retired_susan mcdonald

A student undertook a project to sample how many phone numbers were stored on people's mobile phones, and whether this was related to the age of the phone owner. He stood in Lancaster town centre on a Saturday morning and asked friendly-looking people passing by if they'd mind answering three simple questions.

Many people refused to stop, but if they did the first question was "Do you have a mobile phone with you?". If they answered "no", he let them go; if they answered "yes" he asked them their age and the number of phone numbers stored on the phone.

Some people, especially older people, refused to divulge their age; others refused to count the number of phone numbers on the grounds that there were too many; but eventually he had a sample of 13.

The sampling method used was:

simple random sampling
systematic sampling
quota sampling
cluster sampling
none of the other options

-------------------------------

is it random or systematic? or another

retired_nini59

very useful for estimating any number of things. I'm not even going to do a list.

retired_Fit

sounds like that student has waaay too much free time ....

... it's you isnt it, Mikey ?

retired_nini59

I would say cluster, he interviewed people all in the same area, thus it's likely they'll have similar characteristics, and many of the older samples were no good, so the majority of his sample will have been young, and not with many numbers as those with loads refused to count, thus skewing his data quite a lot, and making it useless tbh, but I'm not sure. Not really random, definitely not systematic, nor quota I don't think.

retired_nini59

I would guess random would be allowing a computer to randomly select people from a list of names of the whole population, systematic would be taking the list, putting it in alphabetical order and asking every 10,000th person, quota would be surveying a certain number of people from each age group/sex etc, cluster would be finding a load of people in a group and surveying them. Don't think I've studied this though, but that's what common sense would suggest.

retired_khabmth

I'd call it crap but as that isn't an option I think you'd have to go with "none".

There are serious flaws in the return rate for answers and in the completeness of answers and in the total number (well rather one of the last two but still) so I can't see the results holing any credibility/scientific value at all. My old maths teacher would have laughed and sent us to do it again (and do it right).

retired_Angela Adams

What course is this?

retired_khabmth

Think they wanted the simple random sampling as the answer though.

retired_susan mcdonald

i've got the 1st quartile as 6. Now my notes are a bit wrong I think, dunno. but to find the 3rd quartile would i do 5+((6*5)/5)=11

is that correct? I've used the L+K*W/F formula

retired_khabmth

Got to be some basic statistical course.

Seems to be given at a uni in Lancaster too