Please provide me with references or sources where authors have used email and social media in the same study to collect data and clearly explain how the two sources of participants should not have created sub-group differences within the sample.