Unit 1 Flashcards
Experimental Design, Frequency Distributions and HIstograms, and Descriptive Statistics
Data
A collection of observations
“Datum” - Plural
Statistics
A field of study using math to collect, describe, and analyze data
Population
The complete or whole group of the observations of interest
Sample
A subset of observations from the population of interest
Descriptive Statistics
Using measures to summarize numerical data
Inferential Statistics
Creating estimations on populations based on limited information
Parameter
A single value summarizing the observations in a population data set
Statistic
A single measure summarizing sample data
Variable
An observation or measurement record that describes a characteristic of individuals in a sample
Quantitative
Recorded as a number
Categorical
Recorded as a text/description
*Numbers that function as labels = Categorical
Experiment
A study that usually compares two or more values in highly controlled procedure in order to determine cause/effect relationships
Observational Study
A study which measures or describes one or more groups w/ limited interaction
Experiment Unit
Individuals in an experiment being observed/measured
People = “subjects”
Treatment
The casual factor being studied
Control Group
The group not receiving the treatment
Placebo
A fake treatment
The Placebo Effect
A phenomenon where subjects respond to a fake treatment
Confounding Variable
A characteristic of a group that wasn’t measured in the study and wrongly attributed change to the treatment
3 qualities for effective experiments
- Control: Avoid confounding variables
- Randomization
- Replication
Completely randomized design
randomly assigning all experimental units with no organizing, sorting, or categorizing beforehand.
*Works best with no confounding variables within the experiment units
Randomized block design
Experimental units randomly assigned to stratas
Matched pairs design
When they’re meaningful differences from a number of intrinsic characteristics, initially participants are paired to “closet match” according to important characteristics.
Survey
A poll, questionnaire, or interview in which the responses are to answer questions
Simulation
A study where observations are generated from realistic representations or recreations of events, rather than real life
* easy to gather quick data
Random Sampling
A sample size n is constructed such that every possible sample of the same size n has the same probability of being chosen
Systematic random sample
With a list of the members of the population, a cutoff by a certain number k such that every kth member of the population is selected.
*relatively prime k = representative percentage
Cluster sample
populations are separated into groups, then a certain number of the groups are randomly chosen
Stratified random sample
grouping the population into stratas, then from every group a number of individuals are chosen random