Depth and Breadth of Pie Menus for Mid-air Gesture Interaction

ABSTRACT

Mid-air gesture interaction has become one of the most promising human–computer interaction manners. However, the principles of designing pie menus for mid-air gesture interaction are largely lacking. Currently, no study has addressed the breadth and depth of the pie menu for mid-air gesture interaction. Here, in two experiments, we found that the breadth and depth of a pie menu had distinct influences on the operation performance in mid-air gestural interaction: breadth affects both the accuracy and reaction time of the operation, while depth only affects reaction time. Overall, when operation accuracy is the main concern, the breadth of the pie menu will be the key factor, and it should not exceed eight items (Experiments 1 and 2). However, when the operation also emphasizes efficiency, the depth of the pie menu should be considered: fewer layers lead to faster responses, with two layers yielding the best performance (Experiment 2).

Acknowledgments

This research was supported by Science and Technology Innovation 2030 program (2018AAA0101605), Fundamental Research Funds for the Central Universities National, Natural Science Foundation of China Grants (31771202), and Key Program of Natural Science Foundation of Zhejiang Province (LZ20C090001). We thanked Ziwei Liu, Jinnan Song, Ke Wang for the help in data collection.

Disclosure of potential conflict of interest

There is no financial interest or benefit that has arisen from the direct applications of your research.

Notes

1. It could be argued that the characters/words used in current study were chosen from daily life rather than from commands for gesture interaction (e.g., copy and paste), and thus lacked ecological validity. In fact, commonly used words and commands should not significantly affect the results. Corroborating this view, most previous studies on menu design adopted common words instead of real commands (e.g., Kurtenbach & Buxton, 1993; Miller, 1981; Urbina et al., 2010).

2. Studies (e.g., Norman & Chin, 1988) have indicated that a menu structure offers better performance when the breadth of the underlying layer is large. For instance, a menu structure 2 × 8, wherein the first and second layer contains two and eight items, respectively, is better than a menu structure 8 × 2, wherein the first and second layer contains eight and two items, respectively. Therefore, only one menu structure was explored for a depth–breadth combination (e.g., only menu structure 2 × 8 was explored).

Additional information

Notes on contributors

Wenmin Li

Wenmin Li is a graduate student at Zhejiang University, Hangzhou. He majors in applied psychology, obtained Bachelor of Science from Zhejiang University, Hangzhou, China, 2018.

Xueyi Wan

Xueyi Wan is a graduate student at Zhejiang University, Hangzhou. She majors in cognitive psychology, obtained Bachelor of Science from Zhejiang University, Hangzhou, China, 2017.

Yanwei Shi

Yanwei Shi is a graduate student at Zhejiang University, Hangzhou. He majors in applied psychology, obtained Bachelor of Science from Zhejiang University, Hangzhou, China, 2016.

Nailang Yao

Nailang Yao is a graduate student at Zhejiang University, Hangzhou. He majors in cognitive psychology, obtained Bachelor of Science from Zhejiang University, Hangzhou, China, 2015.

Ci Wang

Ci Wang is a lecturer at Zhejiang University, Hangzhou. He majors in engineering psychology, and obtained Master of Science from Zhejiang University, Hangzhou, China, 2011.

Zaifeng Gao

Zaifeng Gao is a professor at Zhejiang University, Hangzhou. He majors in cognitive psychology, and obtained Ph.D from Zhejiang University, Hangzhou, China, 2009. His research focuses on engineering psychology and cognitive psychology.