Hi I am doing a dissertation at the moment and I was wondering what is the best way to present qualitative data. I have done some maths on population size and margin of error, but what else can I do?
I sometimes use a US 95% measure for qumtitative results, but can I doo this with qulitative results?
I have already displayed it as bar charts and in percentagesof total vots etc. Need someting a bit meaty if possible.
CogitoErgoCogitoSum
Favorite Answer
You really cant. Anyone who does, tries, says you can, or is in the habit of doing so... is a BAD statistician and is only trying to be manipulative.
Qualitative data is qualitative specifically BECAUSE it lacks quantitative values. There is no getting around that. Questions such as "in your opinion, rate this and this from 0 to 10".... is completely arbitrary and subjective... and holds no true statistical value except to produce results that sway stupid people.
If you are truly doing a dissertation... might I recommend doing some original work... commence a study, develop a theory... publish your own paper... describing just how we might be able to accurately utilize subjective data.
Anonymous
There are some valid statistical methods you can use. Google for "non parametric methods". Sometimes you can convert qualititative data into binary (yes/no) choices which can then be treated as quantitative.
The thing NOT to do is to "invent" some arbitrary numerical representation of the data (like scores our ot 10) and then treat it the same as if it was a genuine numerical quantity like height, weight, etc.
See http://www.reading.ac.uk/ssc/workareas/participation/Quantitative_analysis_approaches_to_qualitative_data.pdf for some ideas.
Anonymous
I would say to place it into a bar or line graph.