Here are some headlines that were written about a study on job satisfaction:
You'll recognize that the verbs in these claims ("lead to" and "boost") are causal ones, and you should know by now that it takes an experiment to support causal claims. Before you read on, ask yourself:
a) What are the two variables in the claim, a "healthy sex life boosts job satisfaction?"
b) Would it be possible to conduct an experiment on this claim?
Now read this description of how the study was conducted (source):
To understand the impact of sex on work, the researchers followed 159 married employees over the course of two weeks, asking them to complete two brief surveys each day. They found that employees who engaged in sex reported more positive moods the next day, and the elevated mood levels in the morning led to more sustained work engagement and job satisfaction throughout the workday.
c) Did the study manipulate a variable, as would be required for an experimental design? (Oh, and did you notice the journalist's verb, "led to," when describing these results?)
Even though this wasn't an experiment, you might notice that it took place longitudinally. By testing how sex one one day was associated with mood and work engagement the following day, the study appears to have temporal precedence.
d) Re-read the study description above. The results are describing mediation. What are the three variables in the mediation diagram? Sketch a figure with three boxes (similar to the mediation diagrams in Chapter 9) that represents this mediation.
Mediation or no, this is still a correlational study. The study supports an association claim (of the form, "Sex the night before was associated with better mood the following morning, and morning mood was associated with evening reports of job satisfaction"). But it cannot support a causal claim.
You can read the peer-reviewed article here.