My Statistics professor started his first lecture with a quote that went: Lies are of three kinds: lies, terrible lies, and Statistics!
I was reminded of this while reading this ToI report. It’s a report about the eating habits of teenagers, who apparently skip meals out of choice because they think they eat too much.
I agree – it’s not a good trend. And yet, when a ‘study’ churns out numbers, what exactly do those numbers mean? For example:
The survey conducted by the Schools Health Education Unit, found that one in 10 teenage girls between 14 and 15 years missed two meals a day thinking they are overweight.
One in Ten? Among what strata of the population? The filthy rich kids? The lower middle class? The chai-wala‘s daughter? The kind of people who read the Times of India? The kind of people who cannot afford a daily newspaper? Who?
It’s incredible how anyone and everyone churns numbers out – without the slightest attempt at a qualifying statement. The numbers are from the chosen sample population, not the entire population. When the numbers are extrapolated, it is assumed that the population is identical in diversity to the sample population. Are these extrapolations even valid? Do the researchers know? Should we know when they churn out numbers at us?
Maybe they are – maybe the study was thorough in its research and actually sampled from all possible strata of society. But even then – shouldn’t there be a qualifying statement?
I’m not taking away from the problem of concern here. Sure, it’s a serious issue that needs to be looked at. Sure, having a study with concrete numbers helps bring attention to the concern. And yet, I can’t help but think that – forget skipping meals – more than One in Ten children of 14-15 years of age in India would thank their respective Gods if they got one full meal a day.
(link via India Uncut.)