Hello Sandy,

can you explain the nature of the qualitative data a bit more? In essence, for any quantitative method, you need numbers. Hence, of you want to use qualitative data, you have to code it numerically. Optimally, you have 2 coders for that to be able to estimate the interrater agreement.

I remember of a colleague a few years ago who had interviews with African business owners. On part of the interview was where the business owners talked about their future perspective and plans. My colleague numerically coded these segments on a number of theoretically relevant dimensions, e.g., "is there a vision" (yes/no), clarity of the vision, ambition of the vision etc.

HTH

Holger

Holger Steinmetz thank you for the response. The qualitative data I have is an ordinal data, which is used to rate the reaction of plants to certain diseases from susceptible to intermediate, then resistant.

Ok, then the question (as in any other case with ordinal data) is whether you can defend a quasi-interval characteristic of these data in order to use it in a linear fashion. An alternative would be to use a set of dummy variables. If your measure is an independent variable than this is very easy to apply and to interpret. I always refrained from using ordinal/dummy variables as dependent variables as this can be very cumbersome to interpret. But i guess this is a question of experience and getting used to it.

Best,

Holger

Holger Steinmetz I am planning to include it as one of the explanatory variables. I still need to read more on how to properly handle this kind of data though.

As I said, either you rely on a quasi-interval nature and you use it as one predictor or you create a set of dummies.

This is easy.

Best,

Holger

Article On Quantitizing