Question

An investigator examined cases of fetal death in 27,000 pregnancies and classified mothers according to whether they had experienced sexual intercourse within 1 month before delivery. It was found that 11% of the mothers of fetuses that died and 2.5% of the mothers of fetuses that survived had had sexual intercourse during the period. It was concluded that intercourse during the month preceding delivery caused the fetal deaths.

Explain why this statement is inaccurate

Answer 1

Correlation does not imply causation

Explanation:

Correlation is a statistical technique which shows how strongly a pair of variables are linearly related and change together.

Causation says that the change in one value will cause the other to change, making one variable dependent on the other.

Examples of correlation with causation are:

a) ice cream sales and summer temperatures

b) forest fires and summer temperatures

An example of correlation without causation is ice cream sales and forest fires. While they both increase as temperatures increase, ice cream sales are not CAUSING the forest fires.

From the information available it can be suggested that there is a correlation between sexual intercourse within a month prior to delivery as there is a higher number of fetus deaths to mothers who had sexual intercourse within a month prior to delivery. However, there is no information which shows the sexual intercourse caused the deaths. Additionally, 89% of the mothers of fetuses who died DID NOT have sexual intercourse within a month prior to delivery, thus even the correlation is weak in this case.