Exam Flashcards
(37 cards)
A type of research design involving one time collection of information from any given sample of population elements is called…
Cross-sectional design
All research designs
Descriptive, Exploratory, Experimental
Options: longitudinal, cross-sectional
Cross-sectional design
(Descriptive)
Single Cross-Sectional Designs
* Sample: One distinct group of respondents
* Data Collection: Occurs once from this group
* Purpose: Offers a snapshot of a specific group at a particular point in time
* Example: Surveying employees’ job satisfaction in a company in 2023
Multiple Cross-Sectional Designs
* Sample: Several distinct groups of respondents
* Data Collection: Occurs once from each group, often at different times
* Purpose: Compare and contrast different groups at different times without repeated measures on the same group. It’s like taking
multiple snapshots
* Example: Surveying employees’ job satisfaction in the same company in 2023, 2025, and 2027 using different employee samples
each time
Commonality: Both designs give insights into a specific time point, without tracking changes in specific individuals over time
Longitudinal Designs
(Descriptive)
Definitions
* Cohort: A group experiencing a shared event or characteristic in a specific timeframe
* Example: Individuals who entered the workforce in 2020
* Longitudinal Design: Research methodology collecting data on the same subjects repeatedly over time
* Example: Surveying a group of 100 people in 2020 about their job satisfaction, then re-surveying the same
group in 2022, 2024, and 2026 to track changes
Relationship
* Cohort Analysis are a type of longitudinal study
* While cohort specifies whom you’re studying, longitudinal describes how you’re studying them
A panel is…
* … a survey of individuals, households, companies etc. to obtain data on a single subject at regular
intervals over a longer period, using the same sample and carried out using the same methods each
time.
The key differences between longitudinal and cross-sectional designs are
- Timeframe
- Longitudinal: Studies the same subjects over a period of time.
- Cross-sectional: Examines different groups at a single point in time.
- Purpose
- Longitudinal: Tracks changes and development over time.
- Cross-sectional: Compares differences between groups at one moment.
- Data Collection
- Longitudinal: Multiple observations over time.
- Cross-sectional: One-time data collection.
- Strengths
- Longitudinal: Captures cause-and-effect and developmental trends.
- Cross-sectional: Quick, cost-effective, and easy to conduct.
- Weaknesses
- Longitudinal: Time-consuming, expensive, risk of participant dropout.
- Cross-sectional: Can’t track changes over time or establish causation.
3 research designs with their goal
- Exploratory: Typically small sample, no statistical tests,
Goal: generate first insights - Descriptive: Representative sampling, infer from sample onto population,
Goal: describe the population as it is - Experimental: Manipulation of independent variable in controlled setting,
Goal: measure causal effect of independent on dependent variable
Example: independent variable: advertising spot A (vs. spot B) dependent variable: sales
Explain the difference between correlation and causality
Correlation does not imply causation, and many misleading statistical conclusions arise when people assume a causal relationship from a mere correlation. Empirical research aims to establish causality through careful study design, such as experiments where variables are controlled and manipulated.
Correlation indicates an association between variables, while causality proves that one variable directly influences another (cause and effect relationship).
Describe the difference between difference and coherence hypotheses
A difference hypothesis checks if two things are different from each other (e.g., “One ad campaign leads to more sales than another”), while a coherence hypothesis looks at how two things are connected (e.g., “Spending more on ads leads to higher sales”).
A difference hypothesis compares two or more groups to determine if there is a significant difference between them (e.g., “Ad campaign 1 results in lower sales than Ad campaign 2”), while a coherence hypothesis examines the relationship between two variables, suggesting that changes in one correspond to changes in the other (e.g., “The higher the ad spending, the higher the sales”).
Name all types of hypotheses
difference hypotheses
coherences hypotheses
one-tailed
two-tailed
Define deductive/inductive approaches
Deductive approach: classical approach, start with problem discovery, confirmatory.
Starts with a general theory or idea, then tests it with data (top-down reasoning). Example: “If customer satisfaction leads to loyalty, then happy customers should return more often.”
Inductive approach: data approach, start with data processing (observations), exploratory.
Starts with observations or data, then develops a theory based on patterns found (bottom-up reasoning). Example: “Many loyal customers seem happy; maybe satisfaction leads to loyalty.”
Describe the concept of panel design
A panel design is a type of longitudinal study where the same individuals, households, or companies are surveyed repeatedly over a longer period using consistent methods.
While panel designs provide valuable insights into how variables evolve, they face challenges such as panel mortality (dropout of participants), selection effects (non-representative initial samples), and panel participation effects (respondents altering answers due to survey experience)
Concept of panel designs in descriptive research
Problems in panel design
panel mortality (dropout of participants)
selection effects (non-representative initial samples)
panel participation effects (respondents altering answers due to survey experience)
Information collected is predetermined.. hard to make changes!
Difference in primary and secondary data
Primary data is collected firsthand by researchers for a specific study or purpose (e.g., surveys, experiments, interviews), while secondary data is pre-existing data collected by someone else for a different purpose (e.g., census reports, company records, published studies).
Explain the Bradley effect + give possible explanation
The Bradley Effect refers to the observed discrepancy between voter opinion polls and actual election outcomes for African-American candidates.
Social Desirability Bias – Some voters may have told pollsters they would vote for the Black candidate to avoid appearing racist but voted differently in private.
What research design is used when a problem exists but you don’t know why?
Exploratory research design is used when a problem exists but the cause is unknown.
Objective: Discovery of ideas and insights
Characteristics: Flexible, versatile, Often the front end of total research design
Goal: To gain initial insights and understand the problem.
Methods: Often qualitative, such as interviews, focus groups, or observations.
Characteristics: Flexible, small sample size, no statistical tests.
Example:
A company notices a drop in customer satisfaction but doesn’t know why. It conducts focus groups and in-depth interviews to explore potential reasons, such as poor customer service or product quality issues.
Discuss the pros/cons (benefits/risks) of LLMs in qualitative research
Pros (Benefits):
1. Automated Text Analysis – LLMs can process large volumes of qualitative data (e.g., interviews, open-ended survey responses) quickly and efficiently.
2. Sentiment Analysis – They help identify consumer emotions and attitudes in social media, reviews, and survey responses.
3. Thematic Coding Assistance – LLMs can assist in qualitative coding by identifying common patterns and suggesting themes, speeding up analysis.
4. Scalability – They enable large-scale qualitative studies that would otherwise be too resource-intensive for manual analysis.
5. Reducing Human Bias – LLMs provide a neutral perspective, minimizing personal biases in qualitative research interpretation.
Cons (Risks):
1. Context Misinterpretation – LLMs may misinterpret nuances, sarcasm, or cultural-specific meanings in text.
2. Bias in Training Data – If biased data is used for training, the LLM may reinforce and replicate those biases in its analysis.
3. Loss of Depth and Nuance – AI-generated summaries might overlook unique, outlier insights that a human researcher would find valuable.
4. Over-Reliance on Automation – Researchers might depend too much on LLMs, neglecting the importance of human interpretation and validation.
5. Ethical and Privacy Concerns – Using AI to analyze sensitive qualitative data raises privacy issues, especially with proprietary or personal data.
Example:
A company analyzing customer complaints using LLMs can quickly categorize concerns (e.g., product defects, delivery issues) but might misinterpret sarcasm in reviews (e.g., “Great service—if you enjoy waiting three weeks for delivery!”).
What is typical for focus groups
Typical Characteristics of Focus Groups
1. Group Size – Usually consists of 6-12 participants to ensure diverse opinions while allowing for discussion.
2. Homogeneous Composition – Participants are often pre-screened to share common characteristics relevant to the study (e.g., similar demographics or consumer behavior).
3. Moderated Discussion – A skilled moderator guides the conversation, ensuring all voices are heard while keeping the discussion focused.
4. Relaxed and Informal Setting – Conducted in a comfortable environment to encourage open and honest discussions.
5. Interaction-Driven – Participants engage with each other, reacting to others’ viewpoints, which adds richness to the data.
6. Recorded for Analysis – Typically audio or video recorded to capture verbal and non-verbal cues for later review.
7. Time Duration – Lasts between 1-3 hours, depending on the depth of discussion needed.
8. Used for Exploratory Research – Helps uncover opinions, perceptions, and motivations rather than statistical validation.
9. Application in Marketing and Social Research – Commonly used for product testing, brand perception studies, and policy discussions.
Example:
A company launching a new soft drink might conduct a focus group with young consumers to discuss branding, taste preferences, and packaging appeal.
What is the purpose of qualitative research
Purpose of Qualitative Research
1. Gain Deep Understanding – Explores experiences, behaviors, and social phenomena in their natural context.
2. Generate New Insights – Identifies emerging trends, attitudes, and motivations that may not be captured through quantitative methods.
3. Explore Complex or Unstructured Topics – Investigates topics that are difficult to measure numerically, such as emotions, beliefs, or social interactions.
4. Develop Theories and Concepts – Helps build or refine theories based on real-world observations and patterns.
5. Provide Context to Quantitative Data – Explains the “why” behind numerical trends and unexpected findings in statistical research.
6. Understand Social and Cultural Meanings – Examines how people’s backgrounds, environments, and interactions shape their views and decisions.
Example:
A company using in-depth interviews to explore why customers feel emotionally connected to their brand, rather than just measuring customer satisfaction scores.
Differences in sampling in qualitative vs. quantitative research
Qualitative Research (Exploratory & Interpretative)
* Purpose: Understand meanings, experiences, and social phenomena.
* Sampling Method: Non-random (purposive, theoretical, or snowball sampling).
* Sample Size: Small (focuses on depth rather than breadth).
* Selection Criteria: Participants chosen for their rich insights and relevance.
* Flexibility: Sample can evolve as themes emerge.
* Data Collected: Text-based (interviews, focus groups, observations).
Quantitative Research (Statistical & Generalizable)
* Purpose: Measure variables, test hypotheses, and generalize findings.
* Sampling Method: Random or probability-based (random, stratified, systematic).
* Sample Size: Large (ensures statistical significance).
* Selection Criteria: Participants represent a larger population.
* Flexibility: Fixed sample, determined before data collection.
* Data Collected: Numerical data (surveys, experiments, structured observations).
What is a characteristic of qualitative research
Characteristic of Qualitative Research
* Exploratory & Interpretative – Aims to understand meanings, experiences, and social phenomena.
* Non-Numerical Data – Focuses on text, images, videos, and observations rather than numbers.
* Small, Purposive Samples – Participants are selected for relevance, not for statistical representativeness.
* Flexible & Adaptive – Research design may evolve based on emerging insights.
* Context-Dependent – Considers social, cultural, and environmental influences on behavior.
* Rich & Deep Data – Provides detailed descriptions rather than broad generalizations.
* Subjective Interpretation – Findings depend on researcher analysis and contextual understanding.
Match the statements with internal/external validity:
* The extent to which the results of an experiment can be generalized – from sample to population.
* The degree to which findings are representative.
* The degree to which a causal conclusion can be drawn.
* The extent to which changes in dependent variables can be explained by experimental manipulation and not by external factors.
External Validity:
* The extent to which the results of an experiment can be generalized – from sample to population.
* The degree to which findings are representative.
Internal Validity:
* The degree to which a causal conclusion can be drawn.
* The extent to which changes in dependent variables can be explained by experimental manipulation and not by external factors.
Internal Validity
Definition: The extent to which a study accurately establishes a cause-and-effect relationship between variables.
Focus: Ensures that changes in the dependent variable are due to the independent variable and not external factors (confounders).
Key Concern: Controlling for extraneous variables to rule out alternative explanations.
Example: A lab experiment testing the effect of a new medication controls for diet, lifestyle, and other health factors to ensure the drug is the only influencing factor.
External Validity
Definition: The extent to which a study’s findings can be generalized beyond the specific sample, setting, or time of the research.
Focus: Ensures results apply to different populations, locations, and real-world settings.
Key Concern: Representativeness of the sample and realism of the study conditions.
Example: If a study on consumer behavior is conducted only with university students, its external validity is low because results may not apply to older consumers.
Name the two means to increase internal validity and explain their purposes
- Controlling Extraneous Variables
- Purpose: Ensures that only the independent variable (IV) influences the dependent variable (DV) by eliminating or reducing alternative explanations.
- Methods:
- Randomization: Randomly assigning participants to groups to evenly distribute confounding factors.
- Matching: Ensuring participants in different conditions have similar characteristics (e.g., age, gender).
- Statistical Control: Measuring extraneous variables and adjusting for their effects using statistical techniques.
- Design Control: Using experimental designs that account for confounders by treating them as additional variables.
- Example: In a drug trial, using random assignment ensures that pre-existing health conditions do not bias results
- Conducting Manipulation Checks
- Purpose: Verifies whether the independent variable was perceived as intended and that it affects the dependent variable as expected.
- How?
- Pre-tests and pilot studies to confirm participants understand the treatment.
- Asking questions that assess whether participants noticed or responded to the manipulated condition.
- Ensuring the IV has a measurable impact before drawing causal conclusions.
- Example: In an experiment testing the effect of brand trust on purchase decisions, a manipulation check could ask participants to rate how trustworthy they perceived the brand after exposure to different advertisements.
By applying these strategies, researchers reduce bias, eliminate confounding factors, and improve causal conclusions, strengthening internal validity.
