Data Vocabulary
The entire group of individuals (people, animals, or things) that a researcher wants to obtain information about.
What is population?
A sampling method where every individual—and every possible group of individuals—has an equal chance of being chosen.
What is a Simple Random Sample (SRS)?
The key action distinguishing an experiment from an observational study.
What is deliberately imposing a treatment?
The type of statistical bias that occurs when certain groups in the population are systematically left out of the sampling process.
What is undercoverage bias?
An inactive or fake treatment, such as a sugar pill, given to a control group in medical experiments.
What is a placebo?
The memory trick taught in class pairs samples with statistics, while populations are paired with these usually unknown numerical summaries.
What are parameters?
Selecting a random starting point and then choosing every k^{\text{th}} individual, such as pulling every 50th bag off a conveyor belt.
What is systematic sampling?
The major advantage of a well-designed experiment over an observational study regarding variable relationships.
What is proving cause and effect (causation)?
The pillar of experimental design that uses chance mechanisms (like pulling numbers from a hat or using an RNG) to assign subjects to groups.
What is randomization (or random assignment)?
Question wording such as "Do you agree with the mayor's selfish decision..." introduces this specific form of bias.
What is response bias?
In statistics, a bare number like 42 is meaningless without this crucial element that answers the "Who, What, Why, When, Where, and How".
What is context?
Dividing a population into homogeneous subgroups based on a shared characteristic (like grade level) and taking an SRS from every group.
What is stratified random sampling?
Hidden variables in observational studies that prevent researchers from proving causation because they influence the response.
What are confounding variables?
The type of bias created when a chosen individual refuses to participate or cannot be contacted.
What is nonresponse bias?
The primary design principle violated when testing one group of runners on an indoor track and another group outdoors on a hilly trail.
What is control?
In a study where a principal surveys 150 students during lunch out of 2,000 total students, the group of 150 represents this.
What is the sample?
Dividing a population into heterogeneous "mini-populations," randomly selecting a few whole groups, and surveying everyone inside those chosen groups.
What is cluster sampling?
A study comparing toddlers who were breastfed versus formula-fed is classified as observational rather than an experiment for this primary reason.
What is the mothers chose the feeding method (the researcher did not impose/assign the treatment)?
The principle of experimental design that requires using enough experimental units in each group to distinguish real effects from chance variation.
What is replication?
Setting up a table outside a cafeteria or approaching shoppers near a mall bench are examples of this convenient, yet highly biased, sampling method.
What is convenience sampling?
If a quality control manager calculates an average lifespan of 9,850 hours from a tested group of 100 LED bulbs, this specific value of 9,850 hours is classified as this type of numerical summary.
What is a statistic?
Large-scale professional studies use this process, which combines multiple sampling methods across successive stages.
What is multistage sampling?
The main reason a researcher would choose an observational study over an experiment when investigating harmful behaviors like smoking.
What is ethical restrictions (or observing natural settings)?
The phenomenon where subjects alter their behavior simply because they know they are being actively watched by an observer.
What is the Hawthorne effect?
In an experiment measuring antacid dissolution times using Cold, Room Temperature, and Hot water, the three water conditions represent these.
What are the treatments?