Errors, Non-/Directional hypothesis, & Non-/Parametric tests

type I error

no effect, H0 true, reject H0

Type II error

effects exist, H0 false, failed to reject H0

false positive

type I error AKA

false negative

type II error AKA

directional prediction

prediction that leans toward a particular group

directional prediction

uses one-tailed hypothesis

directional prediction

prediction that focuses on only one of either positive or negative value on bell-shape curve

non-directional prediction

uses two-tailed hypothesis

directional prediction

prediction that focuses on both positive and negative

directional prediction

example: "group A has higher variable than group B"

non-directional prediction

example: "there is no sig difference in variable between grp A and grp B"

one-tailed hypothesis test

specify either increase or decrease in the population mean

directional prediction

makes a statement about the direction of the effect

non-directional prediction

prediction that doesn't lean towards any particular group

two-tailed test

tailed test that demands more evidence to reject H0 because of the p-value being divided into two

two-tailed test

p < 0.05 turned into p < 0.025

one-tailed test

this type of tailed test is more sensitive

two-tailed test

tailed test that is used when there is no strong directional expectation or when there are 2 competing predictions

one-tailed test

tailed test that should be used when directional prediction is made

one-tailed test

tailed test that should be used if there is a strong justification for making the directional prediction

parametric tests

designed to test hypothesis about specific population parameters

parametric tests

makes assumptions about other population parameters

parametric tests

requires a numerical score for each individual in the sample

parametric tests

test that requires scale interval/ratio

non-parametric tests

test that requires scale ordinal or nominal

non-parametric tests

participants are usually classified into categories

one-tailed test

tailed test to use when researchers want to prove something