How to proceed analysing survey data


New Member
Dear All,
I am new to survey analysis and trying to analyse some survey data based on individual and their knowledge on their local leaders. I have data such as city, nature of employment of the individual, the kind of items (a fridge or a TV) they have at home, the activities they take part in, their religion, names of their local MPs etc and if they are satisfied on the work done in their area, who could be the best MP and who in their view is most devious etc.
the data is seldom hierarchical and the categories are assigned numerical codes.

What is the type of analysis that I can do with this data to infer some predictive results?
I suppose multiple regression, CHAID analysis and cluster analysis is what seems to be recommended. Can i simply do regression in Excel selecting just two columns of data from this entire list? That would be simple regression but how can i do multiple regression?

Apologies for the lengthy question and i really will do appreciate any help here.

Many Thanks and Regards


TS Contributor
Hi ejain,

The appropriate analysis for you will depend with the kind of results you are looking for. Both regression and CHAID are good predictive models, if you wish to predict, yet Cluster is more a classification tool.

The real issue is that survey data cannot be analyzed in the same way as other data, since the samplig scheme is usually special. Therefore, I'd suggest reading a bit about Survey Data analysis: