What can a correlation analyze and what are its limitations?

What can a correlation analyze and what are its limitations?

• A. It analyzes causation; not sensitive to outliers.
• B. It analyzes strength and direction of association but cannot determine causation; sensitive to outliers; only measures linear relationships; third variables may influence results.
• C. It proves a causal link.
• D. It analyzes only mean differences.

Answer: (B)

Correlation examines how two variables move together, capturing both the direction (positive or negative) and the strength of their association. It does not establish causation—finding a strong relationship does not prove that one variable causes changes in the other. This is a key limitation: there can be a third variable driving both, or the relationship could be due to reverse causation or chance. Outliers can have a big impact on the correlation value, potentially exaggerating or masking the true relationship. Also, correlation primarily measures linear relationships; if the association is curved or otherwise non-linear, the correlation may be weak even though there is a meaningful relationship. Finally, the presence of confounding factors can distort the observed association, so correlation alone isn’t enough to infer a causal link. In short, correlation tells you how strongly and in what direction two variables relate, but it doesn’t prove causation, is sensitive to outliers, mostly detects linear patterns, and can be influenced by third variables.

Correlation examines how two variables move together, capturing both the direction (positive or negative) and the strength of their association. It does not establish causation—finding a strong relationship does not prove that one variable causes changes in the other. This is a key limitation: there can be a third variable driving both, or the relationship could be due to reverse causation or chance.

Outliers can have a big impact on the correlation value, potentially exaggerating or masking the true relationship. Also, correlation primarily measures linear relationships; if the association is curved or otherwise non-linear, the correlation may be weak even though there is a meaningful relationship. Finally, the presence of confounding factors can distort the observed association, so correlation alone isn’t enough to infer a causal link.

In short, correlation tells you how strongly and in what direction two variables relate, but it doesn’t prove causation, is sensitive to outliers, mostly detects linear patterns, and can be influenced by third variables.