There's nothing original about saying that perceptions are often far from reality. Nor is there anything original about saying that outward appearances shape the way we perceive others, and vice versa. But here's something I've never seen studied formally but is still quite valid -- dorky does not necessarily equal 'smart.'
It's often that you hear people with traditionally "dorky" characteristics -- glasses, pastiness, lanky frame, socially awkward, non-mellifluous voices and/or speaking patterns, nervous tics, uncoordinated movements, etc. get labeled 'smart' by other people they work and/or live with.
But are they really?
Sometimes, yes. But often, they're not.
A lot of the people I worked with at my last command were very, very intelligent. Yet to a man, almost none exhibited any of the above characteristics, so the average person who just saw them or very briefly met them would not be quick to slap them with the 'smart' label.
I've also spent some time at some civilian three-letter agencies filled with analysts who meet many, or even all, of the above characteristics that I associated with 'dorkiness.' I think they'd get called 'smart' by most people in an instant-impression sort of setting, even without having to 'prove' it an any other way.
But here's the rub: If you're actually listening to people and gauging things like originality/range of thought, intellectual curiosity/flexibility, general situational awareness, etc. (and for the sake or argument, let's assume these are the things that constitute 'smart') you should challenge who you're considering 'smart' based on appearance -- or for that matter, pedigree.
What you find might surprise you -- a random sample from the football team might be a lot smarter than a random sample from the chess club.
And your mailman might be a lot smarter than someone who likes to talk about his three Master's Degrees.