Stats Unit 1 Vocab

Descriptive stats

describe/display sample data

Inferential stats

use sample data to make inference about larger population

Population

• Size N, set of all subjects we want to describe

• defined by researcher

• large, varied, and hard to describe

parameters

• characteristics of a population

• μ- mean, σ- stdev, ρ- correlation

sample

• size n, subset of a population

statistics

• characteristics of a sample (estimates of parameters of a pop)

• X̄- mean, s-stdev, r-correlation

sampling error

happens by chance, reduce by increase sample (size n)

bias

systematic discrepancy, eliminate by taking a random sample

categorical variable

• define membership in a group

• (univariate) describe using frequencies or proportions/%

categorical nominal

categorical variable that has no inherent order (last movie seen, etc)

categorical ordinal

categorical variable that exists in an order (ie fruit ripeness, class)

numeric variable

• quantitative measurement that typically has magnitude and units

• (univariate) describe using a measure of center (mean, median, mode) and a measure of spread (stdev, IQR)

numeric discrete

variable that only exists in individual increments, doesn’t typically have decimals (# of texts sent, shoe size (only have .5 increments))

numeric continuous

exists as any real number (decimals typically make sense); hours slept, height, etc

confounding variable

unmeasured variables that may mask/distract from the causal relationships between variables of interest (homeless people cause car crashes)

univariate

describe 1 variable’s distribution

bivariate

describe relationship between 2 variables

explanatory variable

(bivariate) independent/predictor variable; see if it explains or predicts another variable

response variable

(bivariate) dependent/outcome variable, what you want to explain/predict (variable of interest)

standard deviation

measure of spread, gives average distance for each value to X̄ (mean)

Inner quartile range (IQR)

measure of spread, distance between 1st (25th percentile) and 3rd quartile (75th percentile)

Q3-Q1

percentile

percentage of values in a set of data scores that fall below a given value (25th percentile- 25% of the values are below the given value)

5 number summary

set of numbers that explain distribution

{min, Q1, median, Q3, max}

Describing numeric variables

• include shape (symmetric or skewed) and a measure of center and spread (depends on shape)

symmetric

describe using x̄ (mean) and s (stdev)

skewed

describe using median and IQR

pos/neg based on which direction the tail points

contingency/2-way frequency table

table to display 2 categorical variables

compare the marginal distribution to the conditional distribution

marginal distribution

percentage out of the totals, probability of one variable

conditional distribution

probability of one variable given another

grouped bar chart

used to display 2 categorical variables

mosaic plot

used to display 2 categorical variables, better for when group sizes differ

grouped box plot

used to display one categorical, one numeric variable

grouped histogram

used to display one categorical, one numeric

pearson correlation (r)

tells strength and direction of linear relationships

law of large numbers

probability of an even is what its relative frequency will converge on after infinite trials

or probability

pr(a or b) = pr(a) + pr(b) - pr(a and b)

• removes overlap

mutually exclusive events

pr(a and b) =0

and probability

pr(a and b) = pr(a) * pr(b|a)

• to remove overlap

conditional probability

pr(b|a)

• probability of b, given a has already happened

• reduces sample space