Some Statistical Definitions
When determining what is to be studied, it is important to know what the population to be studied is.
Is it going to be all cats, all long haired cats, or all curly haired dogs?
To study the whole population you would use a census.
However, most of the time, the population is too large, so a sample is chosen.
A sample will be a subset of the whole population. There are several Sampling Techniques, which can also be combined.
- Cluster: the sample is collected based on the location, e.g. all of one zip code.
- Convenience: the sample is convenient to the person conducting the study, e.g. people who walk into the grocery store.
- Random: each person, thing, etc. has an equal chance of being included in the sample, e.g. pulling a number from a hat.
- Stratified: the population is sorted into groups and then a percentage of each group is selected based on the population, e.g. 10% of the sample will be 40 year olds.
- Systematic: every nth item or person is selected, e.g. the 5th caller to the radio station.
When collecting the sample it is important to avoid sampling bias. This is a bias that occurs when the sample does not represent the population. It may represent what the person conducting the study wants to find.