Correlational data

When we talk about interpreting correlational data, we often rightly stress that correlation does not establish cause and effect. However, it is important not to simply dismiss correlations because of this – as this video from the OU’s Joy of Stats series shows, the scientific attitude to an apparent correlation is to work as hard as you can to refute that correlation. If it survives all those attempts, then you might have an association between co-variables that is really worth exploring.