Step by step guide Correlational
Analysis
Goal: How unusual is our observed correlation coefficient in the sample, assuming H 0 is true?
Both the independent and dependent variable are continuous.
Hypothesis:
Null-hypothesis H0 no (linear) relationship with one another.
Alternative hypothesis H1 a (linear) relationship with one another.
! Remember: correlation ≠ causation.
Assumptions of the correlational analysis:
1. Linearity: the relationship between the variables is linear.
a. Violated: transformation, non-linear model, non-parametric correlation (Spearman’s rho)
2. Independence of data
3. Normality (CLT applies)
a. Violated: Spearman’s rho
4. Homogeneity of variance
1
Analysis
Goal: How unusual is our observed correlation coefficient in the sample, assuming H 0 is true?
Both the independent and dependent variable are continuous.
Hypothesis:
Null-hypothesis H0 no (linear) relationship with one another.
Alternative hypothesis H1 a (linear) relationship with one another.
! Remember: correlation ≠ causation.
Assumptions of the correlational analysis:
1. Linearity: the relationship between the variables is linear.
a. Violated: transformation, non-linear model, non-parametric correlation (Spearman’s rho)
2. Independence of data
3. Normality (CLT applies)
a. Violated: Spearman’s rho
4. Homogeneity of variance
1
