Greg Martin Statistics Overview

Page 1

https://www.youtube.com/watch?v=I10q6fjPxJ0&t=437s

Variables can be Categorical

summarize in table or bar chart

Numeric

distribution- range, interquartile range, mean median

box plot

histogram

H1 is hypothesis

H0 is null hypothesis

Probability of relationship testing- p value.

Alpha valiue is cut off pt to reject the null. if p is less than alpha, we reject the null

Chi-squared test (x^2) determines if there is significant difference between expected and observed frequencies in one or more categories

looks at how likely the observed frequencies would be if null hypothesis is true

if more than two categories in one of the variables, do ANOVA instead of t test

correlation test- get correlation coefficient (between-1 and 1) and p value

perfectly negative correlation is -1

no cor is 0

perfectly pos is 1