denoted by independent continuant study group population size experimental unit The smallest unit of an experiment to which a treatment is *independently* applied. It is therefore the entity that receives a treatment (or condition) and whose response is assumed to be independent of other units. The experimental unit defines how many independent pieces of information will be available for inference. It determines the sample size (n) for hypothesis testing and confidence intervals. Mistaking smaller components (like individual plants or students) for independent units can lead to pseudoreplication — overestimating the true sample size and producing misleadingly small p-values. - a plot of land where fertilizer is applied per plot, not per plant. - a patient in a clinical trial when each patient receives a treatment. - a class (not each student) is the experimental unit where a teaching method is assigned per class. STATO Philippe Rocca-Serra ready for release