The study was conducted using the Apple Watch Series 10. Participants ranged in age from 24 to 35 years old (20 males and 25 females), covering the primary demographic of smartwatch users in China. According to recent market research data, smartwatch users in China are predominantly younger adults38. Therefore, selecting participants from this age group helps ensure the relevance and applicability of the findings to the mainstream market. To control for potential confounding variables such as differences in cognitive ability, device operation experience, and vision across age groups, this study did not include older adults or participants with professional athletic backgrounds. All participants had prior experience using smartwatches, with usage durations ranging from six months to five years and a minimum usage frequency of three times per week. Their primary uses included health monitoring (e.g., heart rate, exercise intensity) and daily notifications. Participants were all familiar with basic smartwatch operations and required no additional training to complete the experimental tasks. All had normal or corrected-to-normal vision and no known visual impairments. The experiment lasted approximately 15 min.
Prior to the study, an a priori power analysis was conducted to determine the adequacy of the sample size. Using G*Power software and setting a medium effect size (f = 0.4), significance level (α = 0.05), and power (1–β = 0.8), the required sample size was calculated to be 15 participants per group. Accordingly, participants were randomly assigned to three groups (15 per group), each participating in one of the three movement condition experiments.
Experimental material
Alshehhi et al. found that users prioritize heart rate data when monitoring their health37, which is why heart rate was selected as the data object for the experiment. Dong et al. identified that users’ primary motivation for checking their smartwatches is activity tracking and health monitoring. Among various exercise-related data types, users considered real-time heart rate, exercise intensity, and average heart rate as the most essential. The study further revealed that users showed the highest satisfaction with interface designs that combined two types of data, and that bar charts were the most effective presentation format in terms of clarity and user preference4. Therefore, heart rate and exercise intensity were the main focus of this experiment. Islam et al. identified several common data visualization formats, including text only, chart only, icon only, icon and text, chart and text39. According to Li et al., graphical displays outperform text-based formats on mobile devices31. Based on these findings and prior research on data presentation and visualization, graphic-based data visualization was considered more effective than text-only presentations. Accordingly, the experiment employed text-based and graphical forms of presentation, as shown in Fig. 2. The heart rate and exercise intensity data were presented in a digital table format, with exercise intensity categorized into low, medium, and high according to the American Heart Association. In the text category, exercise intensity was presented directly in text form, while in the chart category, intensity was represented using bar graphs.
Presentation of text and graphics.
The numerical values displayed in the interface represent real-time heart rate, while exercise intensity is categorized into low, medium, and high levels based on the classification standards of the American Heart Association. According to previous studies by Blascheck39, and Dong4 bar charts have demonstrated higher recognition efficiency and user preference on small-screen devices such as smartwatches, particularly when presenting health-related data like heart rate zones and exercise intensity. A preliminary exploratory test was conducted to compare the readability and user preference of three common chart types: bar charts, line graphs, and pie charts on smartwatch screens. Several expert users indicated that bar charts were the clearest and most intuitive format. Therefore, textual information was displayed using direct text, while graphical information was visualized with bar charts. The interface font and color scheme followed Apple Watch design guidelines, and the screen size used in the study was 42 mm.
This study focused on the visualization of health information in exercise scenarios, such as running or walking, which typically take place during daylight hours or under well-lit conditions. As such, the experiment was designed to simulate daytime exercise usage, making the results more relevant to typical user contexts.
The entire experiment took place on a treadmill to control environmental variables across different physical activity conditions, such as running speed, thereby enhancing the consistency and comparability of the data. However, we acknowledge that treadmill-based scenarios differ from real-world outdoor exercise conditions in several aspects, such as limited arm movement, lack of terrain variation, and reduced ambient light interference. These differences may affect users’ visual perception and interaction behavior in dynamic contexts. Although the treadmill setup aimed to simulate a high-intensity movement state, its ecological validity remains limited. Nevertheless, studies such as Mo et al.22 have also adopted treadmills as a controlled simulation method, indicating that this approach is widely accepted in wearable device research.
Experimental design
This study employed a two-factor mixed experiment design, with the between-subjects variable being the motion scenario (static, Low-intensity movement, and High-intensity movement) and the within-subjects variable being the presentation format (text vs. diagrams). The measurement indicators include cognitive comprehension performance (cognitive efficiency) and subjective perception indicators. Cognitive efficiency refers to the total time taken by the user to complete the task, which serves as an indicator of the user’s ability to understand the diagram. Another approach involves subjective measurement, quantifying the impact of different diagram types and interaction forms on users’ subjective perceptions40. In the experiment, the factors affecting users’ ease of reading were categorized into there areas: perceived ease of use, perceived usefulness41, and satisfaction42, as shown in Table 1. The usability questionnaire for the visualizations in this study included three latent variables and eight measurement items. For each latent variable, the mean score of its two corresponding items was calculated as the overall score43. All items were rated using a 7-point Likert scale, ranging from 1 (strongly disagree) to 7 (strongly agree).
Experimental procedure
The experimental procedure consisted of five steps, as illustrated in Fig. 3. First, upon entering the laboratory, participants were informed of the purpose and procedure of the study and signed an informed consent form. Next, the experimenter explained the task objectives and the interface operation, followed by a short practice session to ensure participants fully understood the task requirements. Once participants were familiar with the process, they proceeded to the performance task, in which they viewed the smartwatch interface and completed visualization-related tasks while engaging in physical activity, according to the experimental design. After completing the task, participants filled out a subjective evaluation questionnaire, and some were invited to participate in a brief interview to provide additional feedback. Finally, the experimenter expressed appreciation for their participation, provided compensation, and concluded the session. Photographs of the experimental procedure are presented in Fig. 4.
Photographs of the experimental procedure.
Results and analysisCognitive Understanding of performance indicators
Users performed better with diagram formats than with text formats, and the presentation format had a significant effect on users’ cognitive comprehension performance, F(1,42) = 24.356, p = 0.006, partial η² = 0.367, as shown in Table 2. The effect of the presentation format and motion scenarios on task performance and subjective perception is also presented. The results demonstrate that the chart presentation format helps users access information more quickly, regardless of the scenario. Performance in High-intensity movement scenarios was worse than in static and micro-dynamic scenarios, and different motion scenarios significantly impacted users’ cognitive comprehension performance, F(2,42) = 9.909, p
Subjective perception indicator
The presentation format significantly affects users’ scores on perceived ease of use (F(1,42) = 7.758, p = 0.008, partial η² = 0.156), perceived usefulness (F(1,42) = 8.455, p = 0.006, partial η² = 0.168), and satisfaction (F(1,42) = 11.736, p = 0.005, partial η² = 0.172), with graphical formats outperforming text-based formats in all these subjective perception metrics. Users’ subjective scores in strongly dynamic scenarios were lower than those in static and micro-dynamic scenarios, with significant differences observed only in perceived ease of use (F(2,42) = 4.091, p = 0.024, partial η² = 0.163). Pairwise comparisons revealed significant differences between strongly dynamic scenarios and both static and micro-dynamic scenarios.
Experiment 2: the effect of different rendering animations on user preference and performance in different motion scenariosParticipant
The participant requirements were consistent with those in Experiment 1. Similarly, participants were randomly assigned to three groups, with 15 individuals per group, to participate in the experiment under different movement conditions. To ensure data independence and accuracy, the participants in Experiment 2 were not the same as those in Experiment 1.
Experimental material
Lu et al. categorized data chart animations into three types: traveling ants, gradual emergence, geometric morphing44. The “marching ants” animation is primarily used for presenting multiple sets of data dynamically. Based on the optimal presentation form identified in Experiment 1, this study adopted three animation styles: no animation, gradual appearance, and geometric transformation. The gradual appearance involves chart elements transitioning with color shifting from gray to orange and text fading in from invisible to visible. The geometric transformation involves designated content (such as the orange blocks and text) enlarging temporarily before returning to their original size, thereby drawing the user’s attention to key information, as shown in Fig. 5.
Experimental design and procedure
This study employed a two-factor mixed design, with the between-subjects variable being the motion scene (static, Low-intensity movement, and High-intensity movement) and the within-subjects variable being the presentation form (no animation, gradual emergence, and geometric morphing). The measurement indices and experimental procedures were consistent with those used in Experiment I.
Results and analysisCognitive Understanding of performance indicators
Users’ performance with the animation presentations was lower than that in the unanimated condition, and the presentation animation had a significant effect on users’ cognitive understanding performance, F(2,84) = 32.557, p 3. Pairwise comparisons revealed significant differences between each of the animation presentations, with performance in the unanimated condition significantly higher than in the animated conditions. Furthermore, the performance with the gradual emergence animation was significantly lower than with the geometric morphing animation. Users’ performance in High-intensity movement scenes was inferior to that in static and Low-intensity movement scenes. Different motion scenarios had a significant effect on users’ cognitive understanding performance, F(2,42) = 12.529, p
Subjective perception indicator
The presence of animation had a significant effect on user satisfaction (F(2,84) = 5.979, p = 0.004, partial η² = 0.125) among the subjective perception measures, with animated formats yielding higher scores than non-animated ones. However, no significant differences were observed in perceived ease of use and perceived usefulness. As for movement scenarios, a significant difference was found only in perceived ease of use (F(2,42) = 6.262, p = 0.004, partial η² = 0.23), with scores being lowest under high-intensity movement conditions.
Experiment 3: the effect of different color rendering modes on user preference and performance in different motion scenariosParticipant
The participant requirements were consistent with those in Experiment 1. Similarly, participants were randomly assigned to three groups, with 15 individuals per group, to participate in the experiment under different movement conditions. To ensure data independence and accuracy, the participants in Experiment 2 were not the same as those in Experiment 1.
Experimental design and procedure
In this study, a two-factor mixed design was employed, with the between-subjects variable being the motion scene (static, Low-intensity movement, and High-intensity movement) and the within-subjects variable being the color mode (dark mode and light mode), as shown in Fig. 6. The measurement metrics and experimental procedure were consistent with those used in Experiment 1. The experiment was conducted using an Apple Watch 10, with the color mode set according to the design specifications of iOS.
Two color rendering modes.
Results and analysisCognitive Understanding of performance indicators
Users performed better in dark color mode compared to light color mode, and the color mode revealed no significant effect on users’ cognitive comprehension performance, as shown in Table 4. This may be due to the simplicity of the task, with the color mode having a minimal impact on users’ comprehension. Future experiments involving more complex tasks could further explore the significance of this effect. Users’ performance in strong motion scenarios was inferior to that in static and micro-motion scenarios. The motion scene had a significant effect on users’ cognitive comprehension performance, F(1,42) = 17.609, p
Subjective perception indicator
Color mode significantly affected users’ subjective perception indicators, including perceived ease of use (F(1,42) = 6.468, p = 0.015, partial η² = 0.133) and satisfaction (F(1,42) = 7.243, p = 0.01, partial η² = 0.147), with darker modes consistently scoring higher than lighter modes in both cases. However, no significant effect was observed on perceived usefulness. The impact of the dominant color mode on users’ subjective perception was not significant, likely because the difference between dark and light modes was relatively small, resulting in a reduced effect of different motion scenarios.