I am planning a study on political behaviors to understand factors considered when casting votes. The study is aims to collect data from various socio-economic strata of the community to represent social and economic groups. So, I need experts’ knowledge to select most suitable sampling method to select respondents to this survey as I am planning a questionnaire survey to collect primary data.
I suggest using online survey tool and large sample. You do not need strata in that case since your survey respondents are from every strata. The only important thing is, collect the relevant demographic data (education, income, region or city, etc.).
To analyze that you will use advanced statistics like PCA and segmentation.
To avoid non-response bias, you will apply raking.
Please find my relevant papers below.
Conference Paper Manage sample and population differences by weighting using ...
Article Public perception of bioenergy in North Carolina and Tennessee
Conference Paper Understanding Markets and Distribution (Stakeholder Perceptions)
Article Systematic Review of Bioenergy Perception Studies
Dear Robert Istvan,
Many thanks for your very useful answer with reading materials
Political behavior of people are likely to be varying by age, education, religion, caste and political affiliations. In India, all these factors are known to play a role in political behavior of the people. So, it is expected that your sample should be representative with respect to above variables. Based on your area and country, the factors may vary a bit but you have to anticipate first and then go for the sampling. A random sample may not be a good choice. However, a stratified sampling representing one or more of the above characteristics may be considered for sampling. If a area is uniform with respect to religion or caste then we may go for other factors to represent in your sample. Online survey will give you a sample of only educated people that too who are well versed with computer and its usage. I consider this as a bias sample ignoring people with less education and may be with poverty. An area survey like done in the case of health surveys should be attempted. Even door to door survey using systematic sampling can be a good choice. As per your own statement, you want to consider various social economic groups for the survey so the choice should be of stratified sampling.
Dear Ramnath,
Thanks for your answer. Your insights are very useful. I am also thinking a more representative sample in order to obtain views of various sections.
We have many literature on this matter in internet. I think,some of earlier studies on Political Behaviors shows a clear path to solve problems tied with Method. two of them were quoted below as URL.
http://www.uky.edu/AS/PoliSci/Peffley/pdf/473Measuring%20Public%20Opinion.pdf
https://academic.oup.com/ijpor/article/23/1/24/788078
Depends on Your available resources. Online survey was mentioned previously. difficult though to control who answer your survey if you do not have control Of your sample beforehand (by characteristics of respondents). Stratified sample is way forward by socio-economic strata, age, gender. If you get unbalanced responses, and you have enough responses, you may treat them (each sample) indenpently,
goid luck! Vania
Dear Vania Ceccato,
Thanks for your answer. Yes, I have decided to go ahed with the statified sampling method considering many factors. Will see how the results comes out.
In India random survey wont work. As population is heavily divided demographically economically and they have caste system. It has to be clustered or stratified sample .If population if entire country or state at least 2500 responses would be needed . The behavior is different for urban lower and middle class and rural population Literacy is also important
- Badges
- Science topic
- Similar topics
- Statistics
- Survey