Causality

Describes a cause-effect relationship between two events or phenomena, i.e. that one is caused by the other.

A short circuit in an electrical appliance caused the house fire.

It is important to distinguish the concept of causality from that of correlation, which refers to the co-occurrence of events or phenomena but does not imply a causal relationship.

Description

As a general rule, when phenomena π¨ and π occur in correlation (i.e. either simultaneously or consecutively), the following possibilities must be considered:

1. π¨ causes π (direct causal relationship).
Β
2. π causes π¨ (reverse causal relationship).
Β
3. A third event π causes both π¨ and π (external causal relationship).
Β
4. π¨ causes other events: e.g. π, π, etc, which in turn cause π (indirect causal relationship).
Β
5. π¨ and π are not causally connected, but only occur together by chance (spurious correlation).
Β
6. π¨ and/or π are observational errors and at least one of the phenomena does not occur at all in the form described (e.g. frequency illusion, base-rate fallacy, etc.)

The fact that π¨ and π correlate with each other, i.e. that they regularly occur together, is not sufficient to postulate a causal relationship.

However, the question of when a causal relationship actually exists is rather difficult to answer. Ultimately, different scientific fields have come up with various different β in some cases even contradictory β definitions of causality. These may be more (e.g. in physics) or less strict (e.g. in the social sciences), but they usually contain at least the following minimum requirements:

1. A high correlation of the properties.
Β
2. A plausible mechanism of effect.
Β
3. A comprehensible chronological sequence of events.

It should however be noted that for each of these criteria there are also problem cases and exceptions.