Remember that random sampling is designed to equal-ize the chance of every case, including the odd ones, turning up. Thegeneral method for sampling to avoid the effects of conventionalthinking is quite different: it consists of maximizing the chance of theodd case turning up.
A different sampling method stressed here is actually theoretical sampling, which has a different goal from that of random sampling.
No one will read this description without arriving at a conclusionabout the misery of lives lived in these surroundings
The power of complete description in the example.
You never get beyond the col-lecting, because there is so much music to collect.
The author takes the field of ethnomusicology as an example. He shows that full description may lead to the problem of spending too much time and energy on the collecting process, which explains why sometimes description is also not an appropriate method.
solve the sampling problem by describing
Description may solve the problems of sampling mentioned.
what kind of an organization could be the whole ofwhich the thing we have studied is a part.
Another possible problem of random sampling.
The relation of avariable’s value in the sample to its value in the population is a prob-lem,
One possible problem of random sampling.
The problem with synecdoche, or sampling, seems at first to bethat the part may not represent the whole as we would like to think itdoes, may not reproduce in miniature the characteristics we are inter-ested in, may not allow us to draw conclusions from what we do knowthat will also be true of what we haven’t inspected ourselves.
Three possible problems of sampling.
We need the sample to persuade people that we knowsomething about the whole class
Why sampling?