Note Card Set 1 Flashcards
Variable
A characteristic we measure for each individual or case.
Quantitative (Numerical) Variable
Measured on a numerical scale.
Categorical Variable
Measured on levels (Low/Medium/High).
Discrete Variable
Only specific values, usually integers, are possible.
Continuous Variable
Numerical values can be continuous, where all values in the range are possible.
Nominal Variable
A categorical variable but the levels don’t have a special type of ordering.
Ordinal Variable
A categorical variable but the levels have a
natural ordering.
Population
Entire group researchers are interested in.
Sample
Represents a subset of the cases and is often a small fraction of the population.
Parameter
The overall quantity of interest may be the
mean, median, proportion, or some other
summary of a population.
Statistic
Estimated value of a parameter by taking a
sample and computing a numerical summary.
Observational Study
Collecting data without interfering with how the data arises.
Experiment
Research imposing a treatment to look for
causal relationships between the treatment and the response.
Explanatory Variable
For causal relationships (most of the time) first variable (What happens).
Response Variable
For causal relationships (most of the time)
second variable (affect or product or what
happened).
Treatment Group
A group of subjects that receives a specific
treatment or intervention.
Control Group
A group of subjects in an experiment that
does not receive treatment or intervention
and kept separate.
Levels
Unit of measurement (High, Medium, Low).
Placebo
A seemingly active treatment that contains no active ingredients.
Completely Randomized Experiment
A statistical experiment where all subjects or experimental units are assigned to different treatment groups entirely at random.
Block Design Experiment
An experimental design where experimental units are grouped into smaller subsets called “Blocks” to control for potential confounding variables that could influence the result, allowing researchers to isolate the effect of the treatment being tested more accurately.
Statistical Significance
A result from a stat analysis is unlikely to have occurred by chance, indicating that the observed effect is likely due to real phenomenon rather than random variation.
State three principles which help ensure
significance of an experimental study.
Randomization, Replication, and Local
Control.
How does direct control of confounding
variables increase the quality of an
experiment?
Ensures the accuracy and validity of study
findings.
How does randomization affect the quality of an experiment?
Minimizes selection bias.
Why is blinding potentially important for
experimental design (particularly when
involving humans)?
It prevents bias by stopping participants,
researchers,or other study personnel from
knowing which treatment group a person is assigned to, making sure the experiment is not influenced by conscious or subconscious expectations about the treatment received, giving more reliable and accurate data.
Differentiate between the goals of
experimental and observational studies.
Experimental studies aim to establish a
causal relationship by actively manipulating variables and introducing interventions while observational studies observe existing relationships between variables without intervention.
How are blocking and stratifying analogous? How are they different?
Both involve dividing a population up into
smaller more homogenous groups based on certain characteristics. They differ because blocking focuses more on ensuring an equal distribution of treatment assignments in a group, while satisfying aims to ensure that important characteristics are proportionally represented across different groups.
Differentiate between stratified random
sampling and cluster random sampling.
In stratified random sampling, a population is divided into homogenous subgroups called strata and then a random sample is drawn from that for representation of all groups, while cluster random sampling, the population is divided into groups and a random selection of clusters are chosen.
Why can’t observational studies establish
causal relationships?
This is due to the lack of ability to manipulate variables.
What does it mean for a statistic to be an
unbiased estimator?
When the expected value(average of all
possible sample values) is equal to the true value of the population parameter it is trying to estimate.
