RTI's Commitment to Open Science

RTI is committed to sharing research and the advancement of equitable science. We celebrate the Year of Open Science by highlighting opportunities and collaborations that demonstrate these commitments.

Making research available to all

  • RTI encourages research staff to share knowledge through Open Access channels, such as Open Access journals, RTI Press, and preprint repositories, whenever possible. Open Access is consistent with RTI’s mission and intentions toward more purposeful equity and inclusion. RTI supports Open Access as a means of making science more accessible, by increasing the discoverability, reach, and impact of RTI’s research.
  • Since 2007, RTI Press has shared research through peer-reviewed, Open Access publications. We publish on a broad range of topics reflecting RTI’s multidisciplinary research, expertise in social and laboratory sciences, and extensive international activities. Our publications inform national and international research, scientific discussions, and policy debates. RTI International supports RTI Press as a means of sharing research and practical knowledge to improve the human condition.
  • RTI data scientists have created a range of open source tools that make it easier to work with large data sets and address the needs of our partners. Examples of these tools - such as gobbli, SMART, and Harness-Vue - are detailed in the Open Source Tools section below.
  • RTI's subject matter experts work with partners to make research widely accessible. For example, our natural resource economists have collaborated with the Environmental Protection Agency's Climate Change Division, Climate Economic Branch to support open science around forestry, agriculture, and land use modeling. This work has resulted in numerous Open Access publications aimed at analyzing the sustainable use of natural resources to achieve climate mitigation policies.

Respecting diverse cultures & promoting equity

  • RTI is committed to being at the forefront of the call to action on racial justice and equity. To ensure focused and sustained attention, we have established the Transformative Research Unit for Equity (TRUE). TRUE is turning practice into impact through transformative research that leverages power, influence, and resources to advance equity. TRUE researchers have made their research open through RTI Press publications.
  • With support from the National Institutes of Health (NIH), the All of Us Researcher Academy is designed to provide training, technical assistance and peer-to-peer learning for health researchers at Minority Serving Institutions, including Historically Black Colleges and Universities (HBCUs).
  • RTI researchers work to respect diverse cultures throughout the research life cycle, for example, by reflecting cultural norms of communication in survey research

Maintaining security and privacy

  • RTI is invested in promoting FAIR data principles to meet the needs and challenges for modernizing and securing data systems.
  • In line with the National Science and Technology Council's guidance, RTI is integrating the use of digital persistent identifiers throughout the research life cycle. In particular, RTI has successfully adopted and integrated ORCID iDs, which are researcher-level PIDS, to link scientists and their accomplishments.

Fostering collaborations

RTI researchers collaborate with a range of clients on efforts to promote open science, a number of which are highlighted in the gallery below.

  • In 2023, NIH adopted a new policy to promote the sharing of scientific data. As part of the NIH HEAL Data Stewardship Group, RTI ensures that NIH HEAL data are Findable, Accessible, Interoperable, and Reusable (FAIR) by developing processes for data harmonization, data storage, metadata collection, and sharing within a federated data ecosystem. 
  • Since 2010, improving early grade reading has been a top priority for RTI and for many bilateral and multilateral donors, including the U.S. Agency for International Development (USAID), The World Bank, UK Aid and others. In support of this work and to improve upon paper forms of assessment, RTI developed a mobile assessment and coaching tool we dubbed Tangerine. Open source for all users, Tangerine is a first-of-its-kind software application optimized for offline data collection on low-cost Android tablets. Originally designed to record data from early grade reading and mathematics assessments and to enhance coaching feedback for teachers, students, and school administrators, the Tangerine app has been adopted in other sectors, such as health.

Ensuring reproducibility

  • RTI has adopted DMPTool for researchers to prepare data management and sharing plans that meet funders' expectations and facilitate open data.
  • RTI Fellows and our Scientific Stature Services team are advancing Open Science at RTI through a series of roundtables to address topics especially relevant to our scientists. The purpose is to introduce these topics and prompt researchers to explore the relevant details for their own fields and funders. In 2022-23, topics have included protocol papers, data management and sharing plans, and Open Access publishing.
  • RTI researchers embrace reproducible research as good stewards of research funding and as providers of credible information for policy decision-making across many areas of public concern.

Open Source Tools

A researcher works on coding data at her computer.

SMART, funded in part by a National Consortium for Data Science 2017-2018 Data Fellows grant, leverages active learning, gamification, and a focus on the user experience to help teams reduce time and effort spent coding or labeling data manually. SMART makes labeling data easy by simplifying and allowing monitoring of the labeling process with metrics and visualizations on coder agreement and performance. The project is open-source and hosted on GitHub, has a diverse community of international users, and is one of the top open-source data annotation platforms for text classification on GitHub.

SMART: An Open Source Data Labeling Platform for Supervised Learning

Gobbli is an open-source Python library hosted on GitHub that bridges state-of-the-art research in Natural Language Processing and real-world application to real-world problems and data. Beyond providing a consistent interface to state-of-the-art models, gobbli provides supplementary tools inspired by the types of problems and datasets commonly faced when applying natural language processing to social science and survey research. Gobbli also provides an interface for data augmentation—a powerful tool for generating synthetic text data to aid in model development when there are not many examples to learn from.

Unified Framework Brings Fresh Approach to Text Classification

social media analysis

PushshiftRedditDistiller is an open-source tool hosted on GitHub that makes downloading, organizing, and filtering social media data from the Pushshift Reddit archive much easier. For projects that collect data from reddit, we needed a tool able to run through the full reddit archive dating from 2005 (1TB+) quickly to identify relevant content. This package makes it easier for users to track which data sets they have already downloaded, thereby eliminating unnecessary downloads of large files.

Custom dashboards are a frequent need in federal agencies. Harness-Vue, an RTI-written plugin for the popular open-source vue.js javascript framework, abstracts best practice frontend data architecture into a simple API that drastically reduces the development and maintenance time in writing a custom javascript dashboard. The Harness-Vue suite of software also includes a number of reusable UI components and layout tools for quickly templating a responsive, mobile-friendly and Section 508-compliant dashboard, as well as automated testing, continuous integration and continuous deployment.

Criminal justice research often requires conversion of free-text offense descriptions into overall charge categories to aid in state-to-state comparisons. To address this need, RTI created Rapid Offense Text Auto-Coder (ROTA), a machine learning model accessible in an open source web application on GitHub. ROTA allows for the bulk upload and conversion of offense text into offense codes following the National Corrections Reporting Program (NCRP) coding scheme. This results in the reduction of an hours-long coding task to minutes with an overall accuracy of 93%.

Machine Learning Tool Brings Efficiencies to Criminal Justice Research Practices


Open Science: By the numbers

RTI has made progress towards our Open Science goals with substantial number of Open Access publications, broad uptake of ORCID iDs, and through projects funded by many of the federal agencies participating in the Year of Open Science.

0 +

Number of Open Access publications

0 %

Adoption of ORCID iDs by RTI researchers

0 +

RTI-led projects funded by select agencies

$0 B

Funding for RTI-led projects from select agencies

Featured Stories