parametric and non-parametric statistical tests - RM

5 different types of non-parametric tests

• mann whitney U

• wilcoxon

• chi-squared

• binomial

• Spearmans Rho

4 different things to consider when choosing non-parametric test

• type of data- ordinal/nominal/interval

• experimental design- IMD/RMD

• difference in conditions ?

• relationship/correlation?

test for IMD and nominal

• chi squared

test for RMD and nominal

binomial

test for correlation and nominal

no test

test for IMD and ordinal

mann whitney U

test for RMD and ordinal

Wilcoxon

test for correlation and ordinal

Spearmans Rho

what is nominal data + examples

• can be separated into discrete categories e.g. male/female

what is ordinal data

ranking/ scaling

how do you calculate degrees of freedom

(number of rows -1) x (number of columns -1)

when using chi -squared, if critical value is lower than observed value do we accept of reject null hypothesis

reject

for which tests do we reject the null hypothesis if the critical value is lower

• chi-squared and spearmans rho

rhyme for rejecting null hypothesis

when choosing chi- square or spearmans rho, the critical value is always low(er than observed)

when using all the other tests, the critical value is above the rest