There’s truth in the night –
and bare it is laid
But no light in the night
and it lurks in the shade
We use statistics as a drunkard uses lamp-posts… for support rather than illumination…
A drunkard was looking for something under a streetlamp.
A passer-by asked him: “What are you looking for, mate”?
“My wallet, I lost it”
“Are you sure you lost it here”?
“Come on, certainly not, I lost it over there in the woods”
“So, why aren’t you searching it there”?
“Are you crazy, what do you think the lamps are for”?
The morale or this story applies to most statistics and social research, including pandemic, women and race discrimination, ranking cities, assessing football players and more.
At a recent Conference, I heard an enthusiastic young scholar, who had won awards for an article of hers published in a major journal, saying: “I found wonderful data, so I decided to do the research!” The research consisted in an econometric elaboration of those data. I wonder if we do research just because we have data or because we want to find something useful to do something in which we believe and, in view of that, we look for possible numerical data and other information. Thus, if there is no data available, we should either collect them or adopt different research tools.
We all live in the mediocrity of the world and know that some research is done just for the sake of academic careers. If we want to be even more optimistic, we can add that, using existing data allows social scientists to elaborate models and analytical tools that might be later applied to other situations. Hence, we should not criticize the young scholar’s declaration too severely. I am not complaining that such research was carried out by the young scholar. I do criticize, instead, that this kind of research is encouraged and awarded credit and prizes because it is at the center of academic research and concerns.
As a matter of fact, it is easier to award a prize to falsifiable quantitative research than to risk long discussion about some more arguable argument, no matter how meaningful.