Experimental design is the planned way of investigating how changing one part of a system affects another. This lesson introduces the purpose of design of experiments (DOE), the main types of variables, and the difference between experiments and quasi-experiments.
Design of experiments (DOE) is the construction of procedures intended to explain how changes in one aspect of a system lead to changes in other aspects. Researchers decide what to change, what to measure, and what to hold constant, guided by hypotheses about the sources of variation.
All flashcards
What is design of experiments (DOE)?
A planned procedure for studying how changing selected aspects of a system affects measured outcomes, while managing other sources of variation.
What is an independent variable?
A factor treated as an input or predictor. In an experiment, the researcher may manipulate it to assess its effect on an outcome.
What is a dependent variable?
The output or response measured to determine whether it changes as the independent variable changes.
What is a control variable?
A factor kept constant or otherwise managed so it is less likely to provide an alternative explanation for the result.
How does a quasi-experiment differ from a controlled experiment?
A quasi-experiment studies conditions that are naturally occurring or cannot be fully assigned by the researcher, so causal interpretations may be more limited.
Quiz questions
What does random assignment help researchers do?Answer: Balance known and unknown confounders across conditions on average.
What can a multifactorial experiment reveal that a one-factor-at-a-time approach may miss?Answer: Whether one factor's effect depends on another factor's level.
Which practice is p-hacking?Answer: Trying many analyses and reporting favorable results without disclosing that flexibility.
Which statement best distinguishes validity from reliability?Answer: Validity concerns support for the intended conclusion; reliability concerns consistency.
A study finds that people who exercise more tend to sleep better, without assigning exercise levels. What conclusion is justified?Answer: Exercise and sleep are associated, but the association alone does not establish causation.