Sampling Fundamentals PDF

Ch 14 sampling Fundamentals. Introduction to sampling * * ↓ sampling: Obtaining data from r aubset of a...

Ch 14 sampling Fundamentals. Introduction to sampling * * ↓ sampling: Obtaining data from r aubset of a larger group kamp)Opt populationseye - tion * Census : Obtaining data from every member of the population - entire population is sampled - ex U S -. cenos. every 10 years Final Census vs sampling Objective · Find proportion of people that has watched the TV Now" Stranger Things" in ATL , Sa How to achieve this? Census Approach we would send · a may to "everybody" living in A+ 1 , Ga The proportionop f opomwopopulati · the Parameter · send samplingAppro a Households , in At · The proportion we find that har watched th Now from the sample is called a statistic  * Parameter , statistical value from population # tatistic , tatifical value from spte * We typically measure sample not population Final When is the census approach appropriate? Wheninformationisrequired or needed" for easy · When population is man because researchers would be · able to afford to my every person when consequences of making wrong decisions are · very cotly , so it isbetter to ourvey everyone to reduce uncertainty When · sampling error is high. If you very on just sample it may lead to a a tatific with high divergence from the true value of population When is a app appropriate sampling information · When cost and timing to get from the population ahA every perion have high car · when population size is large be it would not be afforable to urvey everyone · When decisions based on the new information name to be made quickly · when sampling errors are low , adequate representative can be accieved ↑ to the true value of the When tatistics are close  I sampling Types for · sampling error : difference bu a meanum Obtained from a sample of population and the true measure that can be obtained only from the Population non sampling error : all other errors associated with · research project Sampling Process Population MFerment Ta potential list of individuals from where you can draw a sampl X select correct Procedure based on your godis  Step 1: Identify Target Population make the your target population is relatent to · our bereaven objective · consider all alternative populations you can we What is the · appropriate unit of analysis ex. households , individual, neighbornoar ,zipcoder · make we to not ever defin the perlation and consider convience Example · A target population for a videogame Fort has been defined ar > - "all households with children living in ATL > tatement is too - ambigous * How should we define children ? Are they below to yur old , 13 years old or le you old * How do we define Atlanta? Only the metropolitan area or should we include ububs ? * who in the household will be in information Provider Step 1 : Determine sampling Frame · List of that population members to draw could be used a rampt · very rately will the ple frame match the target Population completely Tm will be some individuals that. meet the population target criteria and other won't  The goal is the ampling frame choosen is close a · possible to the target population required > this ensure - selection bias and ener representationer example The videogame store want to urvey "all individuals who · enjoy rose playing gamer" · How could a sampling Frame be developed ? 1) List of all customers who have bought RPG's in the part 2) List of all customers who have vented RPG's in the part 3) Lit of an cutomes who have both rented and bought RPG" in the past - more complete but maybe not completey accurate 2) sampling Franc Issues · sampling fram exists when the ampling frame ales not match the target population #1 That are notpart of the target population and neglects other that houd be included Step 3 : Choose sampling Procedure 1) 2) # i * probability sampling 1) simple random sample : each member of the population har equal probability of being relected 2) systematic random sample : each member is sampled in a yotematic way. Cluster 3 sampl : population is partitioned into a > - > - utuallyexclusiveCute ran selected clusters are all selected members in the Proportionate 4) stratified sample : population is partioned into exclusive Strata trally -elements are randomly rejected from each stratum - Diff bru cluster and Strata is that clusters can be. heterogen our meaning they can have similiar individual While Aratai are nomogenour > meaning individuals in - a stratum nave similiar characteristics Step 2: Stlatify them into homogenous groups to create the fratum 4. I Disproportional stratisfied sample · population is partioned into mutually exclusive strata > elements are randomly selected from each satuch w/o - Paying attention to the proportion of each Arata 2 1. I green for I red minority groups are valued · more than respondent the other · now its red respondents for ↑ green respondent  Probability sampling Advantage · more certainty of having representativeness of the Population · sampling error can quantified be · results are more generalizable to the total population Disadvantage · more expensive (morefrict wher of selection imprying more time and cost · take more time to execute non probability sampler Non probability sampling researchers typically we university · rbjeq Audents as part of th Pool based on uggestions