What criteria can be used to distinguish between correlation and cause & effect
First, it is important to know what correlation and cause & effect means to be able to distinguish between them
Correlation is a mutual relationship or connection between two or more things. Generally it is the degree to which one phenomenon or a random variable is associated with or can be predicted from another.
In statistics, correlation usually refers to the degree to which a linear relationship exists between random variables. Correlation may be positive or negative or inverse:
both variables increase or decrease together.
one variable increases when the other decreases.
Cause & effect definition:
Very often, the correlation and cause & effect get mixed up. This is either due to a misunderstanding or to provide a plausible explanation for a scientific observation. Therefore, it is very important to be able to understand the difference between the two concepts.
Causation involves correlation, this means that if there is a cause and effect then they are correlated. However, correlated events can be caused by a an underlying cause, which means they do not necessarily cause each other, another factor not explicitly mentioned does (a common cause). Just because two events occur together does not imply that one is the cause of the other or that without one event occurring, the other would not happen.
The more solid the correlations, the more likely they are to imply causation.
Eg. the link between smoking and cancer. The correlation between the incidence of cancer and smoking is strong enough that most today consider this to be a cause and effect relationship.
Smoking causes cancer, but cancer does NOT lead to smoking.
Relation of sickle cell anemia and malaria
There is a correlation between high frequencies of the sickle-cell allele in human populations and high rates of infection with Falciparum malaria. Where a correlation exists, it may or may not be due to a casual link.
A causal link is where the Independent Variable has a direct impact on the Dependant Variable.
Sickle cell anemia is when there is a mutation in the genes when the sixth codon is mutated from GAG to GTG. When this allele is transcribed, the mRNA now has GUG as its sixth codon. When this amino acid is coded for, it creates valine instead of glutamic acid. This causes hemoglobin molecules to stick together in tissues with low oxygen concentrations. The hemoglobin molecules distort the red blood cells into a sickle shape.
The frequency of the sickle-cell allele is correlated with the prevalence of malaria in many parts of the world. Therefore, there IS a causal link. There has clearly been natural selection in favour of the sickle-cell allele in malaria ridden areas, despite it causing severe anemia to red blood cells. Natural selection has led to particular frequencies of the sickle-cell and the normal hemoglobin alleles, to balance the risks of anemia and malaria and this is cause & effect.
In conclusion the criteria used to distinguish between correlation and cause & effect is to differentiate whether both effects that are correlated were caused for a particular and logical reason, without any underlying causes (directly related). And if not, then these effects only show correlation and no causality.
All examples of Correlation vs Cause and Effect are shown in the videos below to visually and aurally help you understand the two.
TEDx Talks – The danger of mixing up causality and correlation:
Khan Academy – Correlation and Causality
Correlation Does NOT Imply Causation