Results (
Indonesian) 1:
[Copy]Copied!
whole population, as the population size is quite small (14 students thisyear). Sampling the whole population is known as a census. It is also possibleto (attempt to) sample even a large population given enoughresources. Governments, for example, regularly conduct a census of theirpopulation, although, as recent examples in the United States and theUnited Kingdom have shown, this process is not without problems andnot all members of the population are actually reached. In most cases,we do not have the resources to study the whole population and willneed to sample. It is important to remember that we can only generaliseto a population we have actually sampled from. And therefore somethought about exactly what our population is going to be is warranted.Decide how to sample from the populationIn most cases we will need to take a sample from our population. We willthen usually want to generalise the results we find in our sample to ourpopulation. After all, a survey of the voting intentions of a sample of1,000 people would not be very useful if we couldn’t generalise our findingsfrom that sample to voters as a whole! In order for us to be able togeneralise, we need to have an unbiased sample of the population, whichmeans that we want our sample to be representative of the populationwe are studying, and not skewed towards one group or another. If wewere trying to generalise to all 10-year-olds, for example, we wouldn’twant to sample only all-girls schools. The best way of ensuring that oursample is unbiased is by using probability sampling methods.The most well-known of these is simple random sampling. In a typicalsimple random sample everyone in the population has exactly the samechance of being included in the sample. This is because the sample isdrawn at random from the population (for example, by putting names ina hat or, more typically nowadays, by using random number generators).That makes it the most unbiased form of sampling, and this is themethod used to draw lottery numbers, for example. Saying that this isthe most unbiased sampling method would suggest that it is a good ideato attempt to use simple random sampling at all times. However, whenone looks at actual educational research, it is clear that the majority ofstudies do not in fact use this method. Why is this? There are a numberof reasons, some good, some less so.One good reason is that while simple random samples are excellent forgeneralising to the population as a whole, we might in some cases want
Being translated, please wait..
