Relevance & Research Question: The mobile survey participation is currently a relevant phenomenon, worldwide. Mobile devices have different characteristics in comparison to traditional fixed-PCs or laptops. A key difference is the reduced size of the devices and of their screens. The enhanced portability is counterbalanced by some potential risks. Can the screen size affect the quality and the comparability of collected data? And which is the role that the questionnaire optimization can play?

Methods & Data: Data were collected in 2015, using the Netquest opt-in online panel for Spain, through a two-wave survey proposing twice the same questionnaire to the same respondents (1,608 panelist). Each panelist was randomly assigned, within a specific wave, to a survey condition (i.e. the device used and the questionnaire optimization, for participants using a mobile device). This research focuses only on mobile device respondents. In order to test our hypotheses, we computed a series of indicators -some based on paradata- related to both the participation process (e.g. the completion time, the answer consistency between waves, etc.) and the survey experience evaluation (perceived difficulty, appreciation, etc.).

Results: The device screen size usually does not significantly affect some indicators linked to the data quality/comparability, but can significantly affect variables (such as the survey length or the evaluation of the survey experience) that can determine the respondents’ burden and their willingness in participating again in surveys. However, most of the potential issues linked to the use of small sized devices can be attenuated proposing a questionnaire optimized for the mobile participation. The paper also investigates the interaction between these two variables and the consequences of re-proposing the same survey.

Added Value: Despite the widely acknowledged importance of the screen size, only few studies directly analyzed this key variable. Most of them just considered broader and nonhomogeneous categories of devices (i.e. smartphones and tablets) or used a limited set of indicators (mostly completion time). This work aims at using a wider range of indicators and at providing an updated and more precise view, taking into account the exact screen size (in inches) and its interaction with the optimization.