OVERVIEW

Researchers’ data curation and sharing practices play a crucial role in ensuring research transparency and open science. Our team introduced NTUData, a prototype of a system that potentially allows a research team to internally manage, curate, and share data during the course of their research.

TEAM

Penny Chiang

Jian-Sin Lee

ROLE

Leader

User Experience Designer

User Researcher

DURATION

3 months

(May - July 2018)

* This is a research project sponsored by Ministry of Science and Technology, Taiwan, and led by Professor Wei Jeng at National Taiwan University, Department of Library and Information Science.

DESIGN TOOLS

Sketch

Flinto

BACKGROUND

As with promoting science reproducibility has reached consensus in many disciplines, research data curation and sharing becomes crucial and its benefit include improving research transparency, enabling data reuse, and teaching and learning purposes.

In this study, we use the term “research data infrastructure (RDI)” as an umbrella term to refer to the facilities, services, and tools that support researchers’ data curation-sharing-reuse process. However, there is still a lack of agreement regarding the features and design principles toward developing a better RDI system which can meet researchers’ needs.

To fill in gap, it is necessary to find out how a system could better support and encourage researchers to complete their data curation-sharing process.

Process

EXPLORATION

Literature Review

Since "Research Data Infrastructure" was a new term to me, at the beginning of the project, my goal was to understand the development trend of RDI, the need of researchers, and the suggestion of the system construction.

The main findings about users are:

Exemplar Collections

To know how the existing platforms function to support data curation, sharing, and reposition, I collected exemplars to find patterns and analyze the user experience.

Exemplar collections of existing tools.

From the collections, I concluded that the tasks of submitters include metadata* filling, file uploading, data management in terms of role assignment and embargo setting, and data use permission setting. But there are two drawbacks of the existing system:

*Metadata is data that describes other data. In the submission process, it requires users to fill in the metadata of the research data in order to support the data management and retrieval.

While most of the RDI systems require a submitter fill in detailed information, in contrast, it’s very simple and intuitive to upload, organize and share files in cloud storage platforms like Google Drive and Dropbox. I assumed this could be a good reference for the system's structure.

IDEATION

Early studies led us to two focus areas:

How to make the submission process simpler to save researchers' time?
How can the system help researchers better manage their research data?

Solutions

A project-based system

We introduce NTUData, which is a system that a Principal Investigator (PI) in the institution (National Taiwan University, Taiwan) is able to create a research project and manage all of the team’s data along the way toward the project completion.

Collaborate with team members

The PI of the team can also add other collaborators so that it would be convenient for them to record and communicate (e.g., teammates are able to comment on certain datasets).

Simplify the information

We suggest a clean interface that provides simple messages and clear directions on each page. To initially reduce the complexity, only the most important/used metadata would show in the submission process. Thus, users won't feel confused and struggle for figuring out what metadata to fill.

User Flow

Wireframes

Create Project - Project Metadata

Create Project - Member Settings

My Project - Initial Page

Create Project - Project Metadata

Create Project - Sharing Settings

Create Project - Preview

My Project - Multiple Project

My Data Page

Add Data - Upload Data

My Project - Success

Project Activity Log

EVALUATION

We design the NTUData not only as an institutional repository but also a collaborative platform for a research team to work together and check their data sharing performance actively. We manage to ensure the design in a functional and intuitive system.

To assess the usability of NTUData, we recruited nine participants, including four faculty members and five graduate students in National Taiwan University, Taiwan.

An overview of research participants

Usability Test Procedure

Results

Overall, feedback from participants in terms of the interfaces is positive. Eight of the nine participants consider NTUData user-friendly and easy to learn. Suggestions for improvement were centered in two aspects:

several labels in the system are confusing (e.g., jargons that are used often in library fields)
the interoperability by linking NTUData with external resources (e.g., the project management system of the Ministry of Science and Technology or the University human resource system) so as to reduce the burden of repeatedly typing duplicate fields when creating a new project.

Moreover, two characteristics were both considered as the highest priority of an ideal data curation system in the survey:

Moving forward, we need to create more interfaces for other components of NTUData and make the whole system more robust by conducting interviews with different stakeholders (e.g., staff members of the library or the office of research and development), hoping to turn the prototype into a real product that is more adaptive to the local users' needs in the future.