The results indicate that these three factors are reasonably useful attributes, but not resoundingly so. They indicate that in about 7 per cent of cases, the sign can be differentiated on the basis of these three factors alone.
It must be remembered that on average, if a random guess was taken the
error rate would be just under 99 per cent (since there are 95 signs
-- choosing one at random there is a probability of 
that you would be right).
Thus using the above attributes alone is seven times better than guessing.
At this stage, no statistically significant conclusion about the relative merits of the learners can be made.