Statistics Unit 4- Designing Studies and Collecting Data

population

entire group of individuals we want information about

sample

subset of individuals in the population from which we collect data

convenience sample

sampling individuals who are easy to reach (biased)

voluntary response

people choose to be in the sample (biased)

simple random sample (SRS)

every group is equally likely to be chosen (unbiased)

to conduct- 1. label 2. randomize 3. select

stratified random sample

splits population into groups (strata) and chooses an SRS from each group

• more specific/precise

cluster sampling

split groups based on location (cluster) and randomly select clusters, sample everyone in that cluster

• easier to take sample

undercoverage

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

non-response

when as individual is selected to be in a sample but chooses not to participate or isn’t reached.

response bias

pattern of inaccurate responses

• could be due to wording, interviewer, lying, etc.

observational studies

no treatments imposed, we only observe

experiments

treatments imposed, allows for causation

• needs to be executed in order to confirm/ conclude causation

explanatory variable

used to predict

response variable

outcomes

treatments

what is done to participants

experimental units

what/who the treatment is imposed on

confounding variable

variables not controlled, that affects results

control group

used to provide baseline data for comparison (can include a placebo)

blinding

when subjects (single blind) and/or experiments (double blind) who interact with subjects are unaware of what treatment was assigned

placebo effect

when a fake treatment works

4 key principles of experiments

1. comparison: 2 or more treatments

2. random assignment: use a chance process to assign treatments to experimental units

3. control: keep all other variables besides treatments constant (not the same as the control group)

4. replication: using enough experimental units to distinguish differences

completely random design

1. sample size

2. random assignment

3. what are the groups?

4. treatment types

5. compare data

block

group of experimental units that are known to be similar

randomized block design

separate subjects or blocks and then randomly assign to treatments within each block.

matched pairs

experiment used to compare two treatments that uses blocks of size 2. two very similar experimental units are paired and then randomly assigned to a treatment