The number of online panellists accessing surveys with mobile devices is increasing rapidly. Within the LINK Internet-Panel in Switzerland 28% of the panellists fill in questionnaires with a smartphone (17%) or a tablet (11%). Yet, little is known about the influence of the screen size on survey data quality. How do answers given on a small screen device (e.g. smartphone) differ from answers given on a large screen device (e.g. laptop or desktop computer)?
In the presentation we would like to show our latest scientific results from a methodical study.
Method: Online survey with optimised questionnaire for large and small screen devices.
Sample: Representative Swiss sample from LINK Internet-Panel and LINK Mobile-Panel. These panelists regularely answer questionnaires on both large screen devices and small screen devices (e.g. Smartphone). Applying an identical split sample approach, we examined the effects of the device used for answering on the given answers:
Group 1 had to answer the questionnaire on a large screen device (invitation by e-mail)
Group 2 had to answer the questionnaire on a smartphone (invitation by SMS)
Sample sizes: N=546 for group 1 and N=421 for group 2
Participation rate approximately 45%
Results: Differencies in answering behaviour of panelists using a large screen device compared to panelists using a small small screen device. Results show device effects on various question types such as open questions (partially small significant effects), matrix questions (frequently significant effects), closed questions (no effect), etc.
Added Value: Sharing new and relevant results about the influence of the screen size on the answer behaviour in online surveys which help to better cope with unintended mobile respondents.