Parametric vs Non-Parametric data

what is parametric data

normal data

what is non-parametric data

not normal or non normal data

what is considered normal data 

data that has a bell shaped curve, not to skewed, not to kurtotic, no outliers

why does non normal data matter

normality is the assumption of some statistical models mathematically, if we violate normality and we use parametric test we may not be able to trust the model estimates. 

how do we test for outliers?

test for outliers using box plots 

what to do if data has outliers

run a non-parametric test

remove the value and leave as missing

remove value and replace with nearest acceptable range

remove value and replace with mean

what are ways to test for skew and kurtosis

shapiro wilk test, w statistic, p value (of W statistics)

what are the parametric tests

pearson correlation, t-test (between or within groups), ANOVA (within groups)

what are non-parametric tests

spearman correlation, Wilcoxon test (2 groups/conditions), Kruskall-Wallis test (3 or more groups)

what is degress of freedom

based on the quantities of data in chosen model, degrees of freedom will mostly be the number of participants minus 1. more df means more statistical power