Type I and Type II Errors are Real Possibilities

(image credit)

Statistics, research and analysis often have to work with uncertainties and ambiguities, despite their occasional posturing as precise and accurate. We as human beings are blessed with intelligence, yet it seems we are also cursed with limits and pitfalls in the very thinking capabilities that give rise to that intelligence.

So imagine a crowd of, say, 1000 people. They’re milling about in a mall, heading in and out of shops, as you watch them from a high balcony. Depending on what you subscribe to, as far as the attributes of people are concerned, you may decide that among these 1000 people, a great majority are good guys but a notable minority are bad guys. What’s more, you’re privileged to have varying, but less than complete, amounts of information on every single one of them. Your job is to identify, and of course stop, any would-be bad guy intent on committing harm.

Got the scenario, so far?

You commit a Type I Error, if you dismiss a guy who’s actually bad. You commit a Type II Error, alternatively, if you arrest a guy who’s actually good. Remember what I said about statistics at the outset, and remember that any information about any one guy is bound to be incomplete. So the likelihood of either a Type I or Type II Error is very real, in our shopping mall scenario or any other human situation.

That’s the rub, for me, as I read Google Adds (Even More) Links to the Pentagon about the (even more) intricate relations between Google and the US military.

I can feel reassured that the Pentagon serves its purpose of protecting us Americans, by drawing on Google’s far reaching and sophisticated means for gathering, organizing and analyzing information on the internet. I can also feel afraid that because there is no such thing as perfect, foolproof means of intelligence, a Type I or Type II Error can occur.

You see, emotions as far as our lives are concerned are not either-or, mutually exclusive phenomena. So I can very well feel varying degrees of both reassured and afraid at any given moment. Nonetheless, I can pray, too, that these giant organizations that impact our lives, whether we like it or not, avoid any such errors and do indeed safeguard us, above all.

Thank you for reading, and let me know what you think!

Ron Villejo, PhD