Advanced Empirical Methods Flashcards
Features of the Scientific Method
Observations -> Questions -> Search Literature -> Hypothesis -> Experiment -> Collect Data -> Conclusions -> Share Results -> Develop Interventions -> Ask New Questions
Characteristic of a Research
v Empirical: It means that any conclusions drawn are based
upon hard evidence gathered from information collected from
real-life experiences or observations.
v Logical: research is based on valid procedures and principles.
v Systematic: this implies that the procedures adopted to
undertake an investigation follow a certain logical sequence.
The different steps cannot be taken in a haphazard way. Some procedures must follow others.
v Replicable: research design and procedures are repeated to
enable the researcher to arrive at valid and conclusive results.
v Valid and Verifiable: this implies that whatever you
conclude based on your findings is correct and can be verified
by you and others.
v Critical: research exhibits careful and precise judgment.
Research - Definition
Systematic attempt, using socially approved methods to extend our knowledge and understanding of the world
Social Research
- Human behaviour/ Interpersonal Psychological Research with people at both ends
- Social research attempts to identify, explore, describe, understand, explain, evaluate, and predict social phenomena
involving human behaviour.
Social! - Research! - Methods!
- Social: Human behaviour and experiences, Social Realities
- Research: Finding the Truth; Understanding Facts; Discovering provable Evidence about Social Reality with the assumption that there are Facts/Truths/Evidences out there. Here we mean Scientific Research
- Methods: To have access to the Truth/Facts/Evidence, Methods are used as Procedure/Mechanism/Technique/Medium for communication to systematically get the facts based on evidence. (Statistics; research design; data collection; questionnaire; testing; systematic observation; content analysis) They are called systematic methods.
A study focused on the design and implementation of specific technological products and processes to improve the learning, instruction, and management aspects of education.
Who are your potential target groups?
What variables do you need to create to
respond to the Topic?
What methods would you use?
-
Potential Target Groups
- Students: The primary audience benefiting from improved learning processes and tools.
- Teachers: Individuals who will use the technology to enhance instruction, track progress, and engage students.
- School Administrators: Responsible for managing educational operations and leveraging data insights.
- Parents: To monitor their children’s progress and facilitate better communication with educators.
- Educational Technology Developers and Researchers: To adapt and innovate technology solutions tailored to educational needs.
-
Variables to Create
To address the topic, you may need to create variables such as:
- Student Performance Metrics: Test scores, attendance, participation, and engagement levels.
- Teacher Interaction Metrics: Frequency of technology usage, lesson effectiveness, and feedback quality.
- Technology Adoption Rates: Percentage of target groups using the tools regularly.
- User Satisfaction: Surveys assessing ease of use, satisfaction, and perceived effectiveness.
- Cost-Effectiveness: Comparison of implementation costs versus measurable benefits.
-
Methods to Use
- Surveys and Questionnaires: To gather feedback from students, teachers, and administrators on needs and satisfaction levels.
- Pilot Programs: Implement the technological products in a small group to test their effectiveness and scalability.
- Experimental Design: Compare outcomes between groups using the technology and control groups not using it.
- Data Analytics: Analyze performance metrics and user behavior to assess the impact of the technology.
- Case Studies: Document in-depth examples of successful implementation and the challenges faced.
- Focus Groups: Engage with representatives from target groups to refine the design and address concerns.
Research Design Paradigms and Methods
v Data collection and analysis to address the ‘‘how’’ question.
v We intend to put the ‘‘what’’, ‘‘who’’, and ‘‘how’’ pieces together.
v To conceptualize a useful and feasible empirical study
v To communicate this complex, multifaceted information to the consumers of research
EMPIRICAL RESEARCH
- In its broadest definition, empirical research is a systematic and thoughtful process of planning and implementing observations (McMillan and Schumacher 2001).
- Any sort of deliberate empirical planning involves a careful analysis of what to observe, “whom” to observe, and “how” to observe.
- In the context of educational technology research, the ‘‘what’’ question is typically addressed.
Main Elements of Social Research
- LITERATURE REVIEW
- RESEARCH QUESTION
- THEORY AND CONCEPT
- DATA
- SAMPLING
- ANALYSIS
- WRITING UP
Literature Review
- What is known about the topic?
- What concept and theory have been applied to the topic?
- What research method have been applied?
- What controversies are discussed about the topic and how is it studied?
- What clashes of the evidence exist?
- What is the key contribution to research on the topic?
Research Process
- Phenomenon from
social reality - Task: commercial,
self-imposed - Research question
- Supporting the question by the appropriate theory,
presenting the object of research - Defining the main concepts
- Building the hypothesis
- Research concept: selecting method, measurement, equipment
- Setting up indicators
- Data collection
- Data analysis
- Presenting the results
- Report, publication
RESEARCH PROCESS SIMPLIFIED
- Define the topic
- Narrow the topic
- Gather information
- Create RQ
- Find and cite sources
- Write the paper
Research Question
- Central question
- Prediction of an outcome
- Explaining causes
- Evaluation
Questions to consider when developing the RQ
- WHO
Specific group, gender, sex; age, ethnicity; key figures; socioeconomic status - TO WHAT EXTENT
what are the issues; are there any unanswered questions; are these sub-topics - WHEN
is this a current issue; is it related to a specific [eriod of time; was there an event relayed to this issue - WHERE
Can you narrow to a specific geograpchic location or boundary? - WHY
Why is this issue interesting?
Why should others be interested?
Research Question and Hypothesis:
Important:
Topic: Cultural Influence in the recruitment of students or applicants at Oxford and Cambridge universities. (Zimdars et al. 2009: p. 653.)
Aim: The aim is to inquire or assess whether culture is linked to successful admission or whether there are class prejudices or biases in the admission process/or constrain the life chances of young people from less privileged backgrounds.
PLEASE develop A Research hypothesis from this Topic and Aim. (to be discussed in class).
RQ. 1.: How do Oxford applicants vary in their cultural participation and cultural knowledge, according to parents’ education, social class, gender, and ethnicity?
RQ. 2.: To what extent does cultural capital mediate the effect of social class, parents’ education, private schooling, ethnicity, and gender?
6 components of a Strong Hypothesis
- Empirically testable
- Backed by preliminary evidence
- Testable by ethical research
- Based on original ideas
- Has evidence-based logical reasoning
- Can be predicted
Research Triangle
Clear research questions grounded in Relevant theory informs methodology to generate Appropriate methods to answer Research questions
Research cycle
- Observation -Create or modify the theory- 2. Theory -use to form a hypothesis- 3. Hypothesis -design a study to test a hypothesis- 4. Research -perform the research-
Artificial Intelligence
Ø Natural Science is knowledge about natural objects and phenomena. He interrogates if there cannot be ‘artificial science’ – knowledge about artificial objects and phenomena (p.3.). Today, artificial as a term has a negative connotation. Artificial is defined as “produced by art rather than by nature; not genuine or natural (p. 4.) among others. Artificial - Art; engineering science; computer science – philosophy; political and mathematical science.
Ø John McCarthy - One of the greatest innovators in the field of AI was John McCarthy who got the title of Father of Artificial Intelligence for his contribution to the field of Computer Science and AI. ((1927–2011), an American computer scientist and cognitive scientist; Alan Turing; a British Mathematician)
What is a Conceptual Framework?
A conceptual framework illustrates the expected relationship between your variables. It defines the relevant objectives for your research process and maps out how they come together to draw coherent
conclusions.
* Conceptual frameworks are often represented in a visual format and illustrate cause-and-effect relationships.
Functions of Conceptual Framework
- Justify Research Problem
- Define relevant concepts
- Establish theoretical and empirical rationale
- Select Appropriate Methods
- Interpret Results Relative to theory
What is a Theory?
- Qualitative studies often use specific theories which works somewhat like spectacles, metaphorically. It contributes to qualitative studies alongside the field work and other aspects of the investigators.
- A theory is an attempt to explain why to provide understanding.
Literary Theory
§ Provides rationale
§ Framework within the social phenomena applied to interpret the findings
§ Is an explanation of observed regulatories
Methods
The methods are meant to explain and assess the theoretical findings and considerations in order to achieve valid and verifiable results.
What is a Theoretical Framework?
Structure that can hold or support a theory of a research study
* A theoretical framework consists of concepts and, together with their definitions and reference to relevant scholarly literature, existing theory that is used for your particular study.
* The theoretical framework demonstrates an understanding of theories and concepts that are relevant to the topic of your research.
* The theoretical framework is often not something readily found within the literature. You must review course readings and pertinent research studies for theories and analytic models that are relevant to the research problem you are investigating.
- An explicit statement of theoretical assumptions allows the reader to evaluate their topic critically.
- The theoretical framework connects the researcher to existing knowledge. Guided by a relevant theory, you are given a basis for your hypotheses and choice of research methods.
- Articulating the theoretical assumptions of a research study forces you to address questions of why and how. It permits you to intellectually transition from simply describing a phenomenon you have observed to generalizing about various aspects of that phenomenon.
- Having a theory helps you identify the limits to those generalizations. A theoretical framework specifies which key variables influence a phenomenon of interest and highlights the need to examine how those key variables might differ and under what circumstances.
*You want to research the following topics
*The Influence of Artificial Intelligence on Media Culture
*Media and Information Literacy can combat Misinformation and Disinformation
*Commercial advantages of sensationalizing health topics on YouTube
*What kind of theories could you use to study them?
*Name the theory and explain why it is useful for studying the topic
- Technological Determinism Theory
Founder: Marshall McLuhan
Key Idea: “The medium is the message” — technology, especially communication tools, determines how society is structured and how culture evolves. - Agenda-Setting Theory
Founders: Maxwell McCombs and Donald Shaw
Key Idea: The media doesn’t tell people what to think but what to think about by prioritizing certain topics, thus influencing public opinion. - Uses and Gratifications Theory
Founders: Elihu Katz, Jay G. Blumler, and Michael Gurevitch
Key Idea: Audiences are active participants who use media to satisfy specific needs, such as information, entertainment, or social interaction. - Cultivation Theory
Founders: George Gerbner and Larry Gross
Key Idea: Long-term exposure to media content, especially television, shapes perceptions of reality, often aligning them with the media’s portrayal. - Social Learning Theory
Founder: Albert Bandura
Key Idea: People learn behaviors by observing and imitating others, particularly role models in media, and this is reinforced by rewards or punishments.
The “Two-Step Flow of Communication” concept, developed by sociologists Paul Lazarsfeld, Bernard Berelson, and Hazel Gaudet, emerged from their study of voter behavior during a U.S. presidential campaign, as detailed in their book “The People’s Choice (1944)”. Step 1 - media influences opinion leaders, Step 2 - Opinion leaders influence their social groups by sharing and interpreting information.
The key principle of the Two-Step Flow of Communication relevant to this research is the role of social interaction in spreading information, as people often trust information shared through personal communication more than directly from media.
Instagram functions as a platform where modern-day opinion leaders—such as influencers, activists, and content creators—curate and interpret content related to the conflict. In the first step of the communication flow, these opinion leaders consume and filter information from traditional media or firsthand sources. In the second step, they share this information with their followers, often framing it in ways that resonate with their audience’s values, emotions, and worldviews. By enabling social interaction through comments, likes, and shares, Instagram fosters a participatory environment where information about the conflict is disseminated, interpreted, and reinforced within social networks, ultimately shaping public perceptions in Germany.
Data Collection Methods
§ An experiment requires a specific research design. This makes it possible to identify cause-effect relations.
§ By conducting a survey, we collect mostly attitudes and opinions among a population.
§ Via content analysis, we can systematically describe content.
§ By using the observation method, we can determine people’s actual behaviour.
§ In the field of Communication, researchers use primarily surveys and content analyses.
Data Collection Methods
- Forms and Questionnaries
- Interview
- Observation
- Documents and Records
- Focus Groups
- Oral Histories
- Combination Research
- Online Tracking
- Online Marketing Analytics
- Social Media Monitoring
Data Analysis Process
- Define the question
- Collect the data
- Clean the data
- Analyze the data
- Visulize and share findings
Tree Density exercise
- Evaluate the Graphic
The graphic shows the probability of antidepressant prescriptions based on tree density in a person’s living area.
It highlights three levels: low, medium, and high tree density, with corresponding probabilities of 7 in 100, 6 in 100, and 5 in 100, respectively. - What Topic Is Studied?
The relationship between tree density in residential areas and the likelihood of receiving antidepressant prescriptions.
It explores whether living in greener areas (with more trees) reduces the need for antidepressants. - How Are Cause-Effect Relations Visualized?
Tree density is shown as the independent variable (cause), while antidepressant prescription rates are the dependent variable (effect).
The relationship suggests that higher tree density correlates with fewer antidepressant prescriptions. - Is Anything Missing in This Data Analysis?
The analysis does not account for confounding variables, such as:
Socioeconomic factors (e.g., income or access to healthcare).
Urban vs. rural settings (tree density might differ by location).
Other environmental or lifestyle factors affecting mental health.
It does not explain why tree density might influence antidepressant use (e.g., stress reduction or exercise opportunities). - Can You See Any Bias?
Sampling bias: The graphic doesn’t specify how the data was collected or whether the sample is representative of the broader population.
Interpretation bias: The graphic implies a causal relationship, but the data might only show correlation, not causation.
Suggested Answer
This graphic visually represents the correlation between tree density and antidepressant use but lacks context about other influencing factors. Without additional data on confounding variables and research methods, it’s difficult to draw firm conclusions about causation. The simplicity of the data might oversimplify a complex relationship, potentially leading to bias in interpretation.
Ukraine example
Here’s how you can respond to this task:
1. Evaluate the Graphic
- The graphic illustrates the scale of Ukrainian refugees, showing that 1 million people have already fled Ukraine, with 4 million more expected to flee, representing 11% of the country’s population.
- The visual uses people and blocks to simplify and quantify the data.
2. What Topic Is Studied?
- The graphic focuses on the scale of displacement of Ukrainian citizens due to war, highlighting the proportion of the population affected by the crisis.
3. Is Anything Missing in This Data Analysis?
- The graphic doesn’t provide:
- Context about timeframe (e.g., over what period these displacements occurred).
- Details about where refugees are going or their circumstances after fleeing.
- Demographics of the displaced population (e.g., age, gender, socio-economic status).
- Underlying causes of displacement, such as war zones or other factors.
- It oversimplifies a complex situation without explaining the full impact of the crisis (e.g., humanitarian challenges or international response).
4. Can You See Any Bias?
- Visual bias: The use of human figures might evoke an emotional response, potentially influencing viewers’ perception of the scale.
- Framing bias: The graphic emphasizes a significant number without providing broader context (e.g., how this compares to other refugee crises or Ukraine’s total population size).
- Selection bias: It focuses only on those fleeing, not accounting for internally displaced people or those unable to leave.
Suggested Answer:
The graphic effectively visualizes the scale of displacement but lacks depth. It omits context about the causes, demographics, and global response, which are critical to understanding the full picture. While the use of human figures is impactful, it risks oversimplifying the complexity of the refugee crisis, potentially leading to emotional or skewed interpretations.
Methods for Communication Research
1. Empirical
1.1. Measuring type
and analysis
а) Qualitative
б) Quantitative
1.2. Methods for data collection
Qualitative:
а) Survey
б) Content analysis
Quantitative:
а) Observation
б) Physiological measurement
2. Non-empirical
2.1. Research Designs
Experimental: Theology, Hermeneutics
Nonexperimental
Anatomy of scientific paper
- Introduction
1.1. What is unknown?
1.2. How and why should we fill the gap? - Methods
- Results
- Discussion
- Conclusion
Theory in research
Theory in research follows (the deductive approach) or arises from (the inductive approach) of data collection
The Research paradigm
- Ontology: the nature of reality and of what really exists
- Epistemology: the relationship between the inquirer and what is known
- Axiology: what we value: the ultimate worth of research
- Methodology: the strategy and justifications in constructiong a specific type of knowledge
Conceptualising Research Journey: Iceberg
- Philosophical Foundations: ontology, axiology, epistemology
- Theory: theoretical stance
- Methodology: systematization
- Method: techniques used
QUALITATIVE RESEARCH METHODS
Qualitative research is a scientific method used to gather non-numerical data.
Rather than focusing on measurements or metrics, it seeks to understand concepts, experiences, or phenomena by exploring participant perspectives.
This method is often employed in fields like psychology, sociology, media/communication, and marketing to gain deeper insights into consumer behaviour, motivations, and cultural trends. Neuman, W.L. (2014).
It often takes an inductive view of the relationship between theory and research, which means theory emerges from research. The focus will be data collection methods in qualitative research, sampling, focus groups, semi-structured interviewing, etc… Bryman, A., et al 2022 (pp. 279-303; 305-316; 347-366)
Qualitative features
- Subjectivity
- Contextualization
- Flexibility
- Richness and depth
- Interpretation and meaning-making
- Inductive reasoning
- Naturalistic setting
Types of qualitative research methods
- One-on-one interview: Conducting in-depth interviews (details follow)
- Focus Groups: Answers the “why, what, and how” questions. Used to explain complex processes.
- Ethnographic research: Method applied in naturally occurring environment. (target
audiences’ environments) to understand the cultures, challenges, motivations, and settings that occur. Instead of relying on interviews and discussions, you experience the natural settings firsthand. - Case study research: The case study method has evolved over the past few years and developed into a valuable quality research method.
- Record keeping: This method makes use of the already existing reliable documents and similar sources of information as the data source. This data can be used in new research.
- Qualitative observation: It deals with the 5 major sensory organs and their functioning – sight, smell, touch, taste, and hearing
Types of qialitative reserach methods
- Etnography
- Grounded Theory
- Narrative
- Case Study
- Phemonenology
- Historical
Attributes Qualitative vs. Quantitative
1. Analytical objectives
- This research method focuses on describing individual experiences and beliefs.
- Quantitative research method focuses
on describing the characteristics of a
population.
2. Types of questions asked
- Open-ended questions
- Closed-ended questions
- Data collection Instrument
- Use semi-structured methods such as in-depth interviews, focus groups, and participant observation
- Use highly structured methods such
as structured observation
using questionnaires and surveys
4. Form of data produced
- Descriptive data
- Numerical data
5. Degree of flexibility
- Participant responses affect how and which questions researchers ask next
- Participant responses do not influence or determine how and which questions researchers ask next
Limitations of Qualitative Research Methods
§ Limited generalizability: May be difficult to generalize findings to larger target populations.
§ Subjectivity concerns
§ Time-intensive
§ Resource-intensive: Costly and labour-intensive.
§ Difficulty in replication
§ Researcher bias: The researcher may inadvertently influence the research process and results.
§ Complexity in data analysis: Analysing qualitative data, including coding and thematic analysis, can be complex and time-consuming.
Experimental Methods
- Non-experimental
Measure outcomes before and after program for participants only
No comparison group - Quasi-experimental
Measure outcomes for program participants and non-perticipants withput random assignment
“Control” for bias
Comparison group
- Experimental
Randomized Control Trial (RCT)
Randomize participants to treatment or control group
Measure outcomes for both groups
Explicit comparison group
Empirical vs. Non-empirical
- Empiricism/empirical:
- Defined as information collected in a laboratory or the field, based on specific and systematic analyses.
- The main characteristics of an empirical approach are:
- Collecting experiences (information),
- Questioning
- Observing/Describing
The main characteristics of non - empirical approach are:
* Generally concerned with the reasons underlying human behaviour i.e. the why or how as opposed to the what, where, and when.)
* Library Research
* Literature Research
* Scholarly Research
* Think Pieces
What is intersubjective traceability?
Intersubjective traceability means that the empirical research is conducted independently of the person and the personal preferences of the researcher.
Exploratory and Explanation
Exploratory Research
§ To identify new opportunities
§ Understand consumer behaviour
§ Support in diagnosing problems
§ Help to formulate hypotheses
§ Assist in market segmentation
§ Testing the feasibility of new ideas
§ Helps to prepare for large-scale research
Explanatory Research
§ Exposes If-Then-relations between two or more facts.
§ Explanatory research is defined as a way of connecting ideas to understand cause and effect. Researchers use the variables to explain why or how to detail what is happening between the two variables.
Steps to conduct Exploratory Research
- Define research problem
- Conduct Literature review
- Formulate hypothesis
- Collect data
- Analyze data
- Draw conclusions
Types of explanatory research (+ slde 77 difference)
- Literature research
- In-depth interview
- Case studies
- Focus groups
Difference:
Exploratory: identify problem, hypothesize solution
Explanatory: develop RQ, formulate hypothesis
Purpose of Content Analysis
- Look for the Frequency of Words
- Patterns/Themes
- Sequence Occurrence
What is a content analysis?
Content analysis is an empirical method for the systematically,
intersubjectively transparent description of content and formal characteristics of messages. (Softwares Nvivo, ATLAS.it, Otter)
▪ Content analysis can be applied both in qualitative as well as in
quantitative research.
▪ It can be used for primary and secondary research approaches.
Qualitative content analysis
The qualitative content analysis method focuses on analysing
recorded communication taken from artefacts. For example, extracts from books, newspaper articles, interviews
Common Types of Qualitative Analysis
Content Analysis
Thematic Analysis
Grounded Analysis
Discourse Analysis
Narrative Analysis
Phenomenology/Heuristic Analysis
What is the difference between Content Analysis and Thematic Analysis?
Content Analysis is a research method that quantitatively analyzes texts. It focuses on counting the frequency of words, categories, or themes. The data is broken into smaller units, categorized, and their frequency is measured to identify patterns.
Example: Counting how many times certain words or categories are mentioned in an article.
Thematic Analysis is a qualitative method aimed at identifying and interpreting meanings within the data. It seeks to find key themes that represent ideas, concepts, or experiences expressed by participants.
Example: Analyzing interviews to uncover main themes such as emotional states or interpersonal relationships.
Commonalities: Both methods use coding, identify themes, and work with qualitative data.
Difference: Content Analysis focuses on quantitative aspects (how often something appears), while Thematic Analysis emphasizes qualitative aspects (what meaning lies behind it).
Objects of Content Analysis
▪ Objects of content analysis can be media products:
- articles from newspapers
- lyrics of songs
- TV and radio shows
- movies
- advertisements
- video clips
- blogs
- propaganda material
▪ In the content analysis of the data collected
through interviews, the objects of analysis
are not the interviewed persons, but their
statements (interview transcripts).
Example of Content Analysis
Original Text -> Paraphrasing -> Generalization -> Reduction -> Page
Difference between Quantitative and Qualitative Content Analysis
- Purpose:
Qualitative: Answers the question “Why?”
Quantitative: Answers the question “How many/how much?”
- Data Type:
Qualitative: Focuses on observations, symbols, and words.
Quantitative: Focuses on numbers and statistical results.
- Approach:
Qualitative: Observes and interprets data.
Quantitative: Measures and tests data.
- Analysis:
Qualitative: Groups common data; uses non-statistical analysis.
Quantitative: Relies on statistical analysis.
Quantitative:
You can analyze campaign speeches for
the frequency ofterms such as
unemployment, jobs, and work
* You can use statisticalanalysis to find
differencesover time or between
candidates
Qualitative
* You can locate the word unemployment
in speeches, and identify whatother words
or phrases appearnextto it - such as
economy, inequality or laziness
* Analyze the meanings ofthese
relationships to better understandthe
intentions andtargets of different
campaigns.
Thematic Analysis
Qualitative data analysis method that involves reading through a data set
Identifying patterns in meaning across the data to derive themes
Reflexivity: The researcher’s subjective experience plays a central role in meaning-making from
data
What is a Code in Qualitative Analysis?
- Coding is a qualitative data analysis strategy in which some aspect of the data is assigned a descriptive label that allows the researcher to identify related content across the data
- It is the process of labelling and organizing your qualitative data to identify different themes and the relationships between them
Thematic analysis in 6 steps
- Familiarize yourself with the data
- Generate initial codes
- Search for themes
- Review potential themes
- Define and name themes
- Produce the report
Qualitative Data Analysis - Coding
- Coding means categorizing data into chunks or segments
- Start with a maximum of 8-10 codes – you can develop and add more as you go through the analysis
- Be flexible in refining codes
- If possible, refine codes in discussions with others
- You can expand the code and refine the properties of the code or merge codes into existing codes
What is a Quantitative Content Analysis?
- It is a research method in which the features of textual,
visual, or aural materials are systematically categorized
and recorded so that they can be analysed. - Employed in the field of communication
- Central to content analysis is the process of coding,
which involves following a set of instructions about what
features to look for in a text and then making the
designated notation when that feature appears.
Requires careful attention to
* Unitizing: segmenting the texts for analysis
* Sampling: selecting an appropriate collection of
units to analyze
* Reliability: different researchers making codes
consistently
* Validity:using a coding scheme that
adequately represents the specified
phenomena
Difference between Codes and Categories
- Code: the label you attach to a phrase or
other short sequence of the text you are
analysing. - Example: ‘politician’ for each mention in the text of any political figure.
- Category: a grouping you impose on the
coded segments to reduce the number of
different pieces of data in your analysis
Formal Categories
Formal categories describe the formal characteristics of each particular unit of analysis.
Formal categories, like sociodemographic data in surveys, are not placed at the center of surveys, but
they deliver important additional information.
Requirements for Categories
Complete categories
Precise separation of categories
Inside the category as well as among categories.
▪ Categories are precisely separated when ▪ particular values exclude themselves
mutually and
▪ when all values refer to the same characteristic (variable).
Example:
1 = German Studies,
2 = Humanities…
Benefits of Qualitative Coding
- Increase validity: Qualitative coding provides organization and structure to data so that you can examine itin a systematic way to increase the validity of your analysis.
- Decrease bias: Qualitative coding enables you to be aware of potential biases in the way data is analyzed.
- Accurately represent participants: Qualitative coding allows you to evaluate if your analysis represents your participant base, and helps you avoid over-representing one person or group of people.
- Enable transparency: Qualitative coding enables other researchers to methodically and systematically review your analysis.
Inductive Coding vs. Deductive Coding
- Developing codes based on the Data itself
Here, you, as a researcher, have no preconceived codes. In other words, you read through the text and allow the codes to emerge from the data. - It is suitable for exploratory research from bottom-up and iterative. Especially when there are limited existing theories or understanding of a particular phenomenon.
Deductive
* This approach uses an articulated theory or existing theoretical framework as a basis for a pre-defined set of codes. These sets of codes are designed in advance and contained in a codebook.
* You approach the data with a pre-defined code for deductive coding. This approach works better when testing a theory rather than when exploring a phenomenon.
* Top-down approach.
* Used in assessing or conducting a study on Motivation theory to a unique context
Categories and subcodes
Categories canbeorganized inavariety of ways.
* Within each category, you can group codesthat aresimilar to eachother or pertain to the same topics or general concepts.
Meaning unit
Because at the health
facility there are
trained stall, and in
case you have a
Problem, they would
know and help you
accordingly.
Condensed meaning unit
At the health
facility, there are
trained staff
In case you have a
problem, they would
know and help you
Codes
Health providers are
trained Competent
Heath providers identify
and manage
complications
Categories
It is safer
Theme
Experiences related
to the promotion of
facility childbirth
Qualitative Content Analysis according to Mayring
- systematic (principle-oriented),
- intersubjectively traceable analysis of
- large amounts of text material.
- It includes detailed analyses (focusing on small units of meaning) and aims at an extensive category system, which is the basis of a summarizing interpretation of the material.
Mayring‘s Analysis Concept
Maying’s analysis concept includes three steps:
1. Summarizing content analysis,
2. Explicative (explaining) content analysis and
3. Structuring content analysis.
Summarizing Content Analysis
▪ The source text is reduced to a comprehensible short version, which includes only the most important content.
▪ The aim is to reduce material and keep the essential contents.
What belongs to the work process is:
Paraphrasing:
Crossing out elaborate phrases, transforming expressions into grammatical short forms.
Generalization:
Concrete examples are generalized.
Reduction:
Similar paraphrases are summed up.
Class Task in Groups –Qualitative Content Analysis (Mayring)
Question: How would you address the research question: “How often do male and female university students use a food delivery app monthly?”
Answer for a Study Card:
Method:
Use Interviews to gather data.
Data Type:
Collect Qualitative Data for in-depth insights.
Target Audience:
University students (both male and female).
Categories and Sub-Categories for Interview Questions:
Usage Frequency: Times per week/month.
Reasons for Use: Convenience, cost, or time-saving.
Preferred Apps: Most used apps and reasons for preference.
Barriers: Issues like delivery fees, app reliability, or menu options.
Findings Analysis:
Analyze data using Mayring’s Qualitative Content Analysis by summarizing, explaining, and structuring the responses.
Explicative Content Analysis
▪ Unclear text elements are made comprehensible by using extra materials (e.g. other interview sections, information about the
interviewee).
The objective is to extend the comprehensibility of the text.
Structuring Content Analysis
The summarizing and explicative short version of the text is now arranged and structured according to the theoretical set of questions.
For this purpose, you establish a category scheme and, after testing it once, you improve the category system before the final evaluation follows.
Your aim is:
* to filter out particular aspects
* to conduct a cross-sectional study or
* to assess the material under specific criteria.
Three options for structuring
Content structuring analysis: elaborating specific topics and contents
Standardising structuring: identifying frequently recurring or
theoretically interesting values.
Scaling structuring: values are ranked on ordinal level.
Structuring Content Analysis: Three process steps
1.Defining the categories
* Explicitly define text elements which belong to one category.
2. Standard examples
* Mentioning concrete text examples, which fall under one category and are exemplary for this category.
* These standard examples have a prototype function for this particular category.
3. Coding rules
* Where there are de-limitation problems among categories, formulate rules to allow for a clear correlation.
Problems of Content Analysis
▪ The problem of polysemy of terms is that it can never be controlled by conducting a content analysis. Terms have multiple meanings (e.g., “Bank…).
▪ Inference to the recipients and communicator level is complicated when e.g. there is no extra survey conducted on the part of the recipients
▪ Sequential aspects that the person responsible for coding has to follow when coding evaluation data (for example).
▪ The content of a unit of analysis is constructed, according to the circumstances, across several sentences
▪ It changes the direction of evaluation from positive to negative and comes to a completely different end than its starting point.
▪ Solution: Such issues must be clearly and precisely coded in the codebook.
Types of Generalization in quantitative and qualitative research
- Classic sample-to-population (statistical) generalization
- Analytic generalization (qualitative)
- Case-to-case transfer (transferability)
Analytic generalization (Qualititative)
- Analytic generalization is most often linked with qualitative research, although it is implicitly embedded within theory-driven quantitative research as well.
- In an idealized model of analytic generalization, qualitative researchers develop conceptualizations of processes and human experiences through in-depth scrutiny and higher-order abstraction.
- Qualitative researchers distinguishbetweeninformationthatis relevantto all(or many) study participants, in contrastto aspectsofthe experience that areunique to particular participants.
- Through inductive analysis,together with the use of confirmatory strategies that address the credibility of the conclusions, qualitative researchers can arrive at insightful, inductive generalizations regarding the phenomenon under study.
Transferability- Case to Case
- Case-to-case transfer involves the use of findings from an inquiry to a completely different group of people or setting, is more widely referred to as transferability but has also been called reader generalizability
- Transferability is most often discussed as a collaborative enterprise. The researcher’s job is to provide detailed descriptions that allowreaders tomake inferences about extrapolating thefindings to other settings.
What is a Case Study?
In-Depth Exploration
* Acase study in qualitative research involves a detailed and in-depth exploration of a specific individual, group, event, or phenomenon.It aims to provide a comprehensive understanding of the case in its real-life context.
Contextual Analysis
* The focus of a case study is on the context and the unique characteristics of the case.
Tools commonly used in case studies include:
* Surveys
* Interviews
* Observations
* Artefacts
Holistic Perspective
* Acase study examines the interplay ofmultiple factorsand considers the intricaciesofthe casewithin its natural setting, allowing for a nuanced and comprehensive analysis.
Inductive Approach
Case Studies main rules
- Define Clear Objectives
- Select Appropriate Cases
- Develop a Theoretical Framework
- Use Multiple Data Sources
- Ensure Data Quality
- Employ Systematic Data Analysis
- Maintain Reflexivity
- Develop Rich Descriptions
- Draw Conclusions and Implications
Case study is a Research Strategy
A case study can be descriptive, explanatory or exploratory.
* Descriptive: In a descriptive case study, the purpose is to ‘describe’ a phenomenon in detail in its real-world context. It
is used extensively in sociology and anthropology.
- Explanatory:The study looks for causal factors to explaina particular phenomenon. The primary focus of such a case study is to explain‘why’ and ‘how’ certain conditions come into being,that is, why certain sequence of events occur or do not occur.
-
Exploratory:The purpose is to study a phenomenon with the
intention of ‘exploring’or identifying fresh research questions, which can be extensively used in subsequent research studies.
- You would like to study young people aged 18-21 in the Fridays For Future movement’s perception of their impact during the climate summit.
- What is your research question -considering you want to investigate why and how?
- What kind of case study would you conduct? (Descriptive, Exploratory, Explanatory)
- What kind of data would you need and how would you collect it?
Answer:
-
Research Question:
- Why do young people aged 18-21 in the Fridays For Future movement believe they have an impact during the climate summit, and how do they perceive this impact?
-
Type of Case Study:
- Exploratory Case Study: Since the focus is on exploring perceptions and understanding the reasons behind their views, this type is most suitable.
-
Data Needed:
- Qualitative Data: Personal experiences, opinions, and perceptions of participants.
-
Data Collection Methods:
- Interviews: Conduct in-depth interviews with participants in the 18-21 age range who attended or followed the climate summit.
- Focus Groups: Facilitate discussions among Fridays For Future members to understand shared perceptions.
- Observational Data: Analyze social media posts or activities related to the movement during the summit.
- Surveys: Optional for capturing broader opinions within the movement.
Sampling is relevant in a case study
A qualitative sample plan includes answers to the following questions:
* Is the sampling relevant to your conceptual frame and
research questions?
* Can reliable descriptions or explanations be produced using
the sampling plan selected?
* Is the sampling plan feasible, in terms of time, money, manpower, and access to people under study?
* Is the sampling plan effective enough for its findings to be generalisable to the entire universe of the population from which
the sample is obtained?
An important element in case study strategy is the relation
betweentheory andcasestudy research.
* A case study can aptly be used for testing the
hypothesesandnotthe entire theory.
Data Analysis for Case Studies: principles
Thematic Analysis
Constant Comparison
Coding and Categorization
Member Checking
Triangulation of Data (multiple sources of data)
Contextualization and Thick Description
There are two types of validity: external and internal.
External validity
* Can address the issue whether the findings of a study can be carried over (generalised) to other cases.
External Validity:
Assesses the extent to which the results can be generalized to other groups, situations, or conditions.
Key concept: generalizability.
Tested through replication of results.
Internal Validity:
Determines how accurately causal relationships between variables are established within the study.
Key concept: accuracy and control of factors.
Requires consideration of all external influences to avoid false conclusions.
Levels of measurement or scales of measurement – classification principle of variables I
- Nominal data / variables (e.g. hair colour/nationality)
- Ordinal data (variables have natural, ordered categories: e.g. hot, hotter, hottest)
- Metric data (e.g. age)
Classification principle of variables
Discrete variables - counting numbers with specific values (e.g. age)
* Continuous variables
measurable number e.g. 1.7 feet
Measuring definition
Measuring is the systematic assignment of an amount of numbers or symbols
to the values of a variable, hence also to the objects.
Nominal, Ordinal & Metric Data
Nominal (categorical, qualitative) data
* Variables whose value cannot be organized in a logical sequence and which can only be distinguished by their name. (e.g. gender)
Ordinal data (rank variable)
* Variables that have two or more values which can also be ordered or ranked. That means that variable values have a natural sequence. (e.g. school grades).
Metric data
* Values that can be organized according to size AND can represent a multiple of a unit (e.g. height).
Ordinal Data:
Data with values that can be ranked or ordered, but the exact difference between values is unknown.
Example: School grades (A, B, C) or race positions (1st, 2nd, 3rd).
Metric Data:
Data with values that can be ordered and have meaningful differences, including a measurable unit.
Example: Height (120 cm, 130 cm) or temperature (20°C, 30°C).
Discrete vs. Continuous variables
Discrete variables * A discrete variable is one that has specific values and cannot have values between these specific values. (e.g. school grades, gender, number of mistakes) Continuous variables * There is in theory an infinite number of values between any two values that a continuous variable can take. (e.g. size of screws, height, age).
Nominal, Ordinal, Interval, Ratio
Difference Between Interval and Ratio Scales
Interval Scale:
Measures data where equal intervals between values represent equal differences.
No true zero: A value of zero does not mean “nothing” (e.g., zero degrees Celsius doesn’t mean no temperature).
Examples:
Temperature in Celsius or Fahrenheit.
IQ scores.
Ratio Scale:
Measures data where equal intervals exist, and there is a true zero point, meaning zero indicates “none” or “nothing.”
Allows for comparisons like “twice as much” or “half as much.”
Examples:
Height (e.g., 0 cm means no height).
Weight (e.g., 0 kg means no weight).
Income (e.g., $0 means no income).
Index
An index is formed by summing up several single indicators.
Basis for the ranking:
* An extensive questionnaire
* Different indicators (pressure, corruption), moderators, impact factors
- The index is the sum of all measured variables and hence it is itself a new variable.
- The index must comprehensively and one-dimensionally represent the area of variables, that is the indicators of a theoretical construct.
- The measurements that enter an index must all have the same measuring level, e.g. nominal, dichotomous scales.
Scales
Scales (or rating scales) are measuring instruments that consist of several individual measurements (items), which all cover identical or similar variables and refer to the same theoretical construct
▪ Likert scale:
▪Consists of several at least 5-stage items
(interval scaled), which are combined to
an index by summation.
▪ Semantic differential:
▪Consists normally of about 10 to 20
opposing pairs. Typical usage of a Likert
scale with the aim to empirically determine
the meaning of a concept. Multipoint rating
option.
Principles of good research practices
- Honesty
- Transparency
- Independence
- Responsibility
- Scientific integrity
- Follow guidelines of good research practice
CRAAP test
- Currency: timeliness of the information
- Relevance: consider the audience and compare with a variety of sources
- Authority: origin of source and credentials
- Accuracy: reliability, check evidence and bias errors
- Purpose: identify type of information – fact/opinion- and the intent of the author.
Multivariate Analysis
- Working with relationships * Multivariate analysis methods are used in the evaluation and collection of statistical data to clarify and explain relationships between different variables that are associated with this data
Regression Analysis
- Regression analysis helps in a randomized trial to test for imbalances
- Measure the influence of one variable on another variable
- Helps predict a variable through one or more variables Example: Does your age, level of education, and the amount of hours of your work have a connection to the amount of money you earn? * The salary in this case would be the dependent variable * The age, level of education, and the amount of hours of work would be an independent variable
Regression models provide a mathematical representation between a
set of explanatory variables and a response variable.
- The coefficients in a regression model indicate how much we expect the response to change when the explanatory variable is observed to change.
- Regression-to-the-mean occurs when more extreme responses revert to nearer the longterm average
- Regression models can incorporate different types of response variable, explanatory variables and nonlinear relationships.
Coefficients
- The coefficients in a regression model indicate how much we expect the response to change when the explanatory variable is observed to change.
- Regression-to-the-mean occurs when more extreme responses revert to nearer the longterm average
- Regression models can incorporate different types of response variable, explanatory variables and nonlinear relationships.
Regression Analysis
- Simple Linear
- One independent variable infers dependent variable
- Example: education level to predict salary
- Multiple Linear
- Several independent variables: education level, age, time to
predict salary - Logistic
- Categorical variable. Variable is yes or no
- Is a person at risk of burnout?
- Dependent variable in all cases can be metric, ordinal or normal
Linear model assumption
Allows us to study the relationship between two variables. * Linear regression aims at finding the best-fitting straight line through the points. * The best-fitting line is known as the regression line. * If data points are closer when plotted to making a straight line, it means the correlation between the two variables is higher. Here, the relationship is strong. * Example: the regression line shows the predicted score on ecommerce sales for each possible value of the online advertising costs
Simple Linear Regression
Assesses the relationship
between a dependent variable and an independent variable
* Example: relationship between the age and price for used cars sold in the last year by a car dealership company
Multiple Linear Regression
Assesses the relationship between multiple independent variables
Non-collinearity
* Independent variables need to show a minimum correlation with one another.
* If independent variables are highly correlated with each other, it will be difficult
to assess the true relationships between the dependent and independent
variables.
Example: more than one explanatory variable
* Predicted values of the dependent variable (heart
disease) across the full range of observed values
for the percentage of people biking to work.
* The effect of smoking on the independent variable,
researchers calculated these predicted values while
holding smoking constant at the
minimum, mean, and maximum observed rates of
smoking.
Variable & Cases
- Variable: a logical grouping of attributes
- A variable is anything that can take on different
values. For example, height, weight, age, attitude, and IQ are variables because there
are different heights, weights, ages, races, attitudes, and IQs. By contrast, if something cannot vary, or take on different values, then it is referred to as a constant. - Dependent Variable
- Independent Variable
Cases: are the things that- we are observing and to which we are assigning
values.
Cases: counties; variable: type of voting equipment; values: manual mark,
punch card, optical scan, electronic
Dependent Variable
The dependent variable is called “dependent” because it is
influenced by the independent variable.
* For example: a researcher may be interested in examining the effects of a new medication on symptoms of depression among college students.
* Dependent variable: a measure of depression (because it is influenced by - i.e., is dependent on-
* Independent variable (i.e., the medication).
* The dependent variable is a measure of the effect (if any) of the
independent variable.
Independent Variable
Independent Variable
* The independent variable is called “independent” because it is independent
of the outcome being measured. The independent variable is what causes or
influences the outcome.
* The independent variable is the factor that is manipulated or controlled by
the researcher. (e.g. how much medication?)
* In most studies, researchers are interested in examining the effects of the independent
variable.
Key Reasons for the Manipulability of Independent Variables
❖Control in Experiments:
In an experiment, the independent variable is the one the researcher actively changes or assigns to
participants.
For example, in a study on the effect of sleep on cognitive performance, the researcher can manipulate
the amount of sleep participants receive (e.g., 4 hours vs. 8 hours).
❖Cause-and-Effect Relationship:
By manipulating the independent variable, resulting changes in the dependent variable helps determine
causality. E.g., if increasing study time (independent variable) leads to higher test scores
(dependent variable), a causal link can be inferred/concluded.
❖Distinction from Non-Manipulable Variables:
Not all independent variables can be manipulated directly. Variables like age, gender, or
socioeconomic status can only be measured or categorized but not manipulated. These are often
treated as quasi-independent variables in research. In true experimental designs, manipulation is a
defining characteristic of the independent variable.
Manipulation of Independent Variables
Refers to the deliberate adjustment or control of a variable by researchers in an experiment to determine its causal effect on the dependent variable. The independent variable is the factor hypothesized to influence the outcome being studied, and its manipulation is key to testing causal relationships.
Why Are Independent Variables Manipulated?
To Test Causation:
* Manipulating the independent variable allows researchers to
observe changes in the dependent variable.
✓ For example, adjusting the intensity of light in a room
(independent variable) can help study its effect on reading speed
(dependent variable).
Control of Confounding Factors:
✓ Manipulating only one variable ensures that any observed effect
is due to that variable and not external factors.
Hypothesis Testing:
✓ Experiments often involve predictions about how changes in an
independent variable will affect the dependent variable.
Manipulation directly tests these predictions.
How Are Independent Variables Manipulated?
- Setting Different Conditions or Levels:
- Example: To study the effect of caffeine on memory:
- Group 1: No caffeine
- Group 2: 50 mg caffeine
- Group 3: 100 mg caffeine
- Random Assignment:
- Participants are randomly assigned to groups to ensure
fairness and eliminate bias. - Direct Changes:
- Physically or contextually altering the independent
variable. - Example: Changing the temperature in a room to
observe its effect on productivity.
Types of Manipulation
Direct Manipulation: ❖ Physically changing variables, such as room temperature or noise levels. Instructional Manipulation: ❖ Varying instructions given to participants (e.g., “focus on speed” vs. “focus on accuracy”). Event Manipulation: ❖ Altering the occurrence of events (e.g., showing participants happy or sad videos). Importance of Manipulating Independent Variables: I. It distinguishes experiments from observational studies. II. It enables researchers to make causal inferences. III.It increases the study’s internal validity by reducing alternative explanations. By deliberately changing the independent variable, researchers can identify its precise effect on the dependent variable, leading to robust and reliable conclusions.
Importance of Manipulating Independent Variables:
I. It distinguishes experiments from observational studies.
II. It enables researchers to make causal inferences.
III.It increases the study’s internal validity by reducing alternative explanations. By deliberately changing the independent variable, researchers can identify its precise effect on the dependent variable, leading to robust and reliable conclusions.
A variable’s level of measurement
- Interval/ratio variable
- Ordinal variables
- Nominal variables
- Dichotomous variables
Ratio-Level variables
Here’s the explanation in English:
1. Interval/Ratio Variables:
Definition: Quantitative variables with equal intervals between values. Ratio scales have a true zero point, while interval scales do not.
-
Interval variables: Zero does not mean the absence of the value.
Example: Temperature in Celsius (0°C does not mean “no temperature”). -
Ratio scales: Zero means the absence of the measured quantity.
Example: Weight (0 kg = no weight).
2. Ordinal Variables:
Definition: Qualitative variables that are ordered, but the difference between values is unknown or inconsistent.
- Example: Satisfaction ratings (1 = dissatisfied, 5 = very satisfied). The order matters, but the intervals between levels are not equal.
3. Nominal Variables:
Definition: Qualitative variables that represent categories without any order.
- Example: Eye color (blue, green, brown), nationality, car type.
4. Dichotomous Variables:
Definition: A special case of nominal variables with only two categories.
- Example: Gender (male/female), answers to a question (yes/no).
Key Differences:
- Interval/ratio — quantitative variables.
- Ordinal/nominal/dichotomous — qualitative variables.
Nominal variables
- Nominal and ordinal variables are categorical variables - their
values divide up cases into distinct categories. - The values of nominal-level variables have no inherent order.
The variable eye colour can take on brown, blue, and green eyes;
major— political science, sociology, biology, etc.
Ordinal-level variable
The values of ordinal-level variables have an inherent order.
* These values can be placed in an order that makes sense—first
to last (or last to first), least to most, best to worst, and so on
Why assign values to variables
We want to establish an ordinal order to our variables
* To analyze data we want to organize it
Why do we recode variables?
We want to transform a variable by combining some of its categories or values.
* Example: we want to change a continuous variable into an ordinal categorical variable or merge the categories of a nominal variable.
* In PSPP,this type of transform is called recoding. * Today, we want to recode our variables to merge them
What Does It Mean to Reject Ho?
If the p-value (probability value) from the test is less than the significance level
(α = 0.05):
➢ You reject Ho
.
➢ This means the data provides strong evidence against the null hypothesis.
➢ You conclude that there is a statistically significant effect or
relationship in the data.
Example: Rejecting Ho means there is a significant association between the
variables.
Basic Likert Scale Analysis
With Likert scale data we cannot use the mean as a measure of central tendency as it has no meaning: what is the average of strongly agree and disagree?
The most appropriate measure of it is the mode - the most frequent responses, or the median.
The best way to display the distribution of responses i.e. (% that agree, disagree etc) is to use a bar chart.
How to Calculate the Mode?
Assign numbers to the scale from 1-5, you could assign 1 to ‘very poor’ and assign 5 to ‘excellent’ depending on what the scale measures.
Draw a table for your results, you can have headings on both axes of the table.
The questions would be on one axis and the figure ratings from the last step on the other.
Fill in the number of times each rating appears (mode)
Likert Scale - Ordinal and Interval Data
If the data are ordinal, we can say that one score is higher than another. We cannot say how much higher, as we can with interval data, which tell you the distance between two points.
Quantitative research designs are either
Descriptive
Experimental
- Experimental: measurements before and after intervention
establishes causality - Descriptive: measurements made only once
establishes only correlations between variables
Experimental Research Design
Experimental research is a study conducted with a scientific approach using two sets of variables. The first set acts as a constant, which you use to measure the differences of the second set.
Inferential Statistical Techniques
Inferential statistical techniques are used to analyze the sample’s behavior.
These include the models used for regression analysis and hypothesis testing.
The first-step sample is used to draw conclusions.
Assumptions or predictions about the entire population are used to draw inferences.
Inferential Statistical Techniques are methods in statistics used to draw conclusions about a larger group (population) based on data from a smaller group (sample). These techniques help test hypotheses, make predictions, and estimate population parameters.
Key Points:
1. Purpose: To learn about the entire population by analyzing a sample.
2. Tools: Use probability to determine how accurate the conclusions are.
Examples:
1. T-test: Tests if there is a difference between the means of two groups.
Example: Comparing the average income of men and women.
-
ANOVA: Analyzes the difference in means across three or more groups.
Example: Comparing student performance across different schools. -
Regression Analysis: Identifies relationships between variables.
Example: Studying how education level affects salary. -
Confidence Intervals: A range where the true value of a parameter is likely to fall.
Example: The average height of a population is between 165 and 175 cm with 95% confidence.
Difference:
Unlike descriptive statistics, which simply describe the data, inferential statistics allow you to make predictions and test assumptions about the broader population.
Parametric vs. Nonparametric Tests
Parametric tests assume that the data are continuous and follow a normal distribution.
Although, with a large enough sample, parametric tests are valid with nonnormal data.
The 2-sample t-test is a parametric test.
Nonparametric tests are accurate with ordinal data and do not assume a normal distribution.
There is a concern that nonparametric tests have a lower probability of detecting an effect that actually exists.
The Mann-Whitney U test is an example of a nonparametric test.
Here are examples of parametric and nonparametric tests:
Parametric Tests:
1. T-test (Student’s t-test):
- Example: Compare the average height of men and women.
- Group 1: Heights of men (in cm).
- Group 2: Heights of women (in cm).
-
ANOVA (Analysis of Variance):
-
Example: Compare the average test scores of students from three different schools.
- Group 1: Scores of students from School A.
- Group 2: Scores of students from School B.
- Group 3: Scores of students from School C.
-
Example: Compare the average test scores of students from three different schools.
-
Pearson’s correlation:
- Example: Study the relationship between hours spent studying and exam results.
Nonparametric Tests:
1. Mann-Whitney U test:
- Example: Compare customer satisfaction levels in two stores (1–5 scale, where 1 = very dissatisfied, 5 = very satisfied).
- Group 1: Ratings from Store A.
- Group 2: Ratings from Store B.
-
Kruskal-Wallis test:
-
Example: Compare customer satisfaction levels across three restaurants.
- Group 1: Ratings from Restaurant X.
- Group 2: Ratings from Restaurant Y.
- Group 3: Ratings from Restaurant Z.
-
Example: Compare customer satisfaction levels across three restaurants.
-
Spearman’s rank correlation:
- Example: Study the relationship between stress levels (ranked) and hours of sleep.
Key Difference in Data:
- Parametric: Continuous data with a normal distribution (e.g., height, weight, scores).
- Nonparametric: Ordinal or continuous data without a normal distribution (e.g., ranks, satisfaction ratings).
