According to the results of the meta-analysis by Nielsen and Levy (1994), there is a strong positive association between user task performance and subjective satisfaction. Inconsistent with the findings of Nielsen and Levy (1994), the main finding in the current study (based on the results of the correlation and RM-ANOVA analyses) is that outstanding task performance does not always guarantee a high level of user satisfaction, as suggested by the weak positive relationship between the two variables. In fact, both the fingertip and rubber-tipped stylus, which were faster than the acrylic-tipped stylus, resulted in outstanding performance; there was no significant difference between these two input devices in terms of task performance. However, compared to the rubber-tipped stylus, the finger had a low level of user satisfaction. In terms of display, the elastic display was outstanding in task performance compared to glass, but no significant difference was found in user satisfaction. In general, most smartphone or tablet users are familiar with various styli due to exposure via news articles, hands-on experience in the store, or prior experience with other typical styli; however, the soft-surface touchscreen used in this study might have been less familiar than the styli. Moreover, the novelty of the soft touchscreen might lead to an increase in uncertainty about the product quality. Accordingly, designers should make an effort to consider both the relationship between perceived familiarity and novelty as well as the relationship between user task performance and subjective satisfaction in the interaction design process, thereby handling the conflict between them.
In terms of task performance, the results of the ANOVA revealed an inconsistent performance trend compared to related studies, such that the fingertip and rubber-tipped stylus were found to be faster than the acrylic-tipped stylus regardless of the softness of the touchscreen. Lee and Zhai’s (2009) finding that task performance with a stylus was slightly faster than with a finger (without any auditory or tactile feedback for tapping) is inconsistent with the findings reported herein. While Cockburn et al. (2012) showed similar results, such that the finger yielded faster