Study Types
The variable that is the measured outcome is called the ___ variable
Response
Every group of n individuals has an equal chance of being chosen
Simple random sample (SRS)
"Don't you agree that our amazing principal deserves a raise?"
Question wording bias
The people or things that treatments are assigned to
Experimental units
A fake treatment, like a sugar pill
Placebo
Random SAMPLING lets us ___ results to the whole population
Generalize
The variable that explains or predicts the outcome is called the ___ variable
Explanatory
Picking people you think "look representative"
Judgment sample
People lie about how much they exercise
Self-reporting bias
Assigning treatments to subjects by chance (like flipping a coin)
Random assignment
The group that does NOT get the real treatment
Control group
Random ASSIGNMENT lets us make a ___ conclusion
Causal (cause-and-effect)
Measuring EVERY individual in a population is called a ___
Census
Picking a random starting point, then choosing every 10th person in line
Systematic random sample
An online poll lets anyone who wants to vote
Voluntary response bias
Using many subjects in each treatment group
Replication
Feeling better just because you think you got the real treatment
Placebo effect
Random sampling reduces ___ (bias or confounding?)
Bias
Researchers watch and record what people do without assigning any treatments. Is this an observational study or an experiment?
Observational study
A principal randomly picks 15 students from EACH grade level. Cluster or stratified?
Stratified
Surveys are mailed to 500 randomly chosen people, but only 60 send them back
Nonresponse bias
Keeping other variables the same for all groups, like time of day or room temperature
Direct control
The subjects don't know which treatment they got, but the researchers do. Single-blind or double-blind?
Single-blind
Random assignment reduces ___ (bias or confounding?)
Confounding
Researchers randomly assign half the volunteers to drink coffee and half to drink water. Is this an observational study or an experiment?
Experiment
A teacher randomly picks 3 classrooms and surveys EVERY student in them. Cluster or stratified?
Cluster
A phone survey only calls landlines, so people with only cell phones are left out
Undercoverage bias
A variable that is linked to both the explanatory and response variables and makes it hard to tell what caused what
Confounding variable
Neither the subjects NOR the researchers know who got which treatment. Single-blind or double-blind?
Double-blind
A study used volunteers but randomly assigned treatments. Can we generalize? Can we say cause-and-effect? (two answers)
No to generalize; yes to cause-and-effect