Designing Experiments: Research Questions, Hypotheses, and Variables
In behavioral economics, rigorous experimental design is crucial for understanding human decision-making. This module focuses on the foundational elements of experimental design: formulating clear research questions, developing testable hypotheses, and identifying key variables.
Formulating Research Questions
A research question is the starting point of any experiment. It's a broad inquiry that guides your investigation. In behavioral economics, research questions often explore how psychological factors influence economic choices.
A good research question is specific, measurable, achievable, relevant, and time-bound (SMART), guiding your experimental focus.
Think about a specific behavior you want to understand. For example, 'How does the framing of a choice affect people's willingness to save?' This question is focused on a particular behavior (saving) and a specific influence (framing).
When crafting a research question, consider the phenomenon you're interested in and the potential underlying psychological mechanisms. It should be open-ended enough to allow for exploration but focused enough to be addressed through an experiment. Avoid questions that are too broad or too narrow. For instance, 'Do people make good decisions?' is too broad, while 'Does a blue button increase donations by exactly 5%?' might be too specific for an initial question.
Developing Hypotheses
A hypothesis is a specific, testable prediction about the outcome of your experiment. It's a statement that proposes a relationship between variables.
| Characteristic | Research Question | Hypothesis |
|---|---|---|
| Nature | Inquiry, broad | Prediction, specific |
| Purpose | Guides exploration | States expected outcome |
| Format | Question (e.g., 'How does X affect Y?') | Declarative statement (e.g., 'X will increase Y') |
Hypotheses can be directional (predicting the direction of an effect, e.g., 'Framing a choice positively will increase saving') or non-directional (predicting an effect but not its direction, e.g., 'Framing will affect saving'). In behavioral economics, directional hypotheses are common as we often draw on existing theories about cognitive biases.
A research question is an inquiry, while a hypothesis is a specific, testable prediction.
Identifying Variables
Variables are the factors that you manipulate or measure in your experiment. Understanding the types of variables is essential for designing a controlled and interpretable study.
In experimental design, we distinguish between the Independent Variable (IV) and the Dependent Variable (DV). The IV is what the researcher manipulates or changes to observe its effect. The DV is what the researcher measures to see if it is affected by the IV. For example, in a study on framing effects on saving, the 'framing of the choice' (e.g., presented as a gain vs. a loss) would be the IV, and the 'amount saved' would be the DV.
Text-based content
Library pages focus on text content
Control variables are factors that are kept constant to prevent them from influencing the dependent variable. Confounding variables are those that are not controlled and can potentially influence the dependent variable, leading to misleading results. Identifying and controlling for these is a key aspect of robust experimental design.
A well-designed experiment clearly defines its independent, dependent, and control variables to isolate the effect of interest.
Putting It All Together: An Example
Let's consider an experiment on the 'endowment effect' – the tendency for people to value something more once they own it.
- Research Question: How does ownership influence the valuation of a mug?
- Hypothesis: Participants who are given a mug will value it more highly than participants who are not given a mug.
- Independent Variable (IV): Ownership status (Owner vs. Non-owner).
- Dependent Variable (DV): Stated willingness to pay (WTP) for the mug.
- Control Variables: Type of mug, instructions given to participants, time of day, participant demographics (if controlled for).
Key Takeaways
Mastering the art of formulating research questions, crafting testable hypotheses, and identifying variables are fundamental skills for any aspiring behavioral economist. These elements form the bedrock upon which sound experimental designs are built, leading to reliable insights into human behavior.
Learning Resources
Provides a concise overview of behavioral economics, touching upon its core concepts and experimental approaches.
An article from the American Economic Association discussing the basics of experimental economics and its importance in research.
A lecture from a Coursera course that breaks down the principles of experimental design, including variables and hypotheses.
Khan Academy's comprehensive guide to hypothesis testing, covering null and alternative hypotheses and their statistical implications.
A guide from Purdue University Libraries on how to formulate effective research questions for academic inquiry.
Explains the endowment effect, a key concept often studied using experimental methods in behavioral economics.
A detailed explanation of independent, dependent, and control variables in the context of experimental research.
A seminal paper discussing the application and importance of field experiments in economic research, often involving behavioral aspects.
A foundational book in behavioral economics that illustrates many concepts through experimental findings and choice architecture.
A video explaining psychological biases that influence decision-making, often explored through experimental paradigms.