Statistics Help! Sampling

Associate
Joined
27 Jan 2005
Posts
830
I have research to do for a particular research module.Though, I am rather stuck on which sampling technique and what sample size to choose.

I am wanting the views of IT workers in the UK and their views on a particular issue, but I require the sample to be 50/50 because I will be comparing the male and female views. I.e, 60 male and 60 female in my survey.

However, I do not have access to a known sample frame number (afterall it would not be possible), so therefore am I going to use non-probability sampling with quota sampling? (quota sampling because I want the 50/50 gender rate) I don't want 1000s of respondents as that's not management.

Any help would be great.
 
Associate
OP
Joined
27 Jan 2005
Posts
830
How are you collecting data? Qualitative or quantitative? Also, where are you getting your respondents from?

I will be collected data via a online survey (google docs or something like survey monkey) and will be quantitative.

The survey will be posted to an online forum (no idea if I can name them or not) but it is relevant as the people that post there are part of the research theme. Obviously it is optional for them to visit the forum and click on the link in the post.

However, it's important to state that the aim research is to compare male and female views on a particular issue. So, it will be 50/50 between male and female. I think it's non-probability sampling & quota sampling, but not 100% sure.
 
Man of Honour
Joined
26 May 2012
Posts
16,180
pretty sure if you want a representative view of IT workers, then having a 50/50 m:f ratio inherently introduces bias?
 
Caporegime
Joined
18 Oct 2002
Posts
32,615
I will be collected data via a online survey (google docs or something like survey monkey) and will be quantitative.

The survey will be posted to an online forum (no idea if I can name them or not) but it is relevant as the people that post there are part of the research theme. Obviously it is optional for them to visit the forum and click on the link in the post.

However, it's important to state that the aim research is to compare male and female views on a particular issue. So, it will be 50/50 between male and female. I think it's non-probability sampling & quota sampling, but not 100% sure.

You don't need to keep a 1:1 ratio of males to females to do a age deer comparison, just make sure your have a large sample size. Throwing away male data or specially selecting female data is not a good idea.
 
Soldato
Joined
23 Jul 2009
Posts
14,083
Location
Bath
I have to agree that this might be problematic. If you are using quantative data then there's no reason not to include the surplus male data and average it out, rather than introducing bias by selecting specific data.
 
Back
Top Bottom