As defined in the text random sampling requires sampling with replacement?

Answer 1

Not necessarily. A random sample can occur with or without replacement, depending on what makes more sense. For instance, trying to calculate the odds of a dice roll would require a random sample with replacement (because it is perfectly possible to get a 6 on each and every die); trying to calculate the odds of a Poker hand, however, would require random sampling without replacement (the ace of spades can only show up once in any given round of dealing). when the population size is large enough, the difference between the two is meaningless; people who make national surveys, for instance, usually choose people randomly without replacement (there's no possibility they will survey the same person twice) but treat it as though the were sampling with replacement (because the math is easier). The only requirement for a random sample is that each object that might be chosen has a known and well-defined probability of being chosen at any given moment. For random samples with replacement that probability is always the same; For random samples without replacement that probability is determined by the objects that have previously been selected.