Lessons for Topic 1.13
Blocking: randomized block and matched pairs designsGrouping similar units first so a known source of variation does not hide the treatment effect.Lesson
Completely randomized design and replicationAssigning every unit to a treatment purely at random, with no grouping first.Lesson
Control groups, placebos and blindingA group receiving no treatment or the usual one, to compare the others against.Lesson
Units, treatments, factors and levelsWhat or who receives a condition, and the conditions being compared.Lesson
Observational studies and experimentsWatching versus assigning, and what each can and cannot conclude.Lesson
Random assignment and confoundingUsing chance to decide which unit gets which treatment, so the groups start out alike.Lesson