Module 3 - Descriptive Stats

numerical variables

numeric data

categorical variables

qualitative “yes or no”

continuous numerical

continuous numbers, any value, fraction decimals

discrete numerical variable

integers, only whole numbers

Ordinal categorical variable

qualitative value, on a ranked scale.

• Ex: describe ur happiness with emojis in an ordered scale

• Ex: letter grade on an exam

nominal categorical variable

qualitative, values don’t have order

• Ex: there's no inherent order to apple, oranges bananas, they're types of fruit

central tendency

describes the typical value in sample (ex: mean, median mode)

summary measure that attempts to describe a whole set of data with a single value that represents the middle or centre of its distribution.

dispersion

describes spread of values (variance, range, IQR, sd)

how to calculate variance from data set

1. find mean

2. accumulate sum of (each number - mean)²

3. divide all by how many numbers there are

quartiles

ranked bins of data, divide data in 4 equal parts of 25%

medians

how to calculate quartiles from odd data set

2nd: middle number

1st: omit middle number, find the average of the 2 middle on the left

how to calculate quartiles from even data set

2nd: find average of 2 middle numbers

1: omit middle number, find the middle number of the left.

use mean or quartile w small data set?

small data set - use means. Don’t use quartiles bc those are sensitive to small data sets.

data w outliers

mean changes a lot but median stays the same. Means are sensitive to outliers so use quartiles.

quartiles are resistant to extreme outliers. Helps you understand the central tendency despite the outliers

effect size

evaluate if change in response variables is meaningful. Looks at the magnitude of difference btw groups, strength of relationships btw 2 variables.

purpose of ratios and effect size

ratios help u see proportional change, see the relative change, relative effect size

mean for A/mean for B

age (cat or num)

contin num