Dear all,

I plan to conduct a survey with a sample of approx. 50 to 100 hotels on their innovation behaviour. The model will involve at least 10 latent independent constructs and one dependent construct (innovation behaviour). I was pointed to Support Vector Machines as a methodological alternative. However, I am not sure if this method works also with such small samples.

Furthermore, I wonder if the method implies that the data have a tendency to consist of two groups (i.e. innovators v. non-innovators), or if this requirement can be overcome through the use of a Kernel.

Thank you in advance for any advice.

Kind regards,

Hannes Antonschmidt