Chat with our AI personalities
Random error and sample size have an inverse relationship...As sample size INCREASES random error DECREASES. There's a good explanation at the related link.
Yes. If the sample is a random drawing from the population, then as the size increases, the relative frequency of each interval from the sample should be a better estimate of the relative frequency in the population. Now, in practical terms, increasing a small sample will have a larger effect than increasing a large sample. For example, increasing a sample from 10 to 100 will have a larger effect than increasing a sample from 1000 to 10,000. The one exception to this, that I can think of, is if the focus of the study is on a very rare occurrence.
The margin of error is reduced.
... should be increased by a factor of 4. Note that this implies that the only errors are statistical (random) in nature; increasing the sample size won't improve systematic errors.
It is the number of elements in the sample. By contrast, the relative sample size is the absolute sample size divided by the population size.