Fundamental to any research are the people you wish to ask for information.

If you want quantitative estimates, then you need to be able to define a universe and identify people who fit into that universe. If you are seeking qualitative information, this is less critical as long as you have suitable representation and can interpret the results accordingly.

It is really determined by what you want the information for, but the key issues are:

• will the sample represent the population that you need
• can they be contacted by suitable means (methodology)
• how many responses do we want (sample size)
• how many people do we have to contact to get that number of responses (response rate)
• do we have a population profile which will enable us to select the sample and/or weight the results appropriately
• what can we do to encourage participation (incentives)

Ideally, we want to be able to use the results to make decisions, determine our actions or at least drive our models to help us with that. Anticipating how the information will be used will help you decide whether the sample can provide you with valid information.