The effects of scale direction on response behavior are well known in the survey literature, where a variety of theoretical approaches are discussed, and mixed empirical findings are reported. In addition, different types of survey completion devices seem to vary in their susceptibility to scale direction effects. In this study, we therefore investigate the effect of scale direction and device type on response behavior in PC and smartphone surveys. To do so, we conducted a web survey experiment in a German non-probability access panel (N = 3,401) using a two-step split-ballot design with four groups that are defined by device type (PC and smartphone) and scale direction (decremental and incremental). The results reveal that both PCs and smartphones are robust against scale direction effects. The results also show that response behavior differs substantially between PCs and smartphones, indicating that the device type (PC or smartphone) matters. In particular, the findings show that the comparability of data obtained through multi-device surveys is limited.
