sampling bias
representative sample
biased sample
representative sample
a sample that closely represents the population
#inclusitivity
biased sample
some people have a better chance of being represented in the population
- not all groups are represented/some groups are overrepresented in the sample compared to general population
probability sampling
- every individual has a chance of being selected
- random selection
- every individual in targeted population is identified
nonprobability sampling
does not meet one of the criterias from probability sampling
simple random sampling
probability sampling in which every member of the target population has an equal chance of being selected
- if there are one million people, everyone has a one in a million chance of being selected
systematic sampling
- old fashion approach
1. order all individuals in the population
2. pick a random starting point
3. select each nth individual
example: knocking on every fifth door in the neighborhood
cluster sampling
researcher is interested in studying PSYC 70 students
let’s say there are 2,000 PSYC 70 classes UC wide
cluster sampling is randomly selecting from a few UCs but not all locations
stratified random sampling
- purposefully select particular demographic categories
- randomly select individuals within that category
- proportionate to their membership in the population
example:
survey = campus climate @ UCSD
1. take students by different racial ethnic groups
2. select a proportionate amount of people from each group
- group 1: 5 members = take 1 person
- group 2: 10 members = take 1 person
- group 3: 5 members = take 1 person
cluster vs stratified sampling
cluster
- each cluster is heterogenous
- includes a variety of people
- some clusters are randomly sampled, others are eliminated
stratified
- each stratum is homogenous
- includes one kind of person
- all strata are proportionately sampled
convenience sampling
sampling individuals who are easily accessible
quota sampling
- makes sure to represent all groups
- similar to stratified sampling
- selection is NOT random
- all individuals may or may not be identified
snowball sampling
asking participants to recruit other participants who are similar to them
-useful when you don’t have easy access to a particular group
sample size and sampling error larger samples are more representative. true or false
false. larger samples reduce sampling error but dont reduce sampling bias (systematic error)
