JonathanLB
Nov 26th, 2002, 06:16:10 AM
Ok I was getting bored by my homework and I had some fun with it, I hope you enjoy the second paragraph here:
There are 19,000 students at Oregon State University, but contacting a representative sample of 500 students would most likely (95% certainty) yield a result that is true for the entire student body and not just the limited sample group used in the survey. To conduct a survey about the recreational use of OSU forests by the students of OSU, a random group of people needs to be selected so that no area of campus, such as only dorms, is represented and so that people of various grade levels and both genders are also represented.
Ideally, a database of all social security numbers could be collected from the registrar, by force if necessary, and numbered from 1 to 19,000 and a random number generator on a computer, for instance, could pick 500 of the social security numbers. Then, the chosen numbers would be matched with the students they represent and each of the students contacted by mail with the survey. To ensure that the students respond, one Canadian penny will be inserted into each envelope to pay the students for the value of their time. Because almost every student goes home for fall break, even if their listed address was for their parents’ house, they would still receive the questionnaire in fairly short time. If they failed to send it back by a certain date, for instance January 1, 2003, then a follow-up phone call could be made and the surveyor could even start making threats against the student’s life if he or she fails to return the survey in a timely fashion. In the event of continued non-compliance, warning shots will be fired at the surveyor’s car, apartment, or house, whichever is deemed most appropriate.
I think that deserves a 10/10 personally, but I don't think I should risk it, so I'll turn in my "tame" version.
There are 19,000 students at Oregon State University, but contacting a representative sample of 500 students would most likely (95% certainty) yield a result that is true for the entire student body and not just the limited sample group used in the survey. To conduct a survey about the recreational use of OSU forests by the students of OSU, a random group of people needs to be selected so that no area of campus, such as only dorms, is represented and so that people of various grade levels and both genders are also represented.
Ideally, a database of all social security numbers could be collected from the registrar, by force if necessary, and numbered from 1 to 19,000 and a random number generator on a computer, for instance, could pick 500 of the social security numbers. Then, the chosen numbers would be matched with the students they represent and each of the students contacted by mail with the survey. To ensure that the students respond, one Canadian penny will be inserted into each envelope to pay the students for the value of their time. Because almost every student goes home for fall break, even if their listed address was for their parents’ house, they would still receive the questionnaire in fairly short time. If they failed to send it back by a certain date, for instance January 1, 2003, then a follow-up phone call could be made and the surveyor could even start making threats against the student’s life if he or she fails to return the survey in a timely fashion. In the event of continued non-compliance, warning shots will be fired at the surveyor’s car, apartment, or house, whichever is deemed most appropriate.
I think that deserves a 10/10 personally, but I don't think I should risk it, so I'll turn in my "tame" version.