Class 1: Research Quality

Class 1: Introduction to Research

• Quality in research includes:
  • Objectivity and independence: unbiased and impartial
  • Controllability: transparent research execution and documentation
  • Reliability: consistent results
  • Validity: accurate measurement of concepts
  • Generalizability: representative sample
• Research process:
  1. Formulation of problem definition and research question
  2. Critical literature review
  3. Methodology
  4. Data collection
  5. Data analysis
  6. Report and presentation

Class 2: Research Questions and Hypotheses

• Research question:
  • A specific question that guides the research
  • Formulated in a concrete way, involving variables
• Types of research questions:
  • Descriptive: describes features of a concept
  • Comparative: examines differences and similarities
  • Evaluative: looks at advantages and disadvantages
  • Explanatory: investigates causes and consequences
  • Testing: assesses the effect of one variable on another
• Good research question:
  • Clear, complete, and concise
  • Comprised of a main question and possible sub-questions
  • Feasible and specific
  • Relevant and original
• Hypothesis:
  • A preliminary statement indicating what the researcher expects to find
  • Logically deduced from a theory or developed from observed facts
  • Closely linked to the research question
• Good hypothesis:
  • Precise, testable through research, and formulated before conducting the research
  • Includes variables, the studied group, and the expected outcome

Class 2: Variables

• Variables:
  • Characteristics or features that can vary among respondents or participants
  • Directly linked to the research question and hypotheses
• How to identify variables:
  • Break down the research question into specific elements that can be measured or observed
• Conceptualizing and operationalizing:
  • Linking abstract concepts to measurable variables
  • Conceptualization: developing clear, concise definitions
  • Operationalization: defining how concepts will be measured or observed
• Levels of measurement:
  • Determine how variables are quantified and analyzed
  • Different levels of measurement: nominal, ordinal, interval, and ratio

Class 3: Desk Research

• Desk research:
  • Analyzing existing information from sources like books, news media, and scientific articles
  • Less expensive, but may include outdated, biased, or incomplete information
• How to conduct desk research:
  • Identify key terms
  • Gather information
  • Store and categorize
  • Process information
  • Quick scan and critique
• Evaluating desk research:
  • Check for objectivity, controllability, reliability, validity, and generalizability

Class 3: Referencing

• Referencing:
  • Citing sources of information and ideas used in the research
  • Acknowledging original authors and providing evidence for arguments
• Why do we do referencing?
  • Give credit to original authors
  • Provide evidence and support for arguments
  • Allow readers to verify and follow up on sources
  • Avoid plagiarism

Class 3: Field Research

• Field research:
  • Collecting new data directly from sources like participants
  • Specific to the researcher's needs, providing current, detailed information
• What is the difference between qualitative and quantitative research?
  • Qualitative research: focuses on understanding deep insights, attitudes, and behaviors
  • Quantitative research: involves numerical data to measure and analyze variables

Class 4: Quantitative Research

• Associations between variables:
  • Identifying independent and dependent variables
  • Independent variable: manipulated or categorized to observe its effect
  • Dependent variable: measured and expected to change in response to the independent variable
• Correlation and causality:
  • Correlation: indicates a relationship between variables, but does not imply causality
  • Causality: implies a direct effect of one variable on another
• Experimental research:
  • Involves manipulating the independent variable to observe its causal effect
  • Identifying and controlling for confounding variables

Class 4: Survey

• Survey:
  • A quantitative data collection method that involves spreading a questionnaire
  • Structured, typically consisting of closed questions
• Likert-scale:
  • Used in surveys to measure attitudes, beliefs, and perceptions
  • Assigns numerical values to respondents' levels of agreement
• Criteria for a good survey:
  • Simple and understandable language
  • Free of double negation and ambiguous questions
  • Objective, avoiding socially desirable answers
  • Structured with a clear introduction, organized layout, and logically bundled topics
  • Exhaustive and non-overlapping in answer options
• Types of surveys:
  • Written
  • Telephone
  • Face-to-face
  • Online
  • Panel surveys
• Codebook and datamatrix:
  • Help in organizing and analyzing survey data
  • Codebook links each survey question to a code and each response option to a specific value
  • Datamatrix represents the collected data in a structured format### Eye-Tracking
• Used to examine what captures the most attention of users or how users scan specific information
• Concepts linked to eye-tracking:
  • Fixations: periods when the eyes are relatively stationary, focused on a particular point
  • Heatmaps: visual representations of the most frequently viewed areas on a visual display

In-Depth Interviews

• Qualitative data collection method involving direct, one-on-one engagement with participants to explore their perspectives on a specific topic in detail
• When to use in-depth interviews:
  • When detailed, deep insights are needed about a participant's thoughts, feelings, or behaviors
• When not to use in-depth interviews:
  • When needing to generalize findings to a larger population or when quantitative data is required
• Advantages of in-depth interviews:
  • Provides deep, rich qualitative data
  • Flexibility in exploring new topics that arise during the interview
• Disadvantages of in-depth interviews:
  • Time-consuming and resource-intensive
  • Potential for interviewer bias
• Crucial steps in preparing an in-depth interview:
  • Develop a clear research goal
  • Create an interview guide
  • Pre-test the guide
  • Ensure informed consent is obtained from participants
• Developing an in-depth interview based on the specific research goal:
  • Link research questions to variables, then to indicators, and finally to interview questions

Focus Groups

• Qualitative research method involving guided discussions with a small group of participants to gather diverse perspectives on a specific topic
• When to use focus groups:
  • To explore complex behaviors, attitudes, and motivations
  • To generate ideas and insights for further research
  • When interaction between participants can provide deeper understanding
• When not to use focus groups:
  • When quantitative data is needed
  • When individual privacy and confidentiality are paramount
  • When the topic is too sensitive for group discussion
• Advantages of focus groups:
  • Generate rich, detailed data
  • Allow interaction and discussion among participants, leading to deeper insights
  • Cost-effective for gathering data from multiple people simultaneously
• Disadvantages of focus groups:
  • Potential for groupthink, where participants conform to the majority view
  • Difficult to generalize findings due to small, non-random samples
  • Requires skilled moderation to manage group dynamics

Observations

• Systematic watching and recording of behaviors and events in their natural settings
• Research questions suitable for observations:
  • Questions about how people behave in specific contexts
  • Questions aiming to understand processes, interactions, and environmental influences
• Contexts where observations are applied:
  • In natural settings like schools, workplaces, or public spaces
  • In controlled environments like laboratories
• Structured vs unstructured observations:
  • Structured: predefined categories and systematic recording
  • Unstructured: open-ended and flexible, without predefined categories
• Participating vs non-participating observations:
  • Participating: observer is actively involved in the context
  • Non-participating: observer remains detached and does not interact with subjects
• Open vs covered observations:
  • Open: subjects are aware they are being observed
  • Covered: subjects are unaware of the observation
• Natural vs artificial observations:
  • Natural: observations in real-world settings
  • Artificial: observations in a controlled, experimental setting
• People vs automatic observations:
  • People: observations conducted by human observers
  • Automatic: observations made using technology, such as cameras or sensors
• Direct vs indirect observations:
  • Direct: observing actual behavior as it occurs
  • Indirect: observing evidence of behavior after it has occurred

Ethnography

• Qualitative research method involving immersive observation and participation in the daily life of the study subjects
• Advantages of ethnographical research:
  • Provides deep, holistic insights
  • Captures context and nuances
• Disadvantages of ethnographical research:
  • Time-consuming
  • Potential for observer bias

Qualitative Data Analysis

• Research steps/process of qualitative research:
  1. Data collection
  2. Data analysis
  3. Interpretation of findings
• Inductive vs deductive in qualitative data analysis:
  • Inductive: developing theories based on observed data
  • Deductive: testing existing theories through data
• Three steps of processing qualitative data:
  1. Open coding: identifying and categorizing basic units of meaning in the data
  2. Axial coding: finding relationships between open codes
  3. Selective coding: identifying the core category and integrating other categories around it

Generalizability

• The extent to which the results of a study can be applied to or across different populations, settings, and times beyond the specific conditions of the original study
• Importance of generalizability:
  • Ensures the broader applicability and relevance of research findings
  • Allows findings to be useful in various contexts and for different groups, enhancing the study's impact and value
• When is generalizability important?
  • When research aims to inform policy or practice on a large scale
  • In studies seeking to establish universal principles or theories
• When is generalizability less important?
  • In exploratory or case studies focused on understanding specific contexts
  • In qualitative research prioritizing depth over breadth

Sampling

• Possible ways of sampling:
  1. Simple random sampling
  2. Systematic sampling
  3. Stratified sampling
  4. Cluster sampling
  5. Quota sampling
  6. Targeted sampling
  7. Snowball sampling
  8. Convenience sampling
• Role/importance of a sampling frame:
  • A sampling frame is a list of all members of the population from which the sample is drawn
  • It is crucial for ensuring that the sample accurately represents the population
• Difference between aselect and select sampling:
  • Aselect (random) sampling: every member of the population has an equal chance of being included
  • Select (non-random) sampling: members are chosen based on specific criteria or convenience### Grouped Frequency Tables
• Used for large data sets with continuous variables
• Important aspects:
  • Determine appropriate class intervals
  • Ensure intervals are mutually exclusive
  • Cover all data points
• Calculating frequencies:
  • Absolute frequency: count of occurrences of each value
  • Absolute cumulative frequency: cumulative count of occurrences up to a certain value
  • Relative frequency: proportion of occurrences of each value
  • Relative cumulative frequency: cumulative proportion of occurrences up to a certain value

Cross Tables

• A table showing the frequency distribution of variables simultaneously
• Used to explore relationships between two categorical variables
• Role of X and Y variables: X and Y represent the two variables being compared

Center Sizes (Measures of Central Tendency)

• Values that represent the center of a data set
• Calculated to summarize a data set with a single representative value
• Link to measurement level: determines which measure is appropriate (e.g., mean for interval/ratio, median for ordinal)
• Link to normal distribution: mean, median, and mode are equal in a perfectly normal distribution

Measures of Central Tendency

• Modus (Mode): most frequently occurring value
  • How to find: identify the value with the highest frequency
  • When to use: for nominal data or when the most common category is of interest
• Median: middle value when data is ordered
  • How to find: arrange data in ascending order and find the middle value
  • When to use: for ordinal data or when data is skewed
• Mean: average of all values
  • How to find: sum all values and divide by the number of values
  • When to use: for interval/ratio data with a normal distribution

Data in Daily Life

• Leaving data traces in daily life: digital footprints created through interactions with digital devices and online services
• Examples of a data-driven world: personalized advertising, recommendations on streaming services, smart home devices, and big data analytics in healthcare, finance, and urban planning
• Open data: data freely available for anyone to use, reuse, and redistribute without restrictions

Reporting Quantitative Results

• In-text reporting: used when having a limited amount of numerical data to share
  • Effective for highlighting key numbers directly within the text
• Tables: used when presenting a broad range of data or comparing data points
  • How to use: organize clearly with labeled columns and rows, provide a title and legend, and avoid clutter and excessive use of borders and colors
• Good tables: present data clearly and concisely, with appropriate labels and a clear structure, and without unnecessary detail

Data Visualization

• Heatmaps: graphical representations of data where individual values are represented as colors
  • Used to visualize the distribution and intensity of data across a given space
  • Link to tables: both present data, but heatmaps use color to represent data intensity
  • What to consider: color scale, avoiding too many colors, and normalizing data if necessary
• Interpreting heatmaps: understanding the color gradient and what it represents
• Graphs: visual representations of data designed to show relationships, patterns, and trends
  • Different graphs for different goals:
    • Bar charts for categorical data comparisons
    • Line graphs for showing trends over time
    • Scatter plots for showing relationships between two variables
• Scatterplots: display the relationship between two continuous variables
  • When to use: exploring potential relationships or correlations between two variables
  • Interpreting scatterplots: looking for patterns, trends, clusters, and outliers
• Lineplots: display data points over a continuous range, typically time
  • When to use: showing trends or changes over time
  • Interpreting lineplots: examining the slope of the lines to understand the direction and rate of change
• Bar charts: represent categorical data with rectangular bars
  • When to use: comparing values across different categories
  • Interpreting bar charts: comparing the lengths or heights of the bars
• 100% stacked bar charts: show the relative percentage of multiple data series in stacked bars, with each bar totaling 100%
  • When to use: comparing the proportional contributions of different categories while keeping the total constant at 100%

Data Journalism

• Using data in journalism: providing evidence-based reporting, uncovering trends, and supporting investigative stories
• Being critical when reading data journalism: checking data sources, methodology, and potential biases in data collection and presentation