So what have we learnt?
We found that reading a book in bed before going to sleep improved sleep.
That’s encouraging but, we wanted to check that this finding wasn’t simply a result of there being other differences between the groups.
Using a statistical model we were able to check this wasn’t the case.
This allowed us to compare the sleep quality between the two groups at the end of the trial. The model takes into account how we measured sleep quality in The People’s Trial (i.e. ‘Terrible’ being the worst sleep quality through to ‘excellent’ being the best sleep quality) as well as the influence of:
- initial sleep quality
- knowledge of clinical trials
- whether they worked in healthcare or not
As mentioned early, at the start the two groups were very similar but not exactly the same. The statistical model allows us to see if accounting for these small differences changes our result
In short, it didn’t. Our statistical model also found that reading before sleeping had a positive effect on sleep quality compared to those that did not read at bedtime.
This means that we can now make a reliable claim that in people similar to those who took part in this trial, reading a book in bed before sleeping, in the manner done in this trial, has a positive effect on quality of sleep compared to not reading a book in bed.
Some people sleep better than others for different reasons. Maybe your gender makes a difference to your quality of sleep? Or maybe younger people sleep better in general?
A statistical model is something that uses data to describe any relationship between our outcome (i.e., sleep) and something that may influence it (like age or gender).
In The People’s Trial, a statistical model helps us decide if ‘reading a book in bed’ or ‘not reading a book in bed’ makes a difference to sleep quality while taking into account those things, such as age or gender, that might explain why some people sleep better than others anyway.