Sampling+topic+page

=__Calculator Keys (Casio)__= e.g. numbers from 5 + 20 = 5 + 15 Random. =__Key definitions:__=
 * Into stats mode** - press mode button until you see SD. Select SD
 * Clear Memory** - shift, mode, choose Stats clear
 * Enter data into memory** - use M+ button
 * Find the mean** - shift, 2, select the mean (x with line over it)
 * Random number generator** - shift, decimal point. (this generates numbers from 0 up to 1)
 * Generating numbers from 1 to n** - 1 + n Random
 * Generating numbers from n to another number** - n + (range of numbers) Random

Population
- the group being studied. e.g. surveys on election votes target the population "registered voters". Populations can also be groups of trees, animals, cars, etc. The population is the ENTIRE group you want to know about.

Sample
- the part of the population you actually use in your study. e.g. a survey on election votes might only be asked of 1000 people, even though there are many more registered voters than that.

Bias
- bias is a fault in the METHOD of selecting the sample. It means that some have an unfair advantage, a greater chance of being selected than others.

Representative
- means that the sample has similar proportions to the population. An example of a non-representative sample would be if a population was 20% Maori, but the sample only had 5% Maori. An unbiased method can sometimes deliver a non-representative sample, but good sampling makes this far less likely =__Describing and Justifying Sampling methods__=

Random Sampling - Using random numbers
To describe the method, you should mention that: To justify this method:
 * population is numbered
 * Random numbers generated by calculator are used to select members of the population
 * Say how big your sample is
 * What you enter to get these numbers e.g. 1 + 300 Random
 * That you ignore repeats of numbers and decimal places.
 * It is simple to use, as population is numbered
 * It is unbiased
 * Population should show no sign of being in definite groups.

Systematic Sampling - every nth term
To describe this method you should mention To justify this method
 * How big your population/sample are
 * How you calculated your interval. e.g. a sample of 50 from population of 200 you divide 200 by 50, so that you take **every 4th** item.
 * How you chose the first number of your sample (should be generated at random)
 * It is unbiased
 * It is quick/time efficient
 * If the population is in order, it makes sure you get values from every part of the distribution.

Stratified Sampling - a certain number from each group
To describe this method you should mention If one group is numbers 20-140 of the population, put 20 + 120 Random on your calculator so that only items from this group will be selected. Justifying this method .
 * what proportions there are in the population. e.g. 48% female
 * a calculation of how many from each group should be in the sample. e.g. in a sample of 50, 48% female translates to 24 females in sample.
 * Your method for stratifying. You could use a random method, and then reject numbers that relate to a group you have met your quota for. If your population lists one group at a time, you can do more advanced random sampling. For example...
 * the population contains clear groups. Not representing these groups well may skew the data
 * this method ensures all groups are proportionally represented