In order to scientifically validate a sample, it is necessary that,
The entire population under study should be considered to obtain an average or a mean data. If only a part of the population is considered then the result obtained would be a biased one. A biased data will not represent the entire population and so it would be incorrect.
The sample choice done must be at random covering all the areas of the population to get the accurate result. This is because random selection lowers the possibility of getting biased.
what are the choices? i would think that you keep your temper very low do not raise your voice and try to get them into a seperate room so that you can try to calm them down