Unit 3 covers study design — how to collect data that allows valid conclusions. The distinction between experiments and observational studies determines what conclusions can be drawn.
Simple random sample (SRS): every individual equally likely to be selected. Stratified: divide population into strata, SRS within each. Cluster: divide into clusters, randomly select entire clusters. Systematic: every kth individual. Bias types: voluntary response (people self-select → extreme opinions), convenience (easy-to-reach → not representative), undercoverage (missing groups), response bias (wording influences answers), nonresponse (selected people don\'t participate).
Observational study: observe without intervention. Can show association, NOT causation (confounding variables possible). Experiment: researcher imposes treatments. Principles: control (control group), random assignment (to treatment groups), replication (sufficient sample size). Random assignment → causal conclusions possible. Confounding variable: associated with both the explanatory and response variables. Placebo effect: sham treatment accounts for psychological effects. Double-blind: neither subject nor evaluator knows treatment.
Random sampling is how individuals are selected FROM a population to participate in a study. It allows generalisation of results to the broader population. Random assignment is how participants IN an experiment are allocated to treatment groups. It allows causal conclusions (eliminates confounding by equalising groups). Best case: both random sampling AND random assignment → can generalise AND determine causation. Observational study with random sample → can generalise but NOT determine causation. Experiment without random sample → can determine causation within the sample but NOT generalise broadly.
Book a Trial + Diagnostic session. Get a personalized Learning Path with clear milestones, tutor match, and a plan recommendation — all within 24 hours.
Book Trial + Diagnostic →