Lecture #8 (Correlation) Flashcards
What are we arguing in correlation?
We are arguing that there is a trend between two variables (Shoe size and height)
How do we measure correlation?
Between -1 and +1 where the closer to 1 you get, the stronger the correlation
What is Pearson’s R used for?
Pearson’s R is used for measuring the variability between two datasets
What is the formula for Pearson’s R?
r = (Σ(xᵢ - x̅)(yᵢ - ȳ)) / √(Σ(xᵢ - x̅)² Σ(yᵢ - ȳ)²)
What are some assumptions we have while using Pearson’s r?
Data is normally distributed and is ratio data
In Pearson’s r, is the data independent from eachother?
Yes
What do we use Spearman’s Rank Correlation Coefficient for?
For ordinal and interval data, uses rank instead of values
What is Kendall’s Tau Rank Correlation Coefficient rτ?
Like Spearman’s but has zero assumptions about data distribution
Does Kendalls’s looks at conformity (concordance) and disagreement (discordance) in the rankings?
Yes
What is the Phi Correlation Coefficient (rф)?
This measures the degree of association between data that has been categorized into binary data
What is Phi close to?
Chi Squared