bambam

Experimental Psychology
2005-02-16 17:51:08 (UTC)

Hill 2-16

Sampling

-Must select participants from population in any research

-Population vs Sample
-Population-true interest, but limited (e.g., size &
stability)
-E.g., All college students in U.S.: All students in
experimental class
-Sample-subgroup of group

-Representativness vs Bias
-Goal-sample to reflect attributes of target population as
a whole
-Bias-self-selection problem

-Convenience sample vs Probability sample
(psych info usa) vs (any student)

-Sample selection vs sample assignment
(population) (where to assign)

Sampling Stratagies
-Simple random sampling
-every member of population has = chance of being selected
-Advantage:effective and practical - represenative
-disadvantage:might want some feature of population in
sample, impractical w/ extremly large populations

-stratified sampling
-protect subgroup %s; can lead to a variety of layers
-E.g.,? abut attitudes on abortion;FM in population
-Advantages; helps with 1st issue of simple random
sampling
-Disadvantage; cant deal with large populations either

-Cluster Sampling
-use existing small groups resembling population then
randomly select from groups withen common features for
sample

Types of Samples

-Independant samples
-groups created with simple and stratified sampling
-non equvilant groups not related
-Also include groups created with organismic variables
(e.g., gender vs. age, ethnicity)

-Dependant Samples (related/correlated)
-repeated measures-same subject(Ss) serve in both
conditions
-Matched pairs-matched on some aspect (e.g., GPA trait)
-Natural pairs -e.g., parents and children: husbands and
wives
N paramater population
n statistic samples


Ad:0