Homework #10 Solution

19.1	False.  It is not feasible to do a simple random sample
	of the entire population.  Instead, they will likely 
	conduct a stratified sample.

19.4	A large number of those with college degrees may be 
	living elsewhere in the community, rather than within
	a traditional ethnic neighborhood. 

19.5	Probably too high.  The households with fewer people 
	(e.g., one person who works) are less likely to have 
	someone at home, so they will be underrepresented in 
	this method.

19.6	No.  It is not clear that all Stanford students will be 
	equally likely to walk through this area of campus, so
	the sample could be biased in some way.

19.7	ii. is better.  The percentages 48% and 52% correspond 
	to 480 and 520 heads, respectively, which comprise a 
	larger range than found in i.

19.10	You should pick the numbers 45 to 55 inclusive. Your
	chance of winning is about 73% (see p. 317(b)).

19.12	a) "Who's who" is intended to represent people who are 
	more likely to hold class office, make good grades,
	go to college, etc., than the typical student. 
	b) No.  There is no information given about how
	students were selected to be in the sample, and there
	could be significant non-response bias.