Describe what makes a ‘good’ sample

Question 1

When a sample is drawn for a quantitative study, we usually have one (or both) of two objectives.

Answer 1

1. To estimate certain population characteristics (parameters), such as the average household size, the proportion of children vaccinated, the yield of corn, the height of trees etc. For this a sample is selected, measurements are made on the characteristic of interest (the variable), and these measurements are used to calculate our statistic, our estimate of the unknown population characteristic or parameter from our sample. As we shall see, given the uncertainty due to sampling, all estimates carry with them a certain margin of error, which reflects their precision
2. To test a statistical hypothesis about a target population, eg the hypothesis that at least 45% of the adult population of village X can read and write. Again, we will use the calculated statistic from the sample to see whether evidence contradicts or supports the hypothesis

Question 2

Statistical inference

Answer 2

The reasoning from the observed sample results back to the (unknown) population figures (parameters) is called statistical inference

Question 3

Sample qualities to allow inference

Answer 3

What are the qualities that a sample should have if one is to make inferences from it about a population? A good sampling method (or means of selecting a sample) is one that provides

1. a sample that is representative of the population from which it is drawn as regards the variables being studied and any relationships between them
2. • information about the likely accuracy of sample estimates of population characteristics (where accuracy is defined as the closeness of the sample estimate to the true population value).
3. the above at a reasonable cost.

The most important point to remember about sampling is that the manner in which the sample is drawn determines to what extent we can generalise our findings

Only if the sample studied can be shown to represent a larger population can the results of the sample be taken to give reliable information about the larger population

Question 4

What if the sample is not representative?

Answer 4

If the sample is not representative then the conclusions drawn from the research must be limited to the sample studied.