T1 ETA NORMALITY TEST TERMS

Independent t-test

• parametric

• 2 independent groups

Mann Whitney U test or Fisher’s Exact test

• nonparametric

• 2 independent groups

Dependent t-test or Paired t-test

• parametric

• 2 dependent groups

Wilcoxon’s test or McNemar’s test

• nonparametric

• 2 dependent groups

Analysis of Variance (ANOVA)

• parametric

• > 2 independent groups

One-way ANOVA

• parametric

• > 2 independent groups

Repeated Measures of ANOVA

• parametric

• > 2 independent groups

• REPEATEDLY

Friedman’s test or repeated measures of logistic regression

• nonparametric

• > 2 independent groups

• REPEATEDLY

Dependent

variable from the same group

Independent

variables from different group(s)

Parametric

assume data is normally distributed

Nonparametric

assume data is not normally distributed

• rely on ranks/counts

Ratio Data (height, weight, age)

• difference between measurements

• true zero exists

• quantitative

Interval Data (like temperature, IQ, dates)

• difference between measurements

• NO true zero exists

• quantitative

Ordinal Data

• ranking categories

• qualitative

Nominal Data

• categories

• no ordering or direction

• qualitative

One-way ANOVA

• parametric

• 1 outcome variable

• 1 independent variable

Kruskal-Wallis

• nonparametric

• 1 outcome variable

• 1 independent variable

Two-way ANOVA

• parametric

• 1 outcome variable

• 2 independent variable

Friedman’s test

• nonparametric

• 1 outcome variable

• 2 independent variable

Multivariate ANOVA

2+ outcome variables

Pearson Product Moment Correlation Coefficient (Pearson’s r)

• parametric

• 1st: continuous

• 2nd: continuous

Spearman Rank Correlation or Kendall’s Tau b

• nonparametric

• 1st: continuous/categorical

• 2nd: categorical (ordinal)/continuous

Chi-square Test of Association

• nonparametric

• 1st: categorical

• 2nd: categorical (nominal)

Dichotomous

outcome with 2 possible values