We've simply implemented the recaptcha code as it was given. The goal is to evaluate how well it would do were it used in the wild, so we don't provide any extra instructions.
to be fair, I think that may be part of what they are evaluating: existing captcha systems compared to some others. So the lack of instructions from recaptcha is a flaw in recaptcha, not the survey
Chair is to wall: like 'd00r' is to "mAn" and the An$wer is; bedroom.
I was so flabbergasted I didn't even know where to start, and I had to move on just to see what else was in there---the rest of the captchas all seem pretty mundane.
Edit: found it again, fixed it to be direct quote. I'm looking at it again now and I still have no idea what I'm supposed to do with that.
Without input validation, and without any clear motivation as top why this experiment was important, I was perfectly happy entering garbage or blank forms.
What I'm getting at is that without any real motivation for submitting a form, I don't really care what I enter. This is different than when I enter a captcha "in the wild" - in that case I would try really hard because I care about successfully submitting the form.
I'm not sure what this experiment is trying to prove, but without that motivation, I don't believe it's a proper simulation of real captcha use.
Just kidding. You could always do a drawing. Ask for a donated ipad. everyone who answers 100 or more catchpas correctly are eligible to be in the raffle.
I got about 4 questions in and then it started giving me basic authentication boxes saying the site is password protected, and I couldn't do anything or click any links on the site without logging in or something.
16 comments
[ 2.3 ms ] story [ 47.9 ms ] threadI had little difficulty with the others, but the lack of instructions might bias the results against recaptcha.
Edit: found it again, fixed it to be direct quote. I'm looking at it again now and I still have no idea what I'm supposed to do with that.
What I'm getting at is that without any real motivation for submitting a form, I don't really care what I enter. This is different than when I enter a captcha "in the wild" - in that case I would try really hard because I care about successfully submitting the form.
I'm not sure what this experiment is trying to prove, but without that motivation, I don't believe it's a proper simulation of real captcha use.
Just kidding. You could always do a drawing. Ask for a donated ipad. everyone who answers 100 or more catchpas correctly are eligible to be in the raffle.
(Chrome on OS X with Flash installed, if that makes a difference)
This was my favourite one. Easy to solve -- not at all trivial to parse mechanically.
maybe something like, "How many birthdays has a fourteen year old has?"
Or: how many inches of sandwich do you get if you get a half-foot long?