2. The people surveyed were chosen randomly so give a better insight into the population over all.
The second sampling method is a bit flawed in that it asks only 25 people, and these 25 people were sat near him and so may have similar interests therefore providing a degree of skew.