1.) What kind of data do I have?
Nominal level (non-parametric): category data e.g. yes/no
Ordinal level(non-parametric): data that is a score or ranked e.g. 1st 2nd
Interval data (parametric)– standard unit of measurement e.g. seconds, KG
Criteria of parametric data
- interval data – it’s a standard unit of measurement
- population has a normal distribution
- homogeneity of variance – set of scores in each condition have similar dispersion
2.) A correlation or experimental difference?
Difference between behaviours in 2+ conditions e.g. experiments or research into relationships between to 2 covariables e.g. correlation
3.) What is the experimental design
If it is an experiment, the experimental design affects the statistical test chosen
Independent groups or non-independent groups (repeated measures/matched pairs)
Types of Statistical tests
Nature of Hypothesis | Level of measurement | Experimental design | ||
Independent | Not independent | |||
Test of Difference
(experiment) |
Nominal | Chi squared test | Sign test | |
Ordinal | Mann-Whitney | Wilcoxon Test | ||
Interval (parametric) | Unrelated T test | Related T test | ||
Types of Statistical tests
Relationship
(correlation) |
Ordinal | Spearmans Rho |
Interval | Pearsons R test |