You should undertake a quantitative study on an external validation dataset. Here is a paper that recommends sample sizes for external validation studies

Pl see the following paper. It is simple application.

Regards

Article Cross-validation – An application with volume estimation mod...

Article Cross-validation: A tool for forest modelers

Article Error Index: Another tool for model selection

Article Modelling and Validation for Volume Estimation of Eucalyptus

the kind of method you would use depends on your goals. So, can you tell us more about the topic for your model and what you want to accomplish by "validating" it? 

Thank you very much for your assistance.I developed and published four conceptual frameworks from my PhD Thesis research findings .They focus on different aspects of brand equity management in sport.My intention is now to externally validate , refine and d/eveloped the into full fledged models/theories.

Hi Lysias, 'validation' is such a grey term! like 'evidence-based' and 'saturation'. There are no standard universally accepted guidelines. I would recommend reading widely on it and using references to back up why you think your approach will produce validated results. (Thanks by the way!!)

it is based on the the tool that you had taken to build the model. most frequently it is used classification table to validate the model you drives. So you have to run the estimated model with new data set which is not being used before. you can give some cut points and find the frequencies of correct predication.

I agree that the term "validate" is problematic. One thing that might help is think about how you react to the alternative wording: "test the model." If that sounds good to you, then you probably want to do a quantitative study.

Alternatively, if your model is primarily derived from literature reviews, you might want to "refine the model," and in that case a qualitative study might be more appropriate.

We just finished validating a model so I'll share what I found to be the options. 

There are three primary ways to validate a model such as yours.  I would focus on one at a time if you have multiple models unless they are interconnected.  1) find an existing nationally-representative data set that taps all the constructs in the model; 2) write your own instrument to cover the model and validate it; 3) use validated instruments that tap all the model constructs in a study. 

If you have are lucky enough to find 1), you do data analysis such as Latent Class Analysis or KCluster analysis and you're done; 2) works much the same way only you collect your own data first.  For 3), I would suggest following the classifical measurement theory outlined in DeVellis, Scale Development: Theory and Application to write your own questions etc.  I would also suggest conducting a thorough literature review and including not only questions that tap your model but also questions and constructs that test other variables that have been found to be significant predictors in the research literature.  In our case, we used exploratory factor analysis to reduce the number of items and confirmatory factor analysis to finalize the instrument based on the model.  We then used LCA to detail the relationships among the variables in the model.  3) is a ton of work.  Good luck!

Thank you very much for the valuable and eye-opening answers.