causal inference

Causation describes a relationship where one event (the cause) brings about another event (the effect). This connection involves the cause preceding the effect, and a necessary link where the effect must follow the cause. While we observe constant conjunctions between events, the idea of a "necessary connection" is subjective, stemming from our repeated experiences rather than an observable property of the events themselves.1

In research, establishing causation can be challenging. Controlled experiments, like randomized control trials (RCTs), are considered the gold standard for inferring causation because they allow for manipulation of variables and random assignment of subjects to groups, minimizing confounding factors. However, RCTs are not always feasible or ethical. In such cases, researchers may use observational studies or natural experiments, which exploit naturally occurring events or policy changes to study phenomena. While natural experiments can provide valuable insights, they are quasi-experiments and must be analyzed with caution, as they are subject to threats to causal inference and may not unequivocally determine causation. Measures of association, such as correlation coefficients or odds ratios, quantify relationships between variables but do not inherently prove causation. It is crucial to avoid inferring causality solely from an association, as other explanations may exist.23