Respuesta :
It depends on your population size, and if your sample is biased. If your sample was 40 out of 100, then yes, it would be reasonable. But if your sample was 40 out of 1000, then no because the sample is too small.