Hi,

I am doing Mixed-methods research. I have completed the first phase (interviews) and I am now moving on to the final phase - the survey. This is where I have hit a wall and need some help.

The companies that I need to survey are not in a specific population (e.g. cat food manufacturers). The companies I am interesting in across a variety of industries. They do not have any trade organisations for example that represent them exclusively. I have been compiling a list of companies that manufacture the type of products I am interesting in (and the basis of my PhD) since I started my research.

I want to make my survey as valid as possible but how do I define a population when I do not know what it is? I have done a fair amount of reading on the topic and I think my approach will be Purposive sampling. If this is the case can I just email my survey to specific personnel in the companies that I have identified and then be able to defend my approach?

All advice welcome:)

A.

I have just begun Real World Research by Colin Robson.

Its definition in the glossary of Purposive Sampling is 'choice of sample of participants based on the requirements of the research questions or theoretical considerations. generalisation of findings must be based on non-statistical bases'.

From my reading its all depends on the research questions what type of sample you use. If the particular data collection method is a survey for which you will be having quantitative responses I don't think purposive sampling is correct. It is used where there is qualitative data collected. i.e. those in word form, observation, interviews rather than survey. unless the responses in your survey will be of qualitative form which then would suggest purposive sampling is a correct term to describe your data collection sample.

Hope this helps.

Kitten