Behind the scenes of a year long research project: decision-making, limitations and learnings 

Olivia Johnson

We recently published our report “Working towards healthier information ecosystems: Collective visions from civil society in Latin America and the Caribbean,” as the first stage of our two year project focused on building healthier information ecosystems in Latin America and the Caribbean (LAC). 

Writing and engaging with communities in three languages across LAC presented some specific challenges! We thought we’d share some of our research methods, decisions and learnings from this nearly year-long period of researching and writing, including platforms, translation, interviews, engagement and analysis. 

We hope it will be useful for others working on similar research projects to learn from our practices and limitations, and please get in touch if you’d like to learn more about this process! 

Community calls: building new connections 

Between October 2023 and February 2024 we held three community calls. These calls were facilitated by project researchers, with some featuring panelists from interviewees. Over 60 people participated, with representatives from civil society organizations, philanthropic organizations, journalists, and academics in attendance. 

The goals of the community calls were to share our learnings, further explore key themes, build community with organizations and individuals working on these topics, and introduce our offer of free tech and data support for those interested. 

These calls broadened our networks and introduced us to new people and perspectives as the project progressed. To us, this was an important element of going beyond our immediate networks, especially for a project with such a broad scope. 

Interviews: hearing from both known and new (to us) voices

We wanted to speak to a variety of actors working throughout LAC to showcase some of the inspiring projects and work being done to restore information ecosystems in the region. 

A key component of this research, then, involved interviewing individuals and organizations working on these issues from a variety of perspectives: civil society, researchers, funders, activists, journalists and artists. 

Some of these people were individuals we already knew we wanted to chat with when the project started, and others we became familiar with through our community calls. 

Regular updates: sharing findings and learnings as the project progressed 

In order to create a more participatory and transparent process, it was important for us to share our findings and learnings at regular intervals throughout the course of the research. 

We did so through publishing blogs and holding community calls. Over the course of the research project, we published five blogs that shared our learnings, invited feedback from community members, recapped our community calls for those who weren’t able to attend them, and gave updates on our progress. 

Additionally, we published an article with Branch magazine, looking at climate focused initiatives that are working to restore information ecosystems (both with and without tech).You can catch up on these below: 

Analysis and storytelling  

We knew we did not want to write a solely diagnostic report. Rather, we wanted to focus on showcasing the incredible work being done across LAC in building healthier information ecosystems. In our report, we share some of the success stories that were talked about in our interviews. Integrating storytelling into the report required strong translation and detailed reading (and listening) of interview transcripts. We talk in more detail about this process in the next section.

Language and tech considerations 

Interviews + writing 

Most of our projects at TER are primarily conducted and written in English before being translated into other languages. 

From the start, this was a multilingual project with interviews conducted in Spanish and Portuguese, and research conducted in Spanish, Portuguese and English. Rather than placing the burden of translation on interviewees, we decided to translate all the transcripts into English before coding, analysis and writing. 

But translation introduces error, in terms of missing meaning, and it also slows down the research process. Likewise, the process of translating the report into Portuguese and Spanish after producing the initial English report added a few weeks onto the project timeline. 

In our translated reports, we tried to ensure that our translation was true to the meaning and word choices of the initial version, especially when it comes to gendered words and concepts that are not easily transferable between languages. We hired translators out of house, with two of our researchers reviewing the text for accuracy and consistency. When translating the reports in the primary languages of the interviews, we re-inserted the original quotes where possible. 

Analysis + coding 

For coding, we currently use Taguette, an open source qualitative research software. The fact that the tool is open source means that we can host it ourselves, which offers more control over the data; however, there are tradeoffs: the platform is rather simple and only organizes content with tags, not allowing for analysis on the platform. This meant we had to manually organize and analyze our transcripts, a process that contributed greatly to the writing process and the richness of the interview insights, but was also time consuming. 

Community calls 

The first community call was held in Spanish, with the second held in Spanish and Portuguese. After this call we realized we needed to include interpretation, to accommodate multiple languages. 

We held the first two calls on BigBlueButton, a self-hosted platform we use at TER for video calls. For our third call, we decided to switch to Zoom, so we could offer live interpretation between Spanish and Portuguese. Zoom also allowed for captions. 

For our launch event, we added English interpretation as well, since the report was initially published in English. Figuring out how to balance language access and accessibility needs with tech that is privacy respecting and open source is a constant decision-making process for us – when it comes to calls, we make the decision on which platform to use based on the specifics of the call, intended participants and needs such as language. 

Expanding language accessibility 

While we did what we could given staffing and time constraints, ensuring more equitable distribution of language access is an ongoing challenge. These constraints means that the project was not able to be representative of all 33 countries across LAC, nor the diversity of communities therein, including those who speak indigenous languages like Quechua, Guarani, Aymara; we also did not interview people who speak lesser spoken languages in the region such as French, Creole or English. 

What’s next!

We are currently working on the second phase of this project.  As our work continues to develop we are mindful of some of the learnings from phase 1, such as the importance of trying to meet language and accessibility needs, including underrepresented voices and responding to feedback we collected from our launch event. 

We will continue to post updates as this project progresses and share opportunities and resources on our website and social media, and through our newsletter 

Additionally, we are offering free tech and data support to individuals and organizations who are working to build healthier ecosystems in LAC. If you are interested in receiving support, reach out to our team here

MORE