A screen and a phone call are not the same sample

An online questionnaire is fast, and it is biased toward people who live in their inbox. A phone interview is slower, and it still reaches people who will not open a link.

Neither method is the grown-up one. They answer different coverage problems. If you need both, field them as two samples and show the difference, instead of averaging it away.

Say in the method note who was eligible, how they were reached, and what you did when the two files disagreed. A blended number with no note is not a finding.