Consultancy Service Agreement Terms of Reference – Social media listening – Mixed Migration Centre LAC At Danish Refugee Council

1. Background of the organization

The Danish Refugee Council (DRC) is one of the world’s leading humanitarian NGOs, working in over 40 countries to provide rights-based assistance to refugees, asylum seekers, internally displaced persons, returnees and migrants. Since 2019, DRC has been implementing a humanitarian programme responding to the emergencies of Venezuelan refugees and migrants in the Latin American region and victims of the internal conflict in Colombia.

The Mixed Migration Centre (MMC) is a global research network focused on data collection, analysis and policy development on mixed migration, with regional centres in Africa, Asia, Europe and Latin America. The MMC aims to improve understanding of mixed migration, positively influence migration policies and provide evidence-based responses for people on the move. As part of the Danish Refugee Council (DRC), the MMC benefits from operational support, while maintaining its independence as a source of data, research and policy development for policy makers, practitioners, journalists and the humanitarian sector.

1. Consulting context

Despite the efforts of governments in Latin America and the Caribbean to prioritize the integration of migrants, socioeconomic barriers persist that limit their integration, such as precarious working conditions, legal obstacles and limited access to basic services. In addition, in recent years, the region has experienced increasing xenophobia, which leads to a deterioration of public opinion on migration in the region. This creates the need for better policy responses. To do so, it is important to understand the perception of the host population towards migrants, the perception of the migrant population regarding their experience in the host countries, their integration and the challenges they face and the interactions between both communities.

Since social media is a space where many of these opinions are expressed, it has become an increasingly common source of data to analyze perceptions and feelings around migration and migration policies. Social media is a space for spontaneous expression, which makes it relevant to identify the opinions and feelings of a large number of people in a real and unprecedented way. At the same time, the migrant population uses social media as a means of information.

Uruguay and Argentina represent two contrasting contexts in terms of migration policies and integration programs, which makes them an extremely interesting comparison. While Uruguay has stood out for maintaining an open and receptive approach to migration, promoting inclusive policies and facilitating the integration of migrants , Argentina has adopted a more restrictive approach in recent years, with policies that have limited the entry and stay of foreigners. These differences allow us to analyze how the different regulatory frameworks impact the lives of migrants and their opportunities for integration in both countries.

To this end, MMC LAC aims to analyze the discourse on social media about the reception and integration of the migrant population in Uruguay and Argentina, considering the perspectives of both the host population and the migrant population. Through this analysis, we seek to understand:

• The perception of the host population towards migrants, identifying positive and negative opinions;
• The perception of the migrant population regarding their experience in host countries, their integration and the challenges they face;
• The interactions between both communities, evaluating whether they operate in separate or interconnected spaces;
• The implications of these findings for public policy formulation.

1. General objective of the consultancy.

The purpose of this consultancy is to carry out an analysis of the dynamics of reception and integration of the migrant population in Argentina and Uruguay through the analysis of discourses on social networks.

The analysis will be based on data collected from three Meta social networks (Facebook, Instagram and Threads) and will be carried out by implementing advanced Social Media Listening tools that facilitate the study objective, with the support of the MMC LAC team.

1. Specific objectives of the consultancy

• Developing the research methodology: MMC LAC will provide the consultant with a manual of previous research, along with keyword suggestions and a list of recommended pages, groups, accounts and events. Based on this information, the consultant will develop a methodological proposal (methodological sheet), which will detail the selection of social networks, inclusion/exclusion criteria, keywords, pages/groups/accounts/events selected, variables to be extracted and data collection times. This proposal must be approved by the MMC LAC team before its implementation.
• Implement data extraction and analysis processes: Implement data extraction processes in public groups, pages and events (in the case of Facebook) and public accounts (in the case of Instagram and Threads), using advanced methodologies such as webscraping or the use of APIs to capture relevant posts, comments, interactions and metadata, which will then be organized in a database to ensure the integrity and quality of the information. Any adjustments to the collection methods must be communicated to and approved by the MMC LAC team.
• Qualitative data analysis: Conduct a qualitative analysis using methodologies such as netnography, audience analysis, advanced natural language processing (NLP) and clustering. This will include identifying patterns, trends and relationships between the variables defined in the methodology. The results will be presented in a detailed report of the main findings.
• Collaboration in the analysis: Accompany the Information Management (IM) team in the use and replication of the tools and methodologies developed during the data collection and analysis stage, ensuring the transfer of knowledge and skills necessary for the continuity of the project.
• Prepare technical document: Develop and deliver a technical document that includes a detailed description of the implemented processes, the theoretical basis of the methodologies used, a glossary of variables with categorization, and code developed in Python.
• Prepare results report : Consolidate the main findings of the analysis of the data collected in a report, answering the main research questions presented by MMC LAC.

1. Research Questions
2. How do the attitudes of the host population towards migrants manifest themselves on social networks in Argentina and Uruguay?

• What types of discourse predominate (positive, negative or neutral) and what are the most frequent emotions (acceptance, rejection, fear, empathy)?
• What are the issues that receive the most attention on social media in relation to migrants (such as security, health, education, employment and housing) and how are the opinions of the host population articulated regarding these issues?
• How have these attitudes changed over time or in response to certain events or immigration policies?

1. What main topics do migrants discuss in relation to their integration process in host countries?

• What types of discourse predominate (positive, negative or neutral) and what are the most frequent emotions (satisfaction, discontent, fear)?
• What problems do you highlight regarding access to basic services such as health, education, employment and housing?
• How do migrants perceive the way they are received by the host community in Uruguay and Argentina, and what role does xenophobia play in this perception?
• How have these perceptions changed over time or in response to events or public policies related to migration?

1. How do host communities and migrants interact on social media?

• Are there shared spaces for interaction or are they segregated into separate communities?
• What type of content generates the most interaction between the migrant and host population on social media (e.g. news, personal experiences, migration policies)?
• What emotions predominate in interactions between both groups (for example, empathy, conflict, indifference)?
• Are there influencers or public figures who facilitate or hinder dialogue between migrants and locals (e.g. by promoting xenophobia)?
• What are the factors that facilitate or inhibit interaction between migrants and the host population on social networks?

1. What differences and similarities exist in perceptions of migrant integration between Argentina and Uruguay, according to social network analysis?
2. What implications do perceptions collected on social networks have for the development of public policies that promote the integration of migrants?

• What public policy proposals can emerge from the analysis of these interactions?

1. Research methodology: Collection of data from social networks.

Methodologically, the study will build on a previous study conducted by MMC LAC in collaboration with Disorlab , which used data from Facebook groups. For the present study, three social networks from Meta (Facebook, Instagram, and Threads) will be explored. The previous study underlined the relevance of Facebook for analyses focused on migration and social dynamics, justifying its inclusion in this research. In order to broaden the scope of the previous analysis, additional data from other social networks belonging to Meta, such as Instagram and Threads, will be incorporated. Instagram is selected for its high popularity, particularly among migrant communities. However, it is important to recognize that the use of each social network may vary by country, suggesting the need for a flexible and contextualized approach to data extraction and analysis on each platform.

Data extraction will be carried out in public groups, pages and events (in the case of Facebook) and public accounts (in the case of Instagram and Threads) which will be identified by tracking keywords related to the integration of migrants, events or public policies related to migration and the nationalities most present in the country, for example: migrants in Uruguay/Argentina, Venezuelans in Argentina, Residence for Roots , etc. Publications will be analyzed as well as comments. Selection criteria should consider the level of activity of the group – measured by the frequency of daily posts, as indicated by the search engine – and the presence of messages containing information related to the integration process in the country. Collection and extraction must adhere to the ethical guidelines established by MMC.

Deliverables:

1. Methodological sheet: Detailed summary of the selection of social networks, keywords, groups chosen, variables to be extracted and the time required. This sheet should have a maximum of 3-5 pages and include the tools that will be used for the analysis.

2. Properly structured databases with publications, comments, interactions and metadata extracted from the selected social networks, in accordance with the criteria approved in the methodological proposal by the MMC LAC team. Including a glossary of variables with categorization/clusters of the data and a performance score for each cluster.

3. Detailed technical report: Document describing the complete data extraction process on the selected social networks using web scraping or APIs, together with a description of the use of other tools used during the data analysis stage (15 pages max).

4. Python code: Code used for data extraction and analysis, commented to facilitate replication and adaptation to future projects.

5. Support for the Information Management team: Data collection and analysis session for the MMC-LAC IM team, focused on the practical use of the AI ​​tools and methodologies developed. The proposed training sessions should prioritize practical learning over theoretical learning, with real examples, practical exercises, and the direct implementation of the tools and methodologies on the data collected to answer the project’s research questions. The consultant may propose in the technical proposal the number of sessions and the workload that he/she considers appropriate, according to his/her data collection and analysis plan.

6. Main findings report: Specific summary of the main findings derived from the data analysis, presenting the most relevant results (approx. 20 pages).

Timeline

The project will be implemented between the second half of October and December 2024, with a specific schedule to be agreed upon jointly with the selected consultant or consulting team, to ensure the delivery of all expected products within this period.

The MMC LAC team provides this indicative schedule, for your consideration:

• Methodology (second half of October 2024): MMC LAC will provide the consultant with a manual of prior research, suggested keywords, and a list of recommended pages, groups, and accounts. With this information, the consultant will prepare a methodological proposal (methodological sheet) detailing the selection of social networks, inclusion/exclusion criteria, keywords, selected pages/groups, variables to be extracted, and collection times. The proposal must be approved by MMC LAC before its implementation.
  • Data collection (November 2024): The consultant will extract data from Meta’s social media following approved criteria. He/she will use specialized tools to capture posts, comments, interactions, and metadata, structuring the data in a database for further analysis, ensuring the quality of the information. Any changes to the method must be approved by MMC LAC. The consultant will organize training sessions on data collection with the MMC LAC IM team.

Qualitative data analysis (November – December 2024): Once the database is complete, the consultant will perform the data analysis

• identifying patterns, trends and relationships between variables. The consultant will organize training sessions on the analysis of the data collected with the MMC LAC IM team.
• The results of the analysis will be presented in a report, consolidated by the consultant, which must be delivered no later than the end of December.

Consultant’s experience:

The consultant or consulting team must meet the following requirements:

• Background in computer science, engineering, social sciences, or related fields, with specialization in data analysis, natural language processing, or digital research.
• Experience in data extraction from social networks, using web scraping techniques, use of APIs, Social Media Listening, and qualitative analysis.
• Knowledge of netnography, audience analysis, NLP and clustering, applied to migration and human rights.
• Advanced Python programming skills for data extraction and analysis automation.
• Ability to document and train in technical and methodological processes.
• Experience in Artificial Intelligence and Machine Learning for social media data analysis and NLP.
• Previous experience in humanitarian or migration projects will be valued.

How to apply

Procedure for proposals:

Interested consultants must submit the following items/documents (each document clearly identifiable):

1. Resume of the person or team of consultants.
2. Technical proposal documenting how the consultancy requirements, phases and deliverables would be met.
3. Financial offer in COP, MXN or USD including a cost for each of the six deliverables. The financial offer may not exceed 13,000 USD.
4. A detailed schedule with the delivery dates of the 6 deliverables.
5. Employment references (3).

Evaluation of the proposals received.

After the administrative evaluation to validate that all required documents have been submitted by each candidate, the evaluation of the proposals received will be based on the following scoring system:

• Resume of the person(s) – maximum 10 points.
• A technical proposal, documenting how the consultant would meet the requirements for the consultancy – maximum 10 points.
• Financial offer in USD/COP/MXN including a cost for each of the six deliverables – maximum 10 points .
• Detailed schedule with delivery dates for the 6 deliverables – maximum 10 points.

Maximum points obtainable: 40 points.

Application deadline: 09/29/2024 at 11:59 PM – Colombia Time. Applications must be made through the DRC system, applications received by email will not be considered.

For further information on this consultancy, please contact Simon Tomasi simon.tomasi@mixedmigration.org