Association Flashcards
Association between variables
correlation, regression, chi squared test
association aims to see if
x and y are related
correlation
examines direction/strength of association, both variables are continuous (pearson moment), or one is continuous one is variable is ordinal (spearmans), no outliers, linearity, related pairs
regression
examines how well the line describing association can be used to explain y variation
numeric variables
correlation, regression
chi squared
contingency tables, examines association of 2 independent variables
categoric values
chi squared
correlation analysis
plot data using scatterplot to see outliers/linearity, then work out coefficient (standardise data, observation-mean/standard deviation) then determine if can happen by chance (set null the p value)
negative correlation
as x increases y decreases
positive correlation
as x increases y increases
what does 1/-1, 0 mean
stronger correlation, 0 is weak/no
test equation
r\/(n-2)/ 1-r(2)
however
significant correlation doesn’t mean strong correlation, large samples can have significant correlation when its weak (vice versa), linear relation doesn’t mean there’s a relationship
regression equation
y=b(o) +/- b(1)X
what do they mean
b(o) is intercept, +/- is correlation, b is slope