Why do people get cancer? Why do some young people complete an education while others do not? Why do insects disappear? 

A lot of research is about finding causes. In the search for answers, scientists often find statistical coincidences: developments or phenomena that correlate over time.

A classic blunder among journalists is believing that a statistical coincidence implies cause and effect. But often the coincidence is … well, coincidental.

An Australian statistic, for example, shows that the number of drowning accidents increased with the sale of ice cream.

This does not mean that eating ice cream increases the risk of drowning. 

The underlying cause of the statistical coincidence is probably that people both go swimming and eat ice cream more often in the summer. 

Can family dinners be linked to drugs?

Another example of underlying causes possibly playing an important role: media in several countries have reported on numerous studies linking family dinners with a decreased risk of drug addiction for teens. 

The message seems simple and appealing in the media reports, but the correlation between family dinner and drug abuse does not nescessarily mean that you can prevent drug abuse simply by eating regular dinners with your child. 

There may be several other reasons for the correlations found in these studies. 

For example families having daily dinners together may have strong family connections or deep parental involvement in a child’s life. 

In other words, the dinner itself may not be as ‘magic’ as reported in the popular press.
A lot of research is needed before we can be sure that one is in fact causing the other: that there is a causal relationship.

When scientists conduct statistical studies of, for example, diseases in the population, you will find that they often try to adjust their results for the participants’ financial background, level of education, health and other conditions. 

Ask about the causal connection

This means that the scientists have tried to rule out that the connection they find is due to something other than what they have studied. 

But even if the scientists adjust for these kinds of underlying causes, there may be explanations that they have not been able to investigate, or their data set may be too small to conclude anything unequivocal.

Therefore, always ask the scientists or your independent source how likely it is that their study has found a causal connection.

If you as a journalist confuse a statistical coincidence (correlation) with a causal relationship (causality), you may spread misinformation.

In a worst-case scenario, such misunderstandings can lead readers, listeners or viewers to make choices that can damage their health and well-being.

Read more tips by clicking the blinking icons at the left in the graphic below.