Post-transitioning user performance on cross-device menu interfaces

Abstract

• First study to investigate user performance with multiple menu interfaces across devices with focusing on performance after transitioning from one device to another. • Menu configuration, menu length, and menu item order across devices significantly affected post-transitioning user performance. • Users performed better, had more efficient learning-curve, and required lower NASA-TLX workload with menus designed with consistent item order across devices compared with menus that had inconsistent item order. • Menu layout across devices had no significant effect on post-transitioning user performance. Multi-device user-interaction behavior creates a need to design cross-device menus that support users to re-locate menu items efficiently and effectively after the transition from one device to another. We conducted two laboratory experiments to investigate the effects of specific designs of cross-device menus on post-transitioning user performance (selection time and selection error). In Experiment 1, a total of 72 participants moved between 32 desktop and mobile menu interfaces (representing 16 different designs of cross-device menus) to re-locate menu items. We investigated the effect on user performance of the cross-device menu configuration (complementary versus redundant items), menu length (short versus medium), menu item order (consistent versus inconsistent), and menu format (text-only versus text-and-icon items). The results showed significant main effects of configuration, length and item order with large partial eta-squared sizes. Users performed better with redundant menus, short menus, and menus with consistent item order than they did with complementary menus, medium menus, and menus with inconsistent item order. In Experiment 2, 20 participants moved between eight desktop and mobile menu interfaces (representing four different designs of cross-device menus) to re-locate menu items. We focused on the effect on user performance of the menu layout (horizontal-desktop and vertical-mobile versus vertical-desktop and vertical-mobile) at two different lengths (short versus medium). The results showed no significant effect of menu layout, and a significant effect of length with a small partial eta-squared size. Our study provides a foundation to researchers interested in further investigation of cross-device menu designs, and to practitioners in designing cross-device menus that allow more efficient and effective re-locating of menu items when moving between devices.