I am going to conduct a survey in two different geographical areas. I cannot do interviews as my research and the measure of my themes needs asking many questions for the participants. my supervisor said it is either i conduct survey for at least 300 persons (as it is a PhD ) or you analyse your survey using qualitative methods? Actually I am confused how can we analyse survey data using qualitative analysis?
I am confused as well. Where did the magic number 300 come from?
Best, David Booth
David Eugene Booth David Eugene Booth that is what my supervisor said. Investigating the role of context on social capital and participation using only surveys does not give enough observation to my analysis. I was planning to conduct survey in two areas for 40 young persons , but my supervisor said that will not serve my research questions.
Was this 300 the result of (5-10) times the number of items of the questionnaire you used? I am not sure whether the 300 comes from, but 40 is too small for a survey like your study, as this may reduce the power of your study. Also, if you plan to recruit 40 persons from two areas, what kinds of person you are going to recruit? Those with better socio-economic states and a wide social network or the opposite ones? Do you intend to match the characteristics of persons from the two areas? Additionally, for using qualitative analysis to measure a survey data, there is a method called content analysis which has both qualitative and quantitative methodology (Krippendorff, K.(2004). Content analysis:anintroductiontoitsmethodology. Thousand Oaks,California:SagePublicationsInc) for your consideration. But better to confirm with your supervisor. Good luck on your study.
Hello Fawzia
One rule in the survey is that you have the most closed questions. Processing open questions is long and tedious. Anyway, I point out that there are two procedures to process open questions of a questionnaire:
1) Deductively.- When you know a lot about the topic and you know the categories you will find in the answers. In this way you assign the texts of the respondents to the categories. As a very simplified example, consider that you asked people why they use social networks and know that the reasons are 3: Fun, emotional support and meet people. Then you have to assign the open answers to these categories. Evidently, one person can have more than one reason. In the end it will have looked like a closed multiple-choice question.
2) Inductively.- This procedure is when you do not know the categories that you will find. You start with the answer of the first questionnaire, you give a name to the category. Then you go to the second and see if it is in the same category or should be a new one and so on until you get to group all the answers into categories.
Whether you go along the path 1 or 2, you can process the data quantitatively since you can do frequency counting by categories.
Regards,
Jorge
Talk it through with your supervisor, I imagine they envision a survey based largely on open questions that will provide text that can be analysed qualitatively. Good luck.
300 number of respondents is vague for me , how you took the sample or how your adviser can say at least such number of respondents ? , did he consider the number of attributes in your study? or is there some threshold for saying at least 300 respondents for PHD thesis ? sample as per my knowledge depends on either on your total population that helps to set sample frame than you can select ur sample size or the number of attributes u have commonly ( 10-15 per attributes advisable).
- Badges
- Science topic
- Similar topics
- Statistics
- Survey