Maths

Question 1

Chi-Squared

Answer 1

USED FOR NOMINAL INDEPENDENT GROUPS DESIGN
Test of difference

This is done for every number of data

EXPECTED FREQUENCY = (Row Total X Column Total) / Sum of row total + Column total

Degrees of Freedom = (Number of Rows - 1) (Number of Columns - 1)

Add all calculated values together to get Official Calculated Value

Question 2

Mann-Whitney U

Answer 2

USE FOR ORDINAL (in order) INDEPENDENT GROUPS DESIGN

Rank ALL of data collectively, if same number add together ranks and find mean.

Ra = Added ranks of condition A
Rb = added ranks of condition B

Na = Number of scores in condition A
Nb = Number of scores in condition B

Question 3

Spearman’s Rho

Answer 3

USE WHEN THERE IS A TEST OF CORRELATION OF 2 INDEPENDENT VARIABLES

Rank data of each group SEPERATELY.

Add a ‘d’ column & ‘d^2’ column

Find ‘d’ (difference) by subtracting Rank 1 from Rank 2

Find d^2

Find Sum of d^2

N = Number of ppts.

Question 4

Wilcoxon

Answer 4

USE WHEN TEST OF DIFFERENCE FOR MATCHED PAIRS OR REPEATED MEASURES

Find difference between score on 1 condition with its matched score on other condition

Rank all differences from smallest. Leave out 0s. Treat positives and negatives as the same.

Question 5

What test to use

Answer 5

Never Chuck Chocolate Sponge In My Red Wellies

N C
C S
I M
R W

Nominal = Chi- Square
Correlations = Spearmans Rho
Independent Groups = Mann-Whitney
Repeated Measures = Wilcoxon

Question 6

Significance

Answer 6

0.05 (5%) is what we use for everything other than medical, which is 0.01 (1%)

Question 7

Type 1 & Type 2 errors

Answer 7

False-positive (TYPE 1 ERROR), saying something is there that isn’t e.g. accepting the alternate hypothesis when the null is correct

False-negative (TYPE 2 ERROR), saying there is nothing there when there is e.g. Accepting null hypothesis when the alternate is correct