No, choosing 10 students out of 50 students will not help to assess the IQ level of all the students in the class as it is not necessary that all the students share the same IQ level. There can be discrepancies in framing the sample i.e. it can be possible that the teacher might select the students who have the highest IQ level or may choose the students with the lowest IQ. In this case sample will not represent the whole population and the study may end up with wrong conclusions.