Please enable JavaScript.
Coggle requires JavaScript to display documents.
Selecting a test - Year 1 - Coggle Diagram
Selecting a test - Year 1
DIFFERENCE BETWEEN CONDITIONS
ONE IV
TWO OR MORE CONDITIONS
BETWEEN GROUPS
INDEPENDENT T-TEST (PARAMETRIC = NORMAL DISTRIBUTION)
MANN-WHITNEY U TEST (NON-PARAMETRIC TEST)
WITHIN SUBJECTS
DEPENDENT T-TEST - PARAMETRIC
WILCOXON TEST - NONPARAMETRIC
THREE OR MORE CONDITIONS
WITHIN SUBJECTS
ONE-WAY REPEATED MEASURES ANOVA - PARAMETRIC
FRIEDMAN'S TEST - NON-PARAMETRIC TEST
BETWEEN GROUPS
ONE WAY INDEPENDENT ANOVA - PARAMETRIC
KRUSKAL-WALLIS TEST - NONPARAMETRIC
TWO OR MORE IV LEVELS (DIFFERENCE BETWEEN CONDITIONS) - ALL PARAMETRIC
BETWEEN GROUPS
TWO WAY REPEATED MEASURES ANOVA (THREE WAY ETC)
WITHIN-SUBJECTS
TWO WAY INDEPENDENT ANOVA
MIXED DESIGN
MIXED ANOVA
ADD A COVARIATE
ANCOVA
RELATIONSHIP BETWEEN VARIABLES (CORRELATION)
NORMALLY DISTRIBUTED DATA (PARAMETRIC)
PEARSON'S R CORRELATION
NON-NORMALLY DISTRIBUTED DATA (NON-PARAMETRIC)
SPEARMAN'S RHO CORRELATION
MECHANISMS BETWEEN VARIABLES
Interaction
Moderation
Indirect influences
Mediation
ASSOCIATION
Two nominal / one nominal and one ordinal
independent groups = Chi-square
Conditions are related - McNemar Test
Two ordinal or one ordinal and one continuous
Spearman's Correlation
Two continuous variables -> Pearson's Correlation, if non-normal distribution Spearman's
PREDICTION
DV is nominal
Two binary DV - binomial logistic regression
Non-binary - multinomial logistic regression
DV is ordinal
Ordinal logistic regression
DV is continuous
One IV - simple linear regression
More than one IV - Multiple Linear Regression
Explains how IV and DV are related through third variable = Mediation
Changes strength or direction of relationship between IV and DV = Moderation
NOMINAL DATA
INDEPENDENT / BETWEEN-GROUPS
CHI-SQUARE - EITHER ASSOCIATION OR GOODNESS OF FIT
Can use Fisher's exact for 2 x 2 table
REPEATED MEASURES / WITHIN SUBJECTS
SIGN TEST
ASSUMPTIONS TESTS
BETWEEN SUBJECTS - LEVENE'S TEST OF HOMOGENEITY
WITHIN-SUBJECTS - SPHERICITY
MULTIPLE REGRESSION - LINEARITY, HETERODASTICITY, MULTICOLLINEARITY AND NORMAL RESIDUAL DISTRIBUTIONS
TWO OR MORE DVs
MANOVA