An ideal enhancement for future datasets is the provision of options for data visualization. This significantly eases the process of understanding and analyzing motion data, making datasets more accessible and user-friendly. To aid in this, we publish the code for our motion visualization tool and provide instructions for how to set it up.

Beyond the discussed fundamental specifications, there is additional contextual information that can greatly benefit researchers. For example, providing background information about the used data source, such as whether it was collected from Unity, Steam OpenVR, or other platforms, can give important context information about the dataset’s characteristics. Moreover, awareness of application-specific traits is crucial. For instance, scenarios where users are teleported within a scene can result in abrupt and seemingly inexplicable ‘jumps’ in the data. Similarly, it should be clarified if data for certain peripherals, like hand tracking, are available only under specific conditions (like when hands are visible to the camera). This helps to distinguish between intentional data absences and potential errors. Additionally, understanding whether users might place their controllers or HMDs down during a session, or whether they are seated or standing, can offer valuable insights into the dataset’s dynamics. These nuances, although seemingly minor, can have substantial implications for the accuracy and reliability of research outcomes, underscoring the importance of comprehensive dataset documentation. Creators of dataset should also be aware that their datasets might be used for different purposes, so they should not just focus on the specific requirements of their own research.

We highly recommend that dataset creators disclose and discuss the software used in generating their datasets. This transparency not only allows for the reproduction, verification, and extension of the datasets but also fosters an environment of open collaboration and innovation in the field. By sharing the tools and methods used for data collection, researchers can contribute to a more robust and dynamic understanding of kinematic data in XR environments.

Researchers must document any conversions applied to the motion data. Detailed documentation of these conversions is essential for clarity and integrity of the research. It ensures that each step of data handling is accurately conveyed and that the data is not inadvertently distorted during the process. This transparency is critical for other researchers who may wish to replicate or build upon the work.

In studies utilizing multiple datasets, it is essential to disclose the key differences between them and the measures taken to align each dataset. This includes aligning coordinate systems, normalizing units of measurement, standardizing rotation representations and frame timing. Researchers must clearly outline how they harmonized disparate datasets to ensure consistent and accurate analysis.

Authors must critically analyze and potentially discuss how the specifications of the datasets could have influenced their results. This examination should consider whether the results might be skewed — either overly optimistic, such as when machine learning models overfit to dataset-specific signals, or overly pessimistic, due to erroneous preprocessing or misinterpretation of the data. Such a critical evaluation helps in contextualizing the findings and provides a more nuanced understanding of the study’s implications and limitations.