Third report from Independent Ethics Advisor

Third report from Independent Ethics Advisor Minna Zechner University of Helsinki December 2025 Mapineq deliverable D10.4 Third 2 Mapineq – Mapping inequalities through the life course– is a three-year project (20222025) that studies the trends and drivers of intergenerational, educational, labour market, and health inequalities over the life course during the last decades. The research is run by a consortium of eight partners: University of Turku, University of Groningen, National Distance Education University, WZB Berlin Social Science Center, Stockholm University, Tallinn University, Max Planck Gesellschaft (Population Europe), and University of Oxford Website: www.mapineq.eu The Mapineq project has received funding from the European Union’s Horizon Europe research and innovation programme under the grant agreement No. 101061645. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union, the European Research Executive Agency, or their affiliated institutions. Neither the European Union nor the granting authority can be held responsible for them. Acknowledgement: The content of the document, including opinions expressed and any remaining errors, is the responsibility of the authors. Publication information: This work is licensed under the Creative Commons Attribution-Non-Commercial-Share Alike 4.0 International (CC BY-NC-SA 4.0) license. You are free to share and adapt the material if you include proper attribution (see suggested citation), indicate if changes were made, and do not use or adapt the material in any way that suggests the licensor endorses you or your use. You may not use the material for commercial purposes. Summary history Version Date Comments 1.0 12.12.2025 Manuscript for review 1.1 16.12.2025 Manuscript reviewed for submission Suggested citation: Zechner, Minna (2025). Third report from Independent Ethics Advisor. Mapineq deliverables. Turku: INVEST Research Flagship Centre / University of Turku. DOI: https://doi.org/10.5281/zenodo.17950671 Third 3 Executive summary In this report I chose a three-part approach by 1) checking if the issues from the previous two ethics reports have been addressed, 2) reading the present Data Management Plan (2025) as such and 3) comparing the Data Management Plan (2025) to a template for the same document from the European Commission (Horizon Europe DMP template, 2025). Below I have listed the main observations after concluding the three steps. 1. The aggregating of coarser spatial scales may thus deserve more than consideration, especially then the data may relate to potentially sensitive issues such as racial or ethnic origin, political opinions, sexual orientation, health, religious or religious beliefs. Each of these and more may be present in the data included in the Mapineq Link database. Rule of ten or even hundred is not very coarse when there is a question about sensitive issues. 2. It seems that in connection with the vignette data collection some national workshops were organized (Reality Check Workshops) where possibly some qualitative data was collected. This needs to be clarified and whether this data will be part of the Mapineq inequality database. 3. There are several references to health and genetic data and steps are explained that consortium partners must follow if anonymised, aggregated and licenced for open access indicators from some of these data sets will be added to Mapineq Link database. These steps are essential, but they do not allow a clear picture whether the licencing partners are fully aware of the other data in the Mapineq Link database and the possibilities the combination of the existing and new data may create in for example locating areas and groups of people with certain health issues, such as HIV or other conditions which tend to be stigmatizing. 4. When reading the Data Management Plan (2025) as a stand-alone document, there are still some issues that are not very clear for the reader. On of them is that it is not always obvious which part of the data in the Mapineq Link database is (and will be) available for only researchers in the consortium and which ones are (and will be) openly available for other users. 5. Intellectual Property Rights (IPR) are not discussed. Overall, the Data Management Plan (2025) that this ethics report is based on shows commitment to the All-European Academies (ALLEA) European Code of Conduct for Research Integrity (2023), and the requirements set by the General Data Protection Regulation (GDPR). The share of text and explicit consideration on ethical issues has become more detailed and increasing in amount since the two earlier ethics reports (Zechner 2023, 2024). As the list above shows, some issues need to be still considered. Third 4 Abbreviations API Application Programming Interface DMP Data Management Plan GDPR General Data Protection Regulation IPR Intellectual Property Rights NUTS Nomenclature of Territorial Units for Statistics Third 5 Content EXECUTIVE SUMMARY 3 ABBREVIATIONS 4 1. INTRODUCTION 6 1.1. PREVIOUS ETHICS REPORTS 6 1.2. ASSESSMENT APPROACHES 7 2. COMPARING WITH PREVIOUS ETHICS REPORTS 8 2.1. UNEXPECTED AND INCIDENTAL FINDINGS 8 2.2. MISUSE OF RESEARCH FINDINGS 9 2.3. EVOLVING DATABASE 9 2.4. HEALTH AND GENETIC DATA AND DATA LICENCES 10 3. PRESENT DMP 10 3.1. THE DMP AS AN ENTITY 10 4. COMPARING WITH HORIZON DMP TEMPLATE 11 4.1. COMPATIBILITY WITH HORIZON DMP TEMPLATE 11 5. CONCLUSIONS 12 6. REFERENCES 13 Tables TABLE 1. MAPINEQ DMP (2025) COMPARED WITH THE HORIZON EUROPE DMP REQUIREMENTS __________________ 12 Third 6 1. Introduction Mapineq – Mapping inequalities through the life course– was a three-year project 20222025 (www.Mapineq.eu) that studied the trends and drivers of intergenerational, educational, labour market, and health inequalities over the life course during the last decades. The research was run by a consortium of eight partners: University of Turku, University of Groningen, National Distance Education University, WZB Berlin Social Science Center, Stockholm University, Tallinn University, Max Planck Gesellschaft (Population Europe), and University of Oxford. 1.1. Previous ethics reports The Mapineq project has received funding from the European Union’s Horizon Europe research and innovation programme under the grant agreement No. 101061645. As the project focuses on topics, that are in many ways sensitive and uses complex and detailed data, there is need to pay special attention to ethics. For this purpose, an external ethics advisor (Associate Professor Minna Zechner from the University of Helsinki), was nominated. The first ethics report by the advisor was submitted in September 2023 and the second one in September 2024. Both pointed out that while many ethical issues were already considered by the time of the first report, there were still several concerns, summary of which are listed below (Zechner 2023, 2024). 2023: - It is unclear whether the unexpected findings policy also covers incidental findings. There is also a question how the cases are handled if unexpected or incidental findings are done by others than researchers. - The risk of misuse of research findings has not yet been addressed. This was already mentioned in the original Ethics Summary Report from the application evaluation. - For collecting the qualitative data there is need to prepare for potential discomfort study participants may experience when encountering questions about inequalities such as racism or mental health stigma. Similarly self-discrimination may be an issue when people participate in data collection or use the database. - The intent is to use the pseudonymised qualitative data as an input for the Reality Workshops, but it may be necessary to use anonymized data instead. It is also not clear why there are no intentions to archive the qualitative data for further use. - Regarding the qualitative data, it needs to be established whether it requires a privacy notice. This document is also important when informed consent is sought from the study participants. 2024: - How and by whom the unexpected and incidental findings are recognized and handled after the project has ended and the database is still maintained? - Misuse of research findings has not been addressed. Third 7 - How is it secured that the existing and additional geographic frameworks, combined with socio-economic indicators are not too fine grained so that it would not allow for example identifying and locating small groups of people who might be at risk of persecution or exploitation? - The database is work in progress and the more data there is, the more ethical consideration is needed, so how the entity of the database will be managed in terms of ethics as new data is added? - How is the compliance of data licences ensured as the API (Application Programming Interface) allows free use of the data for anybody without registration? - There are at least plans to use health data from wearables or body monitors and Apps that measure bodily functions. This kind of data has become increasingly detailed, due to technological development. Therefore, it is essential to consider whether incidental findings regarding health are possible via using this data? - Would there be a risk of political persecution when political data on political processes, politicians, voting and the electorate is combined with geographical data? - Please clarify whether genetic data will be used, if so, for which purposes and how. For example, unexpected and incidental findings and misuse of research findings were reoccurring themes. In the first report more emphasis was on qualitative data and in the second one on the risks of combining data, licencing and adding new data as well as political and health-related data. This report is based on the following MapIneq documents: • Erola, Jani, Vono de Vilhena, Daniela, Kilpi-Jakonen, Elina & The Mapineq Team (2025) White Paper on Social Inequalities in Europe: From Early Childhood to Retirement. Findings and Recommendations from the Horizon Europe Project ‘Mapping Inequalities Through the Life Course’ (Mapineq). Berlin: Population Europe. • Figueroa González, Mijail & Leasure, Douglas, R. (2025) Data Management Plan (DMP). Mapineq deliverable D9.3. Updated in December 2025. The White Paper was to give an overview of the project phases and outcomes, and the Data Management Plan is the core document for ethics and thus on the focus of this report. 1.2. Assessment approaches I took three approaches in assessing the latest version (December 2025) of the Data Management Plan by Mijail Figueroa González and Douglas R. Leasure. First, I checked the issues that I had pointed out in my previous Ethics Reports (Zechner 2023, 2024). Focus is on issues that still called for consideration. Second, I read the Data Management Third 8 Plan (DMP) as such to see if it is understandable and covers issues that are essential in regards ethics. Third, I checked the Data Management Plan against a template for the same document from the European Commission (Horizon Europe DMP template, 2025). 2. Comparing with previous ethics reports 2.1. Unexpected and incidental findings Data Management Plans are living documents, and they are updated when new issues emerge. The first issue that I pointed out in both previous reports related to unexpected and incidental findings. In the present DMP (2025) the risk of unexpected findings in considered low on the grounds that data in the Mapineq Link database are already in the public domain and that anybody could have combined these data already (p. 26) 1 . The data is also aggregated and anonymised and does not include any personal or individual level data. It is also stated (p. 25) that “Each new dataset that is added to the Mapineq Link database is evaluated on a case-by-case basis to assess potential of risks in combination with other data in the database—or publicly available from other sources. The risks are significantly mitigated de facto by the coarse spatial resolution of geographic levels available in the Mapineq database (i.e. smallest area is NUTS-3 which is more coarse than municipalities). However, there is the potential for heightened risks with some data that are related to vulnerable groups. We consider additional mitigation strategies in these cases that may include aggregating to coarser spatial scales (e.g. NUTS-0, country level), removing certain filters from the data (e.g. disallowing filtering to specific migrant groups), or omitting the most recent time period from the data.” NUTS means Nomenclature of Territorial Units for Statistics and I pointed out in 2024 that there is a risk that existing and additional geographic frameworks, combined with socioeconomic indicators may be too fine grained so that it could allow for example identifying and locating small groups of people who might be at risk of persecution or exploitation. The aggregating of coarser spatial scales may thus deserve more than consideration, especially then the data may relate to potentially sensitive issues such as racial or ethnic origin, political opinions, sexual orientation, health, religious or religious beliefs. Each of these may be present in the data included in the Mapineq Link database. In addition, it was pointed out (p. 26) that “When including data from other sources, we will take measures to ensure we are not displaying information that would locate small or specific groups (e.g., 10 people or less).” Ten people is a very small group and does rather little to protect individuals if they were targeted for example based on their political views. It is still noted (p. 26) that “The NUTS administrative units used by the dashboard are intentionally larger geographies in sparsely-populated rural areas and smaller geographies in densely-population urban areas. This, again, mitigates risks of geolocating specific 1 The page numbers may not be correct later as the DMP I was reading still had tracked changes visible. Third 9 groups of people.” This is a good approach, but if the rule is ten people or more, one apartment building or a residential quarter could already be recognised in for example ethnically homogenous areas. Finally, the DMP explains that journalists, researchers and data scientists may benefit from the database as it enables “to drill down to very small regional differences within a country” (pp. 10, 11, 12) and “to visually explore geographic differences, from country differences in spending to unemployment or cause of death rates in small regions” (pp. 10, 12). When these small regions are compared, as it is suggested for journalists (p. 11) to “highlight how certain areas perform relative to others within a country or in comparison to similar regions in different countries, shedding light on best practices and lessons learned.”, there is a possibility for presenting certain areas in a very negative light. The DMP suggests focussing on best practices, which is positive. However, these do not reflect very coarse spatial scales and require thus further explanation. 2.2. Misuse of research findings The risk of misuse of research findings or rather the data in the Mapineq Link database was not addressed in the first two Data Management Plans. In the present (2025) one this is pointed out as follows (p. 25): “The primary risks arising from unexpected findings are related to interpretation of the findings: (1) misinterpretation of correlations as causal relationships, and (2) cherry-picking specific results to support a narrative while ignoring potentially greater evidence to the contrary.” I understand this not only as unexpected findings but also misuse of research data by users outside the research team. The mitigation of this problem is to provide disclaimers, explain the difference between phenomena being simultaneous, associated or having a correlation. Also, guidance on how to use the database and that some questions require rigorous scientific methods to be answered. These measures have their limitations, but when the data is open (aggregated and anonymous), advice and support are the main means for mitigation. 2.3. Evolving database The database is work in progress and the DMP promises to maintain the Mapineq inequality database for at least five years after the end of the project. This includes checking the quality of the re-used survey data such as UK Household Longitudinal Study or European Union Labour Force Survey and it is promised that “researchers of the project will use the most recent versions published by data providers” (p. 16). It is not however clear whether these quality checks cover also the public version of the data. Some contingency plans for the future have already been made regarding the possibility of adding aggregated individual level data (p. 27): “Future work after the project period may pursue bespoke aggregation of individual-level data from registries or other sources to derive indicators for the Mapineq database.” This is a positive development when done with care. In the earlier version of the DMP (2023) collection of qualitative data was mentioned. In the present version (2025) it is not mentioned, only vignette data was collected, but it