1/46
Looks like no tags are added yet.
Population
The entire group being studied or described
Sample
A smaller group selected from the population
Parameter
A numerical value describing a population
Statistic
A numerical value describing a sample
How do you tell if a value is a parameter
it describes the entire population
How do you tell if a value is a statistic
it describes a sample
Qualitative data
Data consisting of labels or categories
Quantitative data
Data consisting of numerical measurements or counts.
Observational study
Researchers observe subjects without assigning a treatment
Expirement
researchers assign treatments to subjects and measure the results
Treatment
A condition or intervention assigned to subjects in an experiment
Control group
A group used as a comparison in an experiment
Experimental unit
The individual or object receiving a treatment
Explanatory variable
The variable that may explain or cause the change
Response Variable
The outcome that is measured in a study
Factor
The variable being tested in an experiment.
Treatment combination
A specific combination of levels of the factors
Completely randomized design
subjects are randomly assigned directly to treatment groups
Randomized block design
subjects are divided into similar blocks, then randomly assigned treatments
Block
A group of similar subjects is used to reduce variation
Matched-pairs desgin
An experiment where subjects are paired based on similar characteristics, then each pair receives different treatments.
Single-blind
Subjects or researchers, but not both, don't know the treatment.
Double-blind
Neither the subjects nor researchers know the treatments.
Stratified sample
Divide the population into groups, then randomly sample from each group.
Cluster sample
Randomly select entire groups and survey everyone in selected groups.
Systematic sample
Select subjects using a fixed interval, such as every 3rd person.
Convenience sample
Choose subjects who are easiest to reach.
Undercoverage
Some groups in the population are inadequately represented or missed.
Nonresponse bias
Bias caused when selected people do not respond to the survey.
Response bias
Bias caused by inaccurate or dishonest responses.
Sampling bias
Bias caused by a sampling method that doesn't represent the population.
Why can't sampling methods completely eliminate bias?
Sampling methods can reduce bias but cannot guarantee no bias.
What does randomization do?
Helps reduce bias and balances differences between groups.
What does blocking do?
Groups similar subjects to reduce variation.
What is the purpose of a control group?
To provide a comparison for evaluating the treatment.
What is the purpose of random assignment?
To make treatment groups similar and reduce confounding.
What is a placebo?
An inactive treatment designed to resemble the real treatment.
What is a confounding variable?
A variable that affects results and is mixed with the treatment effect.
What does “compare” mean in an experiment?
Examine differences in the response between treatment groups.
What makes a study an experiment?
Researchers actively assign a treatment.
What makes a study observational?
Researchers only observe or ask; they don't assign treatment.
What is sampling variability?
Natural differences between statistics from different samples.
Statistic vs. parameter shortcut
Statistic = sample; parameter = population.
“Ask your 18 friends” = what method?
Convenience sampling.
“Five countries, sample from each” = what method?
Stratified sampling.
What is the key difference between blocking and stratifying?
Blocking is for experiments; stratifying is for sampling