dezzy said:
they could have used SAM to find out.
Not really. I can't imagine there's really much data which allows someone to correlate aggregates with uais below 30.
Also, most of the data regarding scaling seems to be in smaller increments when it comes to higher marks (i.e. the higher your mark, the more accurate sam can be in determining what it will be scaled to). This seems to fit with the fact that the higher your uai, the more accurate sam usually is in guesstimating it.
e.g. people with uais over 99 usually had guesstimates accurate to about +-0.30, while those over 90 were usually within a couple of points, while those with lower uais could have quite large fluctuations.
I'd hate to think how far off it could be when guesstimating below 30.
Also, what's with comparing average mark to uai? How are they related?