Consider each of the statements below. For each statement, decide whether it is sometimes, always, or never a true statement.
1. In a hypothesis test, the direction of the alternative hypothesis (right-tailed, left-tailed, or two-tailed) affects the way the effect size is estimated.
2. In order to compute Cohen's d, a statistician must directly know the sample size.
3. A hypothesis test that produces a p-value < 0.001 will produce an effect size |d̂| > 0.8
4. If Marcus takes a random sample of a population and conducts a hypothesis test and Caroline takes a different random sample of the same population and conducts the same hypothesis test, they will have the same observed effect sizes.
5. A hypothesis test of a single population mean that produces a t-test statistic t = 0 will produce an effect size d̂ = 0.