Research Design Notes
Cross-Sectional Research
- Definition: a snapshot in time of a particular group; data are collected on one or more variables for a single time-period.
- Primary use: mostly for exploratory research.
- Example: Survey of university students’ attitudes toward climate change and their pro-environmental behaviors.
- Key assumption: there are no dramatic day-to-day fluctuations implied by the design, but results will reflect the time at which the data were collected.
- Time-specificity: findings from a cross-sectional study at a specific time are expected to be unique to that time, especially after major events (e.g., a disaster).
- History effect: the influence of events or circumstances outside the research study on the variables of interest (e.g., national security attitudes measured on September 10 vs. September 12).
- Cross-sectional vs. longitudinal distinction (overview):
- Cross-sectional design collects data at a single time period for one or more variables.
- Longitudinal design collects data for two or more time periods, enabling tracking of changes over time.
Cross-Sectional vs Longitudinal
- Cross-sectional: data collected at a single time period for one or more variables.
- Longitudinal: data collected for two or more time periods, allowing researchers to track changes in the variables.
Longitudinal Research
- Definition: collect data for two or more time periods.
- Two main purposes:
- To investigate changes in variables over time.
- To establish time order (causal relationships) between variables.
- Types of longitudinal designs:
- Trend studies
- Panel studies
- Cohort studies
- Cross-lagged studies
Trend Studies
- Description: gather data from a particular population characterized by a specific variable (e.g., level of education, political position).
- Method: draw a different sample from the population each time to determine trends in data.
- Advantages:
- Easy to maintain sample size across waves.
- Easy to track changes in the study variables.
- Disadvantages:
- Different samples may contain different individual characteristics, introducing error.
- Example: polling data.
Panel Studies
- Description: gather data from a group of individuals who are recruited and retained to answer questions over time.
- Advantages:
- Helps researchers identify factors of influence within the research sample.
- The same sample characteristics tend to persist over time.
- Disadvantages:
- Requires substantial time, energy, and resources.
- High drop-out rate.
- Participants may become familiar with measurements, potentially biasing responses.
- Example: surveying a group of 10 class representatives about their satisfaction with their high school experience over time.
Cohort Studies
- Description: focus on the experiences of people who belong to the same cohort (a cohort is a group with a shared experience).
- Example: All freshmen in Fall 2020 who experienced the transition to online classes and their overall success after college.
- Sampling: the cohort is sampled at each time point (Time 1, Time 2, Time 3), with each time point potentially using a different sample (e.g., 300 random samples of class of 2024 at each time).
- Advantages:
- Maintains a recognizable group across time.
- Easy to track changes in variables within the cohort.
- Disadvantages:
- Typically expensive in terms of time, money, and energy.
- Over time, researchers have less control over variables as the cohort ages.
- Cohort effect: a research result that occurs due to the characteristics of the cohort being studied, rather than due to the variables of interest.
- Example: studying the benefits of using a new AI tool across Baby Boomers vs. Gen Z.
Cross-Lagged Studies
- Design: measure an independent variable (IV) and a dependent variable (DV) at two points in time from the same sample.
- Procedure: IV and DV measured at Time 1, and again at Time 2 (IVT1, DVT1; IVT2, DVT2).
- Purpose: to draw conclusions about causality by establishing time order.
- Significance: the design is often described as the only survey design that permits researchers to assess causality; it is foundational for experiments in social science research.
- Advantages:
- Helps identify potential causal factors within the research sample.
- The same sample characteristics tend to persist over time.
- Disadvantages:
- Requires substantial time, energy, and resources.
- High drop-out rate.
- Variables may be influenced by confounding variables not accounted for in the design.
Example 1
- Dr. Russell’s research team surveys a cohort of children about their television viewing habits and their predisposition to violence at age 5 (Time 1) and again at age 10 (Time 2).
- Objective: assess the relationship between TV viewing and any change in predisposition to violence over time.
Choosing A Design
- Consider available resources and research objectives:
- Exploration and explanation: Cross-sectional study.
- Explanation, prediction, and control: Longitudinal study.
- What do researchers want to do?
- Snapshot status of one or more variables: Cross-sectional.
- Changes in one or more variables, causality: Longitudinal.
- Note: There is little or nothing that cross-sectional research can do that longitudinal cannot.
Practice Questions
Question 1
- Scenario: You are hired at Apple to learn about the current status of existing products and what people want to see in the future.
- Question: Which design is appropriate for this project?
- Options: a. Cross-sectional b. Longitudinal c. Sensational d. Nothing
- Answer: a. Cross-sectional
- Rationale: To capture current status and desires for the future in a single snapshot; longitudinal would be needed only if tracking changes over time.
Question 2
- Scenario: The university will evaluate a course at the end of the semester with one survey.
- Question: This course evaluation survey is which type of research?
- Options: a. Expensive b. Cross-sectional c. Longitudinal d. Meaningless
- Answer: b. Cross-sectional
- Rationale: One survey taken at a single time point fits cross-sectional design.
Question 3
- Scenario: To learn how students’ evaluation of COMG 102 changes throughout the semester.
- Question: What design should be used?
- Options: a. Expensive b. Cross-sectional c. Longitudinal d. Meaningless
- Answer: c. Longitudinal
- Rationale: Observing change over time requires data from two or more time points.
Question 4
- Scenario: What is the least number of times the university should conduct the survey to learn about change over the semester?
- Options: a. Once b. Twice c. Three times d. Four times
- Answer: b. Twice
- Rationale: To assess change, at least two time points are needed (e.g., Time 1 and Time 2).
Wrap-up
- Key takeaway: Research design choices depend on what you want to measure and how many times you collect data.
- Cross-sectional vs longitudinal: One-time snapshot vs two or more time points.
- Longitudinal designs offer several variants (Trend, Panel, Cohort, Cross-lagged) each with specific strengths, weaknesses, and cost considerations.
- Final guidance: Choose a design based on resources and the research goal (exploration, explanation, prediction, control, or causality).