Dr Prithvi Hirani and Robert Trigwell
Please find the full version of the study on the IOM website
The expansion of digital technology and activities in recent years has led to an exponential growth in forced migration data. However, this increase in available data does not always result in more inclusivity nor representation of marginalised individuals/groups. Collecting data in the forced migration field often relies on focus group discussions and key informant interviews. While such methods have proven to be effective, they are susceptible to reproducing existing power-dynamics and systemic biases which undermine diversity, representation, and inclusion. Often such ways to collect data represent the perspectives of those who already have a voice such as community and/or religious leaders, resulting in the inclusion of displaced populations with similar profiles and low participation of demographics like youth or women. Academic researchers and practitioners have pointed towards the limited participation of affected populations as data sources. Data collection practices which target forcibly displaced populations have been criticised for being ‘performative’, unidirectional and extractive. Poor representation within datasets can lead to biased analysis and may detrimentally impact decision-making. At the humanitarian sector level, the Grand Bargain in 2016 promoted localisation, participation, and inclusion as key priorities, however these priorities have mostly remained abstract with limited progress towards implementation. Even though there is momentum towards evidence-based decision-making, issues around power, representation and bias in data must be identified and tackled.
In this blog post, we report on a pilot project by the International Organization for Migration (IOM) leveraging Collective Intelligence to achieve greater representation and diversity in data collection. This pilot project was conducted in Iraq in May 2021. The pilot sought to test ways in which the knowledge, voices and participation of affected populations can be better included through all stages of the data cycle (collection, validation and analysis).
This blog piece illustrates how affected populations can be successfully involved in processes around analysis and validation of data, in a way that is complementary to conventional data collection.
Geo-located results were presented to the affected populations via an interactive platform based on their location, to allow them to vet, validate and comment on the results. Quantitative data and qualitative feedback were used in the next iteration of the analysis. This piece contributes to two ongoing discussions in the forced migration field. First, it explores the challenges of existing data collection tools related to power and voice. Second, it contributes to the growing discussion around improving accountability to affected populations and advancing localisation using new and innovative processes for data collection in forced displacement contexts.
Applying Collective Intelligence and the ‘Have Your Say’ platform
Collective Intelligence combines data, people, and technology. It can be understood as the enhanced capacity of groups, using technology, to mobilise a wider range of information, ideas and insights which can then inform innovation, learning or decision-making. Some well-known examples of Collective Intelligence are Wikipedia and Waze, a navigation app which crowdsources information on traffic to inform other drivers. In the humanitarian sector, the Ushaidi platform shows how Collective Intelligence is used to gather ‘wisdom of the crowds’ in a crisis context by focusing on implementing a crowdsourced public response to crucial decision-making processes. This new branch of Collective Intelligence has been gaining more traction among the COVID-19 response mechanisms.
IOM’s Return Index is a key interview-based tool designed to measure the severity of conditions in locations of return across Iraq. The pilot study sought to engage a broad subset of the affected population in Iraq to explore whether and why their assessment of their neighbourhood or village differed from that of the interviews, as captured in the Iraq Return Index. Collective Intelligence was employed in tandem with IOM’s existing Iraq Return Index to test and validate the findings. The online ‘Have Your Say’ Collective Intelligence platform was built by selecting five out of the sixteen Return Index indicators for validation by the community. They covered the main thematic areas of housing destruction, employment, business recovery, community reconciliation and security which reflect conditions of returns and could be easily cross-referenced. Given the high level of mobile phone and Internet use in Iraq, it is a fertile context for a Collective Intelligence pilot. The survey was disseminated for a seven-week period in English, Arabic and Kurdish via the existing social network of the IOM Iraq Facebook page, with over 93,038 followers. Of these, 25 per cent of users were female and 51 per cent were between the ages of 25-35, a demographic less traditionally surveyed in key informant interviews or focus group discussions. The quantitative analysis tested the indicator results. For the qualitative data, an open feedback platform allowed participants to express their sentiments.
Making room for disagreement: voices of affected populations
While most quantitative findings from the Collective Intelligence platform supported the Iraq Return Index, this section will focus on three out of the six indicators where there was notable disagreement between the traditional interview method and the Collective Intelligence platform. Most participant answers focused on housing (21%) and security (18%), while the smallest proportion of answers related to reconciliation (13%). Interestingly, most respondents who disagreed with the Return Index’s assessment of employment opportunities in their area believed that the situation was worse than previously identified by interviews. Job opportunities are very few because of the lack of support for agricultural and industrial projects. Participants highlighted underinvestment, agriculture and rural issues, social marginalisation, and COVID-19 as sources of unemployment. For example, a female respondent between 18 to 34 years from Ninewa cited ‘favouritism and corruption’ as obstacles to employment. Meanwhile, a male from the same age bracket from Kirkuk quoted that a ‘lack of job opportunities and employers’ failure to appoint people with academic degrees are the reasons behind why they do not have work’.
In terms of business recovery, respondents to the Collective Intelligence platform identified a lack of investment, the scale of destruction and slow returns as important factors influencing business opportunities. Comparatively, the key informant interviews had a much more positive understanding of the business environment. Collective Intelligence respondents noted that cycles of violence and displacement had severely weakened the recovery of businesses in many areas of return. Business owners lacked the financial means to repair premises, infrastructure was damaged, and the fragmented security situation limited the movement of people, capital, and goods.
The inputs and voices of affected populations, as observed through the online platform, have created room for disagreement with the results of traditional data collection.
The third indicator where the Return Index and the Collective Intelligence platform differed was community reconciliation. Many Collective Intelligence respondents noted that peaceful coexistence was possible because of a lack of sources of tension within their community, such as tribal disputes. Many also stated that ethno-religious homogeneity in their area was the crucial determinant of peace within the community. This perception may be threatened by the return of marginalised communities and may also increase the vulnerability of displaced minorities that are currently hosted in those areas. For example, a male in Baghdad between 35 and 59 years noted that ‘All the people of the area are from one ethnic group and get along with the tribes and all other groups in the sub-district. The tribes have an obligation to fight terrorism in the district, and this is what happened after retaking the areas from ISIS’. These findings are valuable to unpack underlying barriers to sustainable returns which go beyond those identified by the Return Index.
These disagreements between the traditional key informant interviews and Collective Intelligence respondents are important for several reasons. Issues around returns, displacement and forced migration are complex and multi-layered. To fundamentally transform how we confront these issues necessitates newer approaches and perspectives which include voices and knowledge beyond the current rubric. Community feedback on sensitive topics such as perceptions of corruption are particularly relevant for understanding obstacles to employment, while insights on community dynamics are important to inform programming.
The success of creating a digital anonymous ‘safe space’ to encourage, aggregate and better understand such perspectives leads us to discuss some of the key methodological takeaways from this pilot. Even though COVID-19 related restrictions in Iraq impeded outreach capabilities to disseminate the platform, it led to the creation of an entirely digital feedback mechanism. This facilitated an opportunity to gain critical feedback in situations where respondents may feel more comfortable being anonymous and are not exposed to the risks associated with identification. Anonymity lent a feeling of protection and a chance for affected populations to voice their concerns without the power dynamics between humanitarian actors and affected populations. The qualitative responses attest to a willingness to provide non-personal information, which also eliminated data protection concerns.
Participatory modes of data collection are ethical and enable affected populations to assert agency over decisions affecting their lives.
Forced migration data is richer when there is disagreement and variance because it contributes towards a more holistic and representative understanding of the situation on the ground.
Limitations and areas for future development
This pilot demonstrates new opportunities for broadening the scope of participation and representation in displacement data, but equally, it poses some barriers. Due to the small number of locations (192 locations) and respondents in the Collective Intelligence process versus in the Iraq Return Index (2,706 locations), the qualitative differences cannot be generalised to the whole population. The digital platform can be seen as a double-edged sword. While participation of newer demographics catalysed critical feedback and disagreement, participation was limited to digitally connected individuals (mostly younger males). It is vital to find the balance between using new technology while recognising other structural asymmetries that may arise due to digital divides and associated disparities. While digital technologies contribute to diversifying participants, the need to involve marginalised groups persists. Overall, even though social media platforms enabled us to improve participation, engage with new demographics, and create a ‘safe space’ for affected populations to participate in analysis, there is scope to improve and to scale up to involve more people. For example, engaging a larger sample of citizens would make it possible to draw statistically significant conclusions and to strengthen the comparison between the two methods.
Potential responses to ongoing challenges
This pilot provided humanitarian and development practitioners one example of how to develop low-tech modalities to engage diverse groups of affected populations.
- The Collective Intelligence methodology or ‘crowdsourcing component’ can be integrated into data collection processes to provide validation and feedback in a regular but sustainable manner.
- Simultaneously collecting data using different methods can help engage affected populations to validate analysis, to identify gaps, and to mitigate assessment fatigue.
- Processes which combine qualitative and quantitative methods can be used to confirm the situation on the ground.
- Tools designed for Monitoring, Evaluation, Accountability, and Learning can benefit from adopting a complementary Collective Intelligence approach to data collection to collect anonymized feedback from affected populations to improve programming.
Overall, with global trends of increasing phone ownership and Internet use, there is emerging need to explore digital participation methods that complement existing processes. For policy actors, the strong participation in the pilot suggests a good appetite for using similar methods in contexts with high mobile phone use. New methods for data collection need to connect with more “marginal” or less traditionally engaged groups who are often excluded from current data collection tools. Finally, this pilot showed a case where macro-level challenges around power, participation and local knowledges of displaced populations were tackled in a local context using technology to complement if not entirely shift data collection methods. Valuing participatory processes is an important step towards promoting inclusivity and shifting power dynamics.
Trigwell, R., J. Phillippo-Holmes, E. Zambrano, J. Bahn, P. Hirani and E. Griesmer, 2022. “Validating Humanitarian Data Analysis Through Collective Intelligence: A Pilot Study”. International Organization for Migration, Geneva. https://publications.iom.int/books/validating-humanitarian-data-analysis-through-collective-intelligence-pilot-study.
Dr Prithvi Hirani
Programme Officer, International Organization for Migration
Robert Trigwell
Senior DTM Coordination Officer, International Organization for Migration
LERRN Blog Series: Knowledge, Voice and Power
This blog series continues the conversation from the recently launched Forced Migration Review 70th issue on Knowledge, Voice and Power, which was supported by LERRN. In this blog series we feature authors from around the world reflecting on questions about involving people with lived experiences of displacement in multiple contexts, as well as scholars based in the Global South, from research to policy discussions.