Correlation and causation
5. The more it rains, the fuller the reservoir gets. → First causes second — Rain runs into the reservoir; the reservoir cannot make it rain.6. The more hours a fridge runs, the more power it uses. → First causes second — Running is what draws the power, not the other way around.Survey, experiment or study
1. A town phones 500 households to ask about a new library. → Sample survey — Questions to a sample, and no treatment given: a sample survey.2. Half the plants, chosen at random, are given the new fertilizer. → Experiment — The researcher ASSIGNS who gets the treatment, at random: an experiment.3. Half the aisles, picked at random, get the brighter lighting. → Experiment — The aisles were assigned their lighting at random: an experiment.4. 200 randomly chosen students are asked how they get to school. → Sample survey — They are ASKED a question. Nothing is done to anyone: a sample survey.5. Runners who already train daily are compared with those who do not. → Observational study — The groups were ALREADY different; nobody assigned the training.6. Lung capacity is measured in people who already cycle to work. → Observational study — No one was told to cycle; the habit came first: an observational study.Margin of error
1. 2,500 fans were asked; 48% watched the match, give or take 2%. What is the lowest the true figure could be? → 46% (48% − 2% = 46%)2. 25 voters were asked about the plan. What is the margin of error? → 20% (1 ÷ √25 = 1 ÷ 5 = 20%)3. 400 users were asked; 66% have tried the app, give or take 5%. What is the lowest the true figure could be? → 61% (66% − 5% = 61%)4. 2,500 shoppers were asked about the new label. What is the margin of error? → 2% (1 ÷ √2,500 = 1 ÷ 50 = 2%)