None of that changes the fact that collecting data in this way for an academic research dissertation is fundamentally meaningless regardless of what her aim is -- it is hopelessly and irrevocably distorted by self-selection bias in ways that make any analysis futile.

If she is using other routes to collect data, then even if those routes are valid, she is seriously polluting her data by including anything from survey results collected in this manner. Including it can ONLY lessen the validity of any conclusions she comes to.

I wish her the best in her efforts, but also want her to base her dissertation results of defensible research -- because she WILL have to defend them in order to pass her course of study.