[mbell]

I'd encourage you to read some of these 'papers', there is often nothing to evaluate, nothing to make predictions on. Usually, just a survey and a statistical correlation test.