Research Design Types

Why This Matters

Your research design is the architectural blueprint for your entire AP Research project—it determines what kind of evidence you can collect, what claims you can make, and how convincing your argument will ultimately be. The College Board expects you to not only choose an appropriate design but to justify that choice in your paper's methodology section. Understanding the strengths and limitations of each design type helps you anticipate examiner questions and defend your approach during your oral defense.

Different designs answer different kinds of questions. Some let you establish causal relationships (Did X cause Y?), while others reveal patterns and associations (How does X relate to Y?). Still others provide depth and context (What does X mean in this particular situation?). Don't just memorize these design types—know which research questions each design can answer, what trade-offs you're accepting, and how to articulate why your chosen design fits your inquiry.

---

Designs for Establishing Causation

When your research question asks whether one thing causes another, you need a design that controls for alternative explanations. These designs prioritize internal validity—the confidence that your independent variable, not some confounding factor, produced the observed effect.

Experimental Design

• Random assignment of participants to treatment and control groups—this is what distinguishes true experiments from other designs and allows you to rule out confounding variables
• Manipulation of the independent variable under controlled conditions lets you systematically test cause-and-effect relationships
• Strongest design for causal inference, but often impractical or unethical for many AP Research topics involving human subjects or real-world phenomena

Quasi-Experimental Design

• Lacks random assignment, making it less rigorous than true experiments but more feasible in real-world research contexts
• Used when randomization is impractical or unethical—for example, you can't randomly assign students to different school systems or communities to different policies
• Still valuable for causal insights when combined with careful analysis of potential confounding variables and transparent acknowledgment of limitations

---

Designs for Tracking Change Over Time

Some research questions require understanding how phenomena develop, evolve, or persist. These designs address temporal relationships—the sequence and duration of events that help distinguish cause from correlation.

Longitudinal Design

• Repeated observations of the same subjects over weeks, months, or years—essential for tracking developmental changes or long-term effects
• Establishes temporal sequence, which strengthens causal arguments by showing that the cause preceded the effect
• Resource-intensive and time-consuming, which may make it challenging for the AP Research timeline unless you use existing longitudinal datasets

Cross-Sectional Design

• Data collected at a single point in time from different subjects or groups—a snapshot rather than a movie
• Efficient for identifying patterns and relationships among variables, making it practical for most AP Research projects
• Cannot establish causation because you lack information about which variable came first; useful for exploratory research or when time constraints exist

---

Designs for Depth and Context

When your research question requires understanding complexity, meaning, or context rather than broad generalizations, these designs prioritize rich qualitative data—detailed information that captures nuance and particularity.

Case Study Design

• In-depth analysis of a single case or small number of cases—ideal for exploring complex, context-dependent phenomena that surveys can't capture
• Provides rich qualitative data through multiple sources like interviews, documents, and observations within one bounded context
• Limited generalizability to broader populations, but strong for generating hypotheses or illustrating theoretical concepts

Observational Design

• Systematic recording of behaviors or events in natural settings—can be structured (with predetermined categories) or unstructured (open-ended exploration)
• Captures real-world dynamics without the artificial constraints of laboratory settings or self-report biases of surveys
• Researcher presence may influence behavior (the Hawthorne effect), requiring careful consideration of your role and potential bias

---

Designs for Breadth and Patterns

When your research question requires data from many participants or seeks to identify broad patterns, these designs prioritize external validity—the ability to generalize findings beyond your specific sample.

Survey Research Design

• Questionnaires or interviews with large samples—efficient for collecting quantitative data on attitudes, opinions, and self-reported behaviors
• Requires careful design including attention to sampling method, question wording, and response bias to ensure validity and reliability
• Strong for generalizability when using representative sampling strategies, but limited by participants' willingness and ability to accurately report

Correlational Design

• Examines relationships between variables without manipulation—useful for identifying patterns and associations in existing data
• Cannot establish causation—correlation does not imply causation is a phrase you'll need to use when discussing limitations
• Often serves as preliminary research to identify relationships worth investigating through more rigorous experimental designs

Comparative Design

• Systematic comparison of two or more groups or cases—useful for understanding how context, policy, or other factors influence outcomes
• Can be qualitative or quantitative depending on your research question and the nature of your comparison
• Strengthens arguments through contrast—showing how different conditions produce different results illuminates underlying mechanisms

---

Designs for Comprehensive Understanding

Some research questions benefit from combining approaches. These designs prioritize triangulation—using multiple methods to cross-validate findings and address the limitations of any single approach.

Mixed Methods Design

• Combines qualitative and quantitative approaches in a single study—for example, surveys followed by interviews, or case studies with statistical analysis
• Leverages strengths of both paradigms to provide comprehensive understanding that neither approach could achieve alone
• Requires explicit justification for how the methods connect: Does qualitative data explain quantitative patterns? Does quantitative data test qualitative themes?

---

Quick Reference Table

---

Self-Check Questions

1. Your research question asks whether a new teaching method causes improved student outcomes, but you cannot randomly assign students to classrooms. Which design should you use, and what limitations must you acknowledge in your methodology section?

2. Compare and contrast longitudinal and cross-sectional designs: What can each establish about the relationship between variables, and when would feasibility concerns push you toward one over the other?

3. A classmate claims their correlational study "proves" that social media use causes anxiety. What's wrong with this claim, and what additional research design would strengthen their argument?

4. Which two designs would you combine in a mixed-methods study examining why students choose certain career paths, and how would each component contribute to answering your research question?

5. You're studying a rare phenomenon that has only occurred once in your community. Which design is most appropriate, what are its strengths for your situation, and how would you address concerns about generalizability in your oral defense?