: researcher participates in key activities, but refrains from committing ot values, goals,

and attitudes

● Complete membership: researcher is an active and engaged member of the group and adopts the views of the group

To ensure Validity

1. Use multiple observers or teams

2. Analytical induction can be employed (search for negative cases → if no negative cases then propositions can be assumed to be universal)

3. Verisimilitude techniques: writing strategies used to make findings feel real, vivid, and believable, as if the reader was there themselves.

The quality of conclusions can be determined by assessing

1. The objectivity/confirmability: the degree to which conclusions flow from the information that has been collected, and not from any biases on the part of the researcher.

2. Reliability/dependability/auditability: degree to which the process of research has been consistent and reasonably stable overtime and across various researchers and methods

3. Internal validity/credibility/authenticity/truth value: degree to which the conclusions of a study makes

sense, if they are credible to the people studied as well as to the readers of the report, and if the final product is an authentic record of whatever it was it was observed.

4. External validity/transferability/ fittingness: degree to which the conclusions of the study have relevance to matters beyond the study itself.

5. Utilization/application/action orientation/pragmatic validity: degree to which programs or actions result from a study's findings and or the degree to which ethical issues are forthrightly dealt with.

Minimizing observer bias

1. Ensure observation is as natural as possible

2. Make sure the research is emergent, that there is room for creativity

3. Triangulation: process of combining observational research with other techniques for the collection of information

Qualitative data analysis approaches

Ethnographic accounts

Life histories

Narrative analysis

conversation analysis

discourse analysis

Analytic induction

Grounded theory

Policy and evaluation analysis

ethnographic accounts

detailed descriptions of peoples lives, groups, or organizaions, often based on long term observation

life histories

use personal stories/narratives, either as standalone cases or grouped by theme, to understand experiences over time

Narrative analysis

focuses on how a story is told, including plot, purpose, and audience - not just content

conversation analysis

analyses structure and flow of conversations, like turn taking and paired responses

Discourse analysis

looks at how language produces knowledge and meaning within a particular context/discourse

Analytic induction

tries to build universal explanations by constantly comparing cases, updating the hypothesis untul all cases fit

Grounded theory

aims to develop theory directly from data, using codes, categories, and the constant comparative model

Policy and evaluation analysis

focuses on understanding how social programs work, their impact, and context —> guided decision making

Computer assisted qualitative methods

1. Text retrievers: search large volumes of text for specific words/phrases

2. textbase managers: data management systems that structure your data and make it searchable

3. Code and retrieve programs: label/tag passages of text that can later be retrieved according to codes applied

4. Code based theory builders: help build concepts and often include tools to create hyperlinls or connect ideas within your data

5. conceptual network builders: allow visual mapping of concpets, relationships of meanings

Three stages in analysing text

1. Data management

• Raw data, meaning is assigned to things, data is labeled/sorted/brought together

• themes and concepts identified

2. Descriptive accounts

• During this phase the aggregated data to identify key dimensions is used

• codes can be merged, then a descriptive account is linked to it

3. Explanatory accounts

• looks for patterns of association and relationships within data, examines why these associations exist

2 approaches to text(content analysis)

1. Quantification of qualitative data (e.g. words, gestures,) for the purpose of statistical inference —> counting entities such as codes words phrases — turnign words into numbers — categorized based on how often a category is mentioned

2. Grounded theory / the constant comparative method

two coding strategies

A. A-priori coding — deductive; you start with predefined themes/concepts (e.g. taste, packaging) and find codes in the text that support them, though new unexpected concepts can still emerge along the way.

B. Grounded theory — inductive; you start with no predefined concepts at all, and all themes emerge purely from coding the data itself (this is grounded theory).

2 coding procedures

1. Theoretical coding

2. Thematic coding

Theoretical coding

the process of coding and categorizing qualitative data with the goal of developing a theory, in an emergent manner

—> 3 procedures open, axial, selective coding

Atlas ti

Software used to execute grounded theory/text analysis (coding, building categories, visualizing relationships).

Advantage: user-friendly with handy visualization techniques (tree-like diagrams of codes/categories).

Disadvantage: those tree-like visualizations become unclear/unreadable once you have large texts or many codes.

Descriptive accounts

Organizing data so it's conceptually clear and meaningful,

1. detection (identify the content/dimensions of a phenomenon),

2. categorization (refine categories, assign data to them),

3. classification (group categories into higher-level classes).

Explanatory accounts

about detecting patterns, doing associative analysis, and identifying clustering.

Associative analysis: a lucrative form of qualitative data investigation, as it nearly always brings deeper understanding.

2 Main Ways to Analyze Group Data

1. Whole group analysis: treats the group as a single unit.

2. Participant-based analysis: focuses on individual contributions within the group context.

Conceptual model

Visual representation of a theory

• looks at predictions (propositions/hypothesis)

• only attributes and cause-effect

Case study research (CSR)

Research technique aimed at examining a contemporary phenomenon in its natural setting

Types of case study research

1. Descriptive case study: to describe the incidence or prevalence of a phenomenon of interest

2. Explanatory case study: to trace causal linkages among actions, decisions, and events over time

3. Exploratory case study: to develop pertinent hypothesis and propositions for further inquiry

How do I know if something is a case study?

Test: can you describe the phenomenon without describing the specific setting it happens in? If stripping away the context makes the phenomenon meaningless (not just "harder to generalize"), the phenomenon and context are fused → it's a case study. If the phenomenon shows up across many contexts and you'd sample across them, it's not.

Single vs. multiple case design

Single = one case studied. Multiple = several similar cases studied for comparison/replication.

Holistic vs. embedded case design

Holistic = the case studied as one whole unit. Embedded = the case broken into multiple nested sub-units you compare within it.

The 4 case study design combinations

Holistic + single: "Why did the new method succeed with company X?"

Holistic + multiple: "Why did the new method fail with companies Y1, Y2, Y3?"

Embedded + single: "How does the new method work in department 1 and 2 of company X?"

Embedded + multiple: "How does the new method work in department 1 and 2 across companies Y1, Y2, Y3?"

Reasons to choose single-case design

Critical case	The case is decisive for testing a well-formed theory — if the theory holds (or fails) here, it confirms, challenges, or extends that theory in a meaningful way
Extreme/unique case	The case is rare or unusual, deviating sharply from the norm — worth studying precisely because it's atypical
Representative/typical case	The case captures everyday, ordinary circumstances — chosen because it's a good stand-in for the common situation, not an outlier
Revelatory case	The researcher gets access to a phenomenon that was previously inaccessible or unobserved — the value is in finally being able to study something nobody could examine closely before
Longitudinal case	The same single case is studied at two or more points in time, to see how it changes

Reasons to choose multiple-case design

Stronger evidence — pilot case

pilot case

A trial case selected for being very accessible, convenient, unusually information-rich, or the most complicated of all — run first to clarify concepts, pretest/refine the research plan and questions, narrow the research focus, and work out methodological issues before the main data collection.

Closed vs. flexible case study design

Closed = design stays fixed throughout active data collection. Flexible = design can adjust mid-study based on emerging info — though swapping cases or changing theory is a bigger move than just tweaking setup.

5 case study analysis techniques

1. Pattern matching — compare your predicted outcome pattern against what you actually found; with one case, you check it across multiple DVs.

2. Explanation building — build a how/why explanation, refining it case by case, comparing as you go.

3. Time-series analysis — track trends, structural changes, and chronological order over time.

4. Logic models — map a deliberate cause→effect→cause→effect chain across time (e.g.more pc-tutorials and assignments (intervention) leads to more cooperation with fellow students (immediate outcome) lead to increased understanding of the topic (intermediate outcome) leads to higher exam grades (final outcome).

5. Cross-case synthesis — used with multiple cases; check if similar conditions give similar results, and opposite conditions give opposite results.

Action research

the researcher deliberately intervenes in a situation while simultaneously studying the effect of that intervention — it's an interactive inquiry process balancing collaborative problem-solving with data-driven analysis, aimed at understanding underlying causes and predicting future outcomes.

One-liner: "You don't just observe the problem — you try to fix it, and study what happens when you do."

5 phases of action research

1. Diagnosing: identification of problem(s) and issue(s) related to phenomenon of interest

2. Action planning: development of an action plan to specify activities for resolving identified problem(s)

3. Action taking: implementation of planned actions

4. Evaluating: assessment of actions taken and their impact on identified problem(s)

5. Specifying learning: consolidation of knowledge from preceding phases to determine future courses of action