[ | Next | Previous | Up ]

Re: ZIMBABWE MBA STUDENT NEEDS MENTOR

From: lloyd_neilson@bigfoot.com
Date: 23 Dec 1999
Time: 09:46:12

Comments

Hello Stan,

Thank you for your posting and offering to assist me - I desperately need it.

What I want to achieve is possibly a little out of the ordinary but I would assume that there must be theory to back same. I need to sample Internet users only in the UK population of 18 and above via the Internet.

The methodology that I want (actually need) to use is as follows:

1: Mail my contacts in UK that I know will go to the necessary site and complete the survey presented.

2: Request that they forward my request to their friends and associates only in the UK. That way I would hopefully get the replies that I needed without simply soliciting people by mail. Once the required number of responses have been received, less any parties 17 and younger, I would then be able decompose my data.

Using this approach it appears impossible to select a frame or a list from a frame, except an initial contact list, and from there everything is purely random and allowed to ramp until the desired number of responses have been received?

I know I don't need to quote the Central Limit Theorem or other theory to you but possibly for the sake of others and my own confusion it is necessary to try and clarify the situation.

"Central Limit Theorem concludes - If we take random samples of size n from a population, the distribution of sample means will approach that of the normal probability distribution. This approximation will become closer the larger is n." (Wisniewski 1997)

"This is a particularly important conclusion and underpins much of statistical inference. It implies that - no matter what type of population we are examining - if we take random samples, the distribution of these samples will approximate to the Normal Distribution. Similarly, the larger the sample sizes we use, the closer this approximation will become. The sampling distribution will approximate to the Normal distribution as long as we take sufficiently large samples. Sufficiently large is generally accepted to be a sample size of at least 30."

1: If the above approach is acceptable how many results would I ideally need to return? Obviously 30 is not acceptable in light of what I have read in other literature?

2: Is the above approach acceptable when considering the multitude of sampling techniques available?

3: For my dissertation I need to comment on the methodology used in terms of theory and its potential failings, which sampling theory/ies will I actually be working to?

As you can see I am without any real direction at present and need help! The upside is that I have time as I am working at least three months ahead of schedule. Any assistance would greatly be appreciated.

You are aware that: "Questions beget answers, which beget questions":) Thanks again for offering your help.

Kind Regards

Lloyd Neilson (Zimbabwe)


Last changed: November 20, 2007