Source-linked AI summary
Usability Comparison of Mouse, Touch and Tangible Inputs for 3D Data Manipulation
Lonni Besançon, Paul Issartel, Mehdi Ammi, Tobias Isenberg
TL;DR
The paper addresses limited comparative evidence on mouse, touch, and tangible input for 3D manipulation. It compares the modalities in a docking task, finding similar precision but different completion times, with tangible fastest, touch slower, and mouse slowest.
Problem
Method
The authors compared the three modalities in a 3D docking study with repeated trials and measures of precision, time, fatigue, workload, and preference.
Source: §1 · p. 1, §3.1 · p. 2Results
Takeaways & Limitations
The findings provide insights for further development and evaluation of these techniques, particularly for 3D visualization.
Source: §1 · p. 1The shared setup constrained the study, and for more complex practical interactions mouse and touch are likely better suited than tangible input.
Source: §6.7 · p. 8
Abstract
from arXiv · showhide
We evaluate the performance and usability of mouse-based, touch-based, and tangible interaction for manipulating objects in a 3D virtual environment. This comparison is a step toward a better understanding of the limitations and benefits of these existing interaction techniques, with the ultimate goal of facilitating the integration of different 3D data exploration environments into a single interaction continuum. For this purpose we analyze participants' performance in 3D manipulation using a docking task. We measured completion times, docking precision, as well as subjective criteria such as fatigue, workload, and preference. Our results show that the three input modalities provide similar levels of precision but require different interaction times. We also discuss our qualitative observations as well as people's preferences and put our findings into context of the practical application domain of 3D data analysis environments.
1 INTRODUCTION
The paper compares mouse, touch, and tangible interaction for 3D manipulation as a step toward an interaction continuum for 3D data analysis. In a docking study, the modalities achieved similar precision but differed in completion time and preference.
1 INTRODUCTION
- The study examines how mouse, touch, and tangible inputs can support 3D interaction, with the longer-term aim of enabling transitions between techniques as tasks change.The authors measured precision, fatigue, workload, preferences, and general feedback across 15 docking tasks per modality.Source: §1 · p. 1
- All three modalities achieved similar precision, but differed in task completion times and participant preferences.The study also recorded qualitative observations of how participants used each technique.Source: §1 · p. 1
- The authors present their findings as insights into the techniques’ advantages and disadvantages and as a basis for further development and evaluation, particularly for 3D visualization.Their contributions include analysis of users’ understanding and use of the modalities and a study design that does not focus on a single evaluation parameter.Source: §1 · p. 1
2 RELATED WORK
Prior work rarely compared mouse, touch, and tangible inputs for full 3D manipulation while considering multiple measures of performance and usability. This study builds on earlier comparisons but extends them to true 3D tasks and a broader assessment of interaction.
2 RELATED WORK
- Few studies had analyzed mouse, touch, and tangible modalities for 3D manipulation, leaving limited comparative evidence for these tasks.Earlier research more often compared novel techniques with established ones or studied pairs of modalities, often in 2D tasks.Source: §2 · p. 1
- Earlier 3D comparisons often focused narrowly on rotation, while tabletop matching constrained interaction to two dimensions and other work assessed only selected parameters.The authors aim for a more holistic view of how modalities affect interaction with 3D shapes or scenes.Source: §2 · pp. 1–2
- The paper focuses on direct interaction with displayed 3D information rather than remote manipulation through separate displays.The authors describe touch as occurring on the displayed data and tangible input as manipulating a physical object where the virtual shape is thought to be.Source: §2 · p. 2
- The study extends Tuddenham et al.’s comparison of mouse, touch, and tangible input by using true 3D manipulation tasks, including for tangible input.The earlier matching task was constrained to two dimensions.Source: §2 · pp. 1–2
3 COMPARATIVE STUDY
The comparative study evaluates three input modalities using a 3D docking task and several usability measures. Its design covers task setup, apparatus and mappings, participant procedure, measured variables, and hypotheses.
3 COMPARATIVE STUDY
- The study assessed effectiveness, efficiency, satisfaction, and ease of learning through precision, completion time, subjective ratings, and changes in completion time; error tolerance was out of scope.The authors used angular and Euclidean distance as two measures of accuracy.Source: §3 · p. 2, §3.6 · p. 4
3.1 Task
- The docking task required translating and reorienting a 3D object, then positioning it precisely to match a wireframe target.Participants completed 15 trials per modality using the same pool of valid start and target positions, for 45 trials total; modality order was counterbalanced.Source: §3.1 · p. 2
3.2 Apparatus
- All conditions used the same touch-enabled screen; the study compared a computer mouse, screen-based capacitive touch, and an optically tracked handheld cuboctahedron.The screen was slightly tilted to provide a comfortable touch setting.Source: §3.2 · p. 3
3.3 Interaction Mappings
- The mappings used mouse-and-keyboard controls, touch gestures, and one-to-one tangible motion to manipulate the virtual object.For touch, the authors chose a finger-based mapping that included rotation, scaling, and translation; the tangible mapping followed the physical object’s motions.Source: §3.3 · p. 3
3.4 Participants
- The study included 36 unpaid participants aged 19–52, with varied self-reported experience in computer-based 3D manipulation.Twelve participants described themselves as skilled and 24 reported no significant prior experience.Source: §3.4 · p. 4
3.5 Procedure
- Participants received minimal instructions for each modality, completed questionnaires on demographics, fatigue, workload, and preferences, and could select a favorite technique.The study also recorded observations of participants’ use of the three modalities.Source: §3.5 · p. 4
3.6 Variables
- The independent variable was interaction modality; dependent variables were completion time, precision, fatigue, workload, and preference.Precision included Euclidean distance to the target and rotational difference.Source: §3.6 · p. 4
3.7 Hypotheses
- The hypotheses predicted faster times for tangible than touch and for touch than mouse, different precision and workload across modalities, greater tangible fatigue, and preferences for touch and tangible over mouse.The proposed precision ordering favored mouse over touch and touch over tangible; the predicted preference reasons included touch’s mappings and tangible input’s similarity to real-world interaction.Source: §3.7 · p. 4
4 RESULTS
Across mouse, touch, and tangible input, task completion times differed while precision was similar. Workload, fatigue, and preferences also varied across the measures reported.
4 RESULTS
- The study collected 1,620 docking trials from 36 participants, measuring completion time and accuracy for each input modality.The analysis reported effect sizes and confidence intervals rather than p-value statistics.Source: §4 · p. 4
4.1 Task Completion Time
- Tangible input completed the docking task in 26 s, versus 47 s for touch and 61 s for mouse; touch was also faster than mouse.Tangible was more than twice as fast as mouse and almost twice as fast as touch; learning appeared for mouse and touch, but not tangible.Source: §4.1 · p. 5, Figure 5 · p. 5
4.2 Precision
- Precision was similar across modalities: Euclidean errors averaged 5 mm for mouse and tangible and 6 mm for touch, while rotational errors were 3.4° for mouse and 3.7° for touch and tangible.The authors found no evidence of a large precision difference, including when analyzing only later trials.Source: §4.2 · p. 5
4.3 Measuring Workload
- Tangible input required lower overall workload than touch, while other overall workload comparisons were less clear.Sub-aspect differences were mostly gradual; tangible had lower frustration and effort ratings, while temporal-demand ratings tracked the timing difference between tangible and mouse.Source: Figure 8 · p. 5, §4.3 · p. 6, Figure 9 · p. 6
4.4 Measuring Fatigue
- Mean fatigue ratings were below 4 on the 0–10 scale, with no evidence of an important difference between conditions.Source: §4.4 · p. 6, Figure 10 · p. 6
4.5 Measuring Preferences
- Tangible was the most frequently stated preference (24 participants) and was chosen by a majority for another 15 trials (22 participants).Among the 16 participants willing to repeat the tasks, 11 chose tangible, 4 mouse, and 1 touch.Source: §4.5 · p. 6, Figure 11 · p. 6
5 QUALITATIVE OBSERVATIONS
Qualitative observations documented how participants used touch and tangible input during the docking task, including difficulties and varied interaction strategies.
5.1 Touch
- Twenty participants had difficulty with two-finger RST, often combining rotation and scaling when they intended to change only one degree of freedom.Eleven participants said they preferred translating along the x- and y-axes without changing distance or z-axis orientation.Source: §5.1 · pp. 6–7
- Participants used different hand configurations for pinch input, and some reported accidental transformations when lifting their fingers from the screen.Twenty used fingers from different hands and 16 used two fingers on one hand; three reported this unintended transformation on finger removal, called exit error in the paper.Source: §5.1 · p. 7
5.2 Tangible
- In the tangible condition, 12 participants docked objects through sequential translations and rotation rather than integrated manipulation.For rotations over 90°, 29 participants used two hands and seven used one; the authors were uncertain why some did not use integrated interaction.Source: §5.2 · p. 7
6 DISCUSSION
The discussion reports faster tangible interaction and similar precision across modalities, while workload, fatigue, preference, and realistic application requirements complicate choosing one universally preferred input.
6.1 Completion Time
- Tangible interaction was fastest, touch slower, and mouse slowest; the authors link this pattern partly to integrated DOF control and fewer mode switches.They conjecture that tangible interaction’s faster approximate docking left more time for fine-tuning.Source: §6.1 · p. 7
6.2 Precision
- The study found no evidence that any modality provided higher measured precision, although participants’ perceived control differed across mouse, touch, and tangible input.Participants reported precise control 22 times for mouse and 8 times for touch, while some described involuntary movements with tangible input.Source: §6.2 · p. 7
6.3 Workload
- Workload tended lower for tangible than touch, then mouse, but raised arms and touch-mapping sensitivity complicated the comparison.Fast tangible movements exceeded the tracking system’s capacity, and camera occlusion interrupted tracking, frustrating participants.Source: §6.3 · p. 7
6.4 Fatigue
- Fatigue ratings were similar across techniques, but the shared setup and unreleased tangible object could affect fatigue in other settings.Participants reported difficulty keeping the virtual object positioned after releasing the tangible; the authors conjecture extended use could increase fatigue without an exit-error-free release option.Source: §6.4 · pp. 7–8
6.5 Preferences
- Participants overwhelmingly preferred tangible interaction, but novelty likely biased that preference; at least 5 of 11 tangible-preferring participants cited the chance to try unfamiliar technology.The authors caution that participants were already familiar with mouse and touch interaction.Source: §6.5 · p. 8
6.6 Realistic Application Scenarios
- Realistic applications may involve longer, more complex tasks, more than 6 DOF, and expert users, potentially favoring mouse or touch for generic interaction.The authors suggest tangible input may suit specialized tasks, while touch may suit integrated approaches.Source: §6.6 · p. 8
6.7 Summary of Limitations
- The shared setup, participant population, tangible novelty, learning effects, and chosen touch mapping constrain how the findings transfer to practice.The authors note that dedicated setups could yield better results and that other applications may require different mappings.Source: §6.7 · p. 8
7 CONCLUSION
The conclusion finds all three modalities equally suited to precise positioning, while tangible interaction was fastest. It argues that each modality has advantages and disadvantages rather than one clear favorite.
7 CONCLUSION
- All three modalities were equally well suited for precise 3D positioning, while task completion time was fastest for tangible, slower for touch, and slowest for mouse.The tangible speed advantage remained when examining only the last third of trials, despite observed learning effects.Source: §7 · p. 8
- The authors conclude that no single modality is a clear favorite for 3D data exploration; each has advantages and disadvantages to consider.They identify further study of touch mapping, exit errors, DOF control, and hand use, with the broader aim of task-dependent transitions between interaction scenarios.Source: §7 · p. 8