2.3 Collecting Samples Types of Samples Random – occurring by chance 1) Simple Random Sampling All selections must have an equal chance of being chosen All combinations of selections must be equally likely Example – How can you choose 3 people from the class? >> put everybody’s name in a hat and draw 3 names >> use a TI83+ calculator Press MATH >>> PRB, choose 5:randInt(start,end,quantity) keep pressing ENTER for more choices 2) Systematic Random Sampling Used when sampling a fixed percent of the population Choose a random starting point and then you select every nth individual person for your 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝑠𝑖𝑧𝑒 study, where n is the sampling interval (found by evaluating 𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒 ) Example – How can you choose 20% of the class? >> do a simple random sampling for the starting point 23 23(0.2) >> determine n { = 23 4.6 = 5} >> from the starting point, pick every 5th person 3) Stratified Random Sampling The population is divided into strata (groups based on age groups, geography, gender, etc.) A simple random sample of the members is then taken The size of the sample for each stratum is proportionate to the stratum’s size Example – How can you choose 8 members from the class so males and females are fairly represented? >> Calculate: 𝑇𝑜𝑡𝑎𝑙 𝑚𝑎𝑙𝑒𝑠 𝑇𝑜𝑡𝑎𝑙 𝑐𝑙𝑎𝑠𝑠 ×8= 𝑇𝑜𝑡𝑎𝑙 𝑓𝑒𝑚𝑎𝑙𝑒𝑠 𝑇𝑜𝑡𝑎𝑙 𝑐𝑙𝑎𝑠𝑠 ×8= >> then do a simple random sample for each gender 4) Cluster Random Sampling Organize the population into groups (schools, communities, companies, etc.) Choose a random sample of the groups ALL members of the chosen group would be surveyed Example – How can you choose people from the school to survey? >> Do a random sample of all the homerooms for a particular grade using the master list >> Then visit each homeroom and survey all members 5) Multi-Stage Random Sampling Organize the population into groups A random sample of the groups is chosen Then a random sample of the members of the chosen groups is taken Example – How to pick 10% from the class? >> Groups the students by row and pick 2 numbers from a hat that represent row numbers (2 out of 5 rows = 40% of the class) >>Then pick 25% of the row randomly (25% of 40% of the rows = 10% of the class) 6) Destructive Sampling Samples from which the selected elements cannot be reintroduced into the population Example – light bulbs being tested for quality control – left on to burn until blown and time of its life is recorded Sample Size: How much is enough? - Size of the population impacts the appropriate sample size - Sample size is related to the reliability of the results o Variability of the population (greater variability = greater sample size) o Amount of precision required for the study o Sampling method chosen “The larger the sample, the better the results!” Homework: page 99 #1, 3, 5, 6, 8