Inferential Stats

When to use a sign test

testing for a difference

repeated measures design

nominal data

Steps to calulate a sign test

work out the difference between two columns

decide whether each difference is positive or negative - results with 0 are cancelled

count the signs

caluculate the S value (smaller of the two signs)

work out if its significant - calculated S value must be equal to or less than the CV to be significant

When to use chi squared

nominal data

independent design

finding difference

testing correlation

When to use spearmans rho

testing correlation

ordinal data

What is the rule of ‘r’

if there is an R in the name, the calculated value needs to be GREATER than the critical value for it to be significant

When to use pearsons R

testing for correlation

interval data

data is normally distributed

When to use an unrelated T test

testing for a difference

independent groups design

interval data

When to use a related T test

testing for a difference

repeated measures design

interval data

When to use Mann-whitney U test

testing difference

independent groups desing

ordinal data

When to use a wilcoxon test

testing difference

repeated measures

ordinal data

What is the T value in a wilcoxon test

the sum of the less frequent sings

What is a null hypothesis

there will be no significant difference in the investigation

What is the significant level

the level of probability at which it is agreed that the null hypothesis will be rejected and the experimental hypothesis accpeted

What is a type one error

False positive - when experiment hypothesis is accepted when the results were chance findings so null should have been accepted

What is a type two error

false negative - when the null is accepted when actually it was a real difference/relationship

Parametric data

certain features (parameters) that the study has to meet to use one

if it doesn’t meet these then a non-parametric test is used

Which tests are parametric

t-test

pearsons r

What tests are non-parametric

sign test

chi squared

mann whitney

spearmans rho

wilcoxon

What is the difference with nominal, ordinal and interval data

nominal - frequency data (tally)

ordinal - ranked data (scales)

interval - anything with a 0 (test score, heart rate)

How to calculate df for an unrelated t-test

N (condition a) + N (condition b) x 2

How to calculate df for a related t-test

N-1

How to calculate df for chi squared

(rows-1) x (columns-1)

How to write a statement of significance

the findings were significant/non-significant at P less than or equal to 0.05. This is because the observed value of ….. is higher/lower than the critical value of …… for a one/two tailed test with n= …..

One-tailed test

checks for a difference in only one direction - directional

Two-tailed test

checks for a difference in either direction - non-directional