STAT 200: Variables, Cases, and Data Types Flashcards

Raw Data: Data formatted in tables where summary representations (such as histograms, pie charts, or smooth normal distributions) synthesize individual observations.
Cases: The individual units of observation represented in the rows of a raw dataset, such as individual Stat 200 survey respondents or countries of the world.
Variables: Characteristics or features recorded about each case, represented in the columns of a dataset, which vary from one case to another.
Identifier/Index Variables: Specialized categorical variables, such as a PSU ID, where each category corresponds to exactly one unique case.

Definition: Variables that divide cases into non-overlapping groups or categories using words, symbols, or numerical labels.
Nominal Categorical Variables: Categories that lack a natural or inherent ordering (e.g., living on campus, Pennsylvania residency, phone area codes like 814, or Nobel Prize categories).
Ordinal Categorical Variables: Categories that possess a natural, logical order (e.g., year in school, letter grades, or bracketed numerical groups such as 0 or 4 to 7 drinks per week).

Definition: Numerical measures or counts for each case where arithmetic operations (e.g., calculating an average or sum) are mathematically meaningful.
Discrete Quantitative Variables: Countable quantities that take on a distinct, limited set of values (e.g., number of children in a household, puppy litter size, or US states visited).
Continuous Quantitative Variables: Measured quantities that theoretically take on any value within a given range at arbitrary precision, typically requiring a measurement tool (e.g., blood alcohol content measured in $\text{g/dL}$ , height, or travel duration).

Explanatory Variable: The variable used to explain, predict, or account for changes in the outcome variable.
Response Variable: The primary outcome variable of interest being measured, explained, or predicted.
Analytical Relationships: Statistical questions evaluate how an explanatory variable impacts a response variable (e.g., in a study of 2,623 schoolchildren in Barcelona, the amount of greenery surrounding schools serves as the explanatory variable predicting student attention span as the response variable).

Audience Query: Are calories considered discrete or continuous, given that nutrition packaging reports whole numbers while calories represent energy units?
Response: Calories represent continuous energy units that can theoretically be measured to arbitrary precision; however, practical data collection methods sometimes record continuous variables in discrete integer buckets.