Originally, I wanted to use random stratified sampling for my survey involving an automotive industry. The sample is based on the value chain comprising of manufacturing, sales, quality finance and HR. The composition of the respondents follows the percentage breakdown of the population. I did not have problems with sampling during the pilot but during the main survey, there were a number of organizational changes that resulting in poor response from the sales division with only less than 5 respondents out of a target 150 respondents. What should I do? I am consideringremoving the sales division and focusing on the balance and considering focusing the study on the supply chain. Alternatively, could I use all respondents minus sales but still consider the survey to be on the automotive industry? Any other suggestions?
Hi Chan,
In my view, whether to drop the 5 respondents from sales dependent on how significant the sales respondents play in your research which goes back to your research objective, conceptual framework / research model, research methodology e.g. why stratified sampling etc. For example, if your study touching on constructs pertaining to sales vs others like manufacturing, HR etc comparison then, soliciting responses from sales respondents deemed important. If not, may be you can exclude sales respondents but still advisable to pen down this as 1 of the limitations at the end of the research thesis / article. All the best.
Regards,
Fung
Dear Tze Leong Chan,
In this type of analytical process, the survey method is one of the best instruments for collecting data necessary for the purpose of objectively conducted scientific research.
I invite you to the discussion.
- Badges
- Science topic
- Similar topics
- Political Science
- International History and Politics
- Globalization