If you want to conclude that a given variable has a causal relation with a second variable...
The degree to which observed values/scores are free from errors of measurement, refers to