The BBC website published a worrying article today about actual vs. self-reported alcohol consumption.
“University College London researchers compared alcohol sales figures with surveys of what people said they drank.
They found there was a significant shortfall with almost half of the alcohol sold unaccounted for in the consumption figures given by drinkers.”
It’s a great reminder of why you’ve got to choose your methodology carefully.
Asking how people much they drink has three major flaws.
First, people find it hard to remember what they drank, and many people don’t understand what a unit of alcohol is.
Second, people want to be viewed favourably by others, and don’t want to admit to socially undesirable behaviours. The social desirability bias.
Third, they may also have a vested interest in under-estimating what they drink. It can be easier not to face up to things.
GPs know this. A medic friend says…
View original post 119 more words