Our August 2019 issue starts with an essay from Carol Smith, “Intentionally Ethical AI Experiences.” In this essay, she explains how bias can enter artificial intelligence applications and the important role user experience professionals can play in reducing bias in the AI development steps of content/curation, model training, and maintenance. In addition to the essay, this issue includes three research papers: one using augmented reality to assess product usability, one discussing different methods for controlling smartphone games, and one comparing rating scale item formats.
The first article is “Applying Tangible Augmented Reality for Product Usability Assessment,” by Young Mi Choi. Her study was designed to explore the use of augmented and tangible augmented reality when evaluating product usability. She reported greater success with tangible than standard augmented reality, especially for products with physical controls, but also reported limitations in the assessment of perceived ease-of-use for both forms of AR.
The second article is “A Comparison Study of the Smartphone Gaming Control,” by Sang-Duck Seo and Sunghyun Kang. They studied the effectiveness between analog and digital game controls for smartphone games, using a mix of qualitative and quantitative user research. They concluded that direct touchscreen interaction was more effective for two-handed input control and indirect physical control was more effective for one-handed control.
In the third article, “Comparison of Four TAM Item Formats: Effect of Response Option Labels and Order,” Jim Lewis used a revised version of the Technology Acceptance Model to measure the experience of use rather than expected likelihood of use. Participants rated product experience with one of four versions of the modified TAM created by crossing two independent variables: response format (labels or numbers) and response order (increasing from right-to-left or from left-to-right). The results indicated that the item format variations did not lead to any important differences in the magnitude or structure of modified TAM measurement, but there were significantly more response errors when the magnitude of agreement increased from right to left.