Unit 3 (Collecting Data)

a population is...

the entire group of individuals we want information about

a sample is...

the subset of individuals in the population from which we collect data

in a convenience sample, you...

sample individuals who are easy to reach

are convenience samples a good or bad type of sampling?

bad

in a voluntary response, people...

choose to be in the sample (biased)

is voluntary response a good or bad type of sampling?

bad

in a simple random sample, or SRS, every group...

is equally likely to be chosen

what are the steps to taking an SRS?

1) label

2) randomize

3) select

is an SRS a good or bad type of sampling?

good

in a stratified random sample, you...

split the population into groups (aka strata) and choose an SRS from each

does a stratified random sample involve heterogeneous or homogeneous groups

homogeneous

in cluster sampling, you...

split the population into clusters and randomly select, then sample all in the chosen cluster(s)

key phrase: "and then interview all from the group"

does cluster sampling involve heterogeneous or homogeneous groups

heterogeneous

in a systematic random sample, you...

randomly choose an individual, then sample every nth individual

benefits of SRS sampling

unbiased, low variability

benefits of cluster sampling

easy to take sample

benefits of systematic sampling

doesn't need to label everyone, but ALL groups are represented

undercoverage happens when...

some members of the population cannot or are less likely to be chosen

nonresponse occurs when...

an individual is selected to be in a sample, but chooses not to participate or isn't reached

a response bias is a...

pattern of inaccurate responses due to wording, interviewer, lying, etc

in an observational study...

NO treatment is imposed, only observation

in a experiment...

treatment IS imposed

if there is more than one explanatory variable, then they are called...

factors, which have levels

treatment is...

what is done to the participants

experimental units are...

who/what receives treatment (often called "subjects" if it's people)

confounding variables are...

variables not controlled for that affect the results

control groups are used to...

provide baseline data for comparison

blinding is...

not revealing what treatment is assigned

in a single-blind...

subjects are unaware that treatment was assigned

in a double-blind...

subjects and those administering treatment are both unaware that treatment was assigned

the 4 key principles of experiments

1. Comparison (2 or more treatments)

2. Control

3. Random Assignment

4. Replication (have enough experimental units to distinguish differences)

a block is...

a group of experimental units that are known to be similar

in a randomized block design, you...

seperate subjects into blocks, then randomly assign treatments within each block

matched pairs are...

an experiment that uses blocks of size 2 (2 similar subjects are paired, then randomly assigned to different treatments)

an inference draws...

influence based on experimental results

a random sample allows...

generalization to whole population

a random assignment allows...

us to make cause-and-effect conclusions

data ethics (3)

1. Institutional Review (ethical)

2. Informed Consent

3. Confidential