Sharing childrens’ data online#
What’s this?
This is summary of Wednesday 25th October’s discussion, where we spoke and wrote about the article How influencer ‘mumpreneur’ bloggers and ‘everyday’ mums frame presenting their children online, discussing the sharing of information about children online. The summary was written by Jessica Woodgate, who tried to synthesise everyone’s contributions to this document and the discussion. “We” = “someone at Data Ethics Club”. Vanessa Hanschke and Nina Di Cara helped with the final edit.
The study compares “influencer mums” to “everyday mums”, evaluating their online behaviour and approach towards privacy concerns. Findings demonstrated that everyday mums demonstrated concern about privacy, but did not understand the complexity of laws, issues and technology around social media. Some influencer mums showed hesitation surrounding privacy of their children but others were less concerned about the rights of their children as they were young.
Q1 What (if anything) concerned or surprised you about the attitudes of everyday mums? Why?#
We felt some surprise and concern about the pressure that “everyday” mums felt to publish. Some participants expressed that they felt concern about publishing photos, yet still went ahead with it. We wondered if this could have been more deeply analysed if the paper had explored participants motivations for publishing on social media. Some of us had personal experience as mums who have gone through periods where we were isolated from support networks, and social media filled a gap. Other motivations include the pull factor of social media from the generation of social capital. The part social media platforms play in incentivising engagement is an important part of this discussion and is underexplored in the paper. Those of us with experience working in social media found that working in the industry does affect your understanding of the kinds of things that end up on social media, and the repercussions of that.
Participants’ understandings of social media seemed incomplete. We found it interesting, although not necessarily surprising, that there was a gap between how informed participants thought they were, and the efficacy of their actions to prevent risks. For example, some participants used pseudonyms to protect their privacy. However, this does not seem sufficient if mums are using their own names or posting a great deal of detail about their lives, including photos of their homes and where they live. We wondered if participants understood where data about their family is going, and how it is being used. There is an important consumerist/commercial aspect to this story, as platforms foster engagement to harvest data. Advertising (which produces a lot of the profit for data harvesting) clearly does work, even if the algorithms aren’t perfect. There is a lack of informed consent about the data that social medias volunteer to capitalist data harvesters klaxon.
Differences in attitudes to engagement with social media are perhaps not as clear cut as the division between “influencer” and “everyday”. The influencer group seemed to be always embracing the performative element of social media. There is a specific persona and brand they are trying to get across, and their approach to privacy is a part of that manufactured image. However, everybody has personal rules about how they engage with social media, and how you present yourself online. We thought that the paper presents it as if only the influencer mums are curating what they put forwards. However, to different extents, this is the case for everybody.
We wondered how attributes such as age, gender, and socioeconomic background might influence the way we present ourselves online. It might be the case that there is a big difference between the social media presence of a 19 year old mum and a 45 year old mum. The article focussed on generally affluent participants, and we wondered how socioeconomic background might affect interaction with and understanding of social media. The article does not discuss the behaviour of Dads, which we were not keen on. We wondered if Dads do interact differently, and if they do, why? Gender stereotypes is an important part of the story; part of the reason there is such a big mum blogger culture may be because it is mums who are more likely undertake childcare. In the landscape of mum bloggers, there is a big “tradwife”/SAHM (stay at home mum) culture. We thought this was interesting as, whilst their image may capitalise on this lifestyle, they are actually running a business through their influencer jobs.
Alongside individual differences in engaging in social media, we wondered if there is a difference in how people interact with different platforms (e.g. Instagram, Facebook, YouTube). Different platforms lend themselves to different types of interactions, for example Instagram encourages shorter comments, whilst people often leave paragraph length comments on Facebook. Certain platforms such as Instagram (primarily images) and TikTok (primarily videos) drive behaviour to be more identifying.
Q2 How effective are the tactics used by each group for maintaining children’s privacy in your opinion?#
As well as concerns about the efficacy of pseudonyms previously discussed, we were worried about the loss of control when images are placed on a public platform. Once they’re out there, it is difficult to prevent images from ending up in other domains or AI training sets. Education around this should be improved; some people may believe that when you upload an image it is locked to that platform, however often this is not the case.
The perpetuity of images elicits questions about who has control over time, and what happens if people want their images to be withdrawn. The internet is not set up for forgetting, as data is constantly being copied and stored in different locations. Similarly, AI has the issue of either catastrophic forgetting, or not knowing how to forget. The right to be forgotten is especially important in the case of children, who will one day grow older and might change their mind about how they want to be presented online.
Children do not have much of a say over what their parents post about them online, raising concerns about autonomy and safety. We wondered at what point a child should get a say in what’s being posted about them. This is especially difficult for young children, who do not have reasoning capacities. Some of our personal experience involved specifically asking our children if we could take a photo of them, who we were taking it for, and how we were going to send it. This way, we are asking their permission, as well as teaching them about digital literacy and consent.
The first generation of the children of influencers have begun speaking out. France is drafting legislation to protect children’s rights to their own images and prevent parents from oversharing. There is also an interesting case from the film “the sound of freedom” – a Hollywood film based on the true story of Tim Ballard, who saved children from sex trafficking. In real life, Ballard adopted some of the children he rescued. Today, that family is in the public domain with their personal stories widely publicised and their photos on Ballard’s social media. We voiced concern about their safety, and the effect publicity might have on children who have gone through trauma. We thought that the Kardashians are another interesting case study about young children being all over social media, and wondered how much of a choice they have, even as adults. When your life has been so public from such a young age, this will undoubtably shape how you behave as an adult.
Behaviour of children when they are being filmed does change, and we wondered what conversations parents are having surrounding agency and long term impact. For influencer parents, could this result in a sense of mistrust and doubt, wondering if situations in their lives arose because it was generating good content for their parents. On a more general basis, we wondered if young people today behave more sensibly and are more prone to anxiety partially because they feel more observed. We thought that there was a surprising lack of specific risks, such as damage to parent-child relationships, mentioned by parents in the article.
Utilising images of children to promote brands and gather revenue seems like a form of child labour, and we wondered how it intersects with child labour laws. How is the allocation of time to influencing balanced with education, and are there other effects to their education? We found the use of the term “play”bour a bit scary; if a lot of the images are of children, yet parents are the ones receiving the money, who should have ownership of those funds? This relates back to questions over ownership of images, for example if they end up in catalogues, and what happens if different parents have deferring views.
Using children to gain publicity is something which mainstream media is guilty of as well. Emotive images of children in distress are often picked in charitable fundraising campaigns, or by journalists to increase engagement with stories. Provocative imagery is effective in bringing stories closer to home, however, we wondered how much consent can be obtained by people in traumatic situations – especially if those people are children. In situations where an organisation (be it a charity or a journalist) from a wealthier country is taking photographs of people from poorer countries, there is an important difference in power dynamics blurring the line around informed consent. The problem of vulnerability is especially apparent in wartime media. There are restrictions around publishing photographs of the children of celebrities, and we wondered why this collective social responsibility does not extend to other children.
Vulnerability also enters the conversation when we turn to the topic of mental health and neurodivergence. In the article, some of the mums’ online presence revolved around sharing their experience of having children with learning difficulties. There are problems here with the ability to give informed consent, yet we could see how this kind of influencing can be reassuring; providing resources to other parents and a means for people to connect with others in a similar position. In order for the audience to connect, people might want to know identifying characteristics like age or gender, presenting a more detailed story that they could relate to. There might be some level of metadata which an audience could reasonably expect to include. However, we wondered if it was necessary for children’s faces and identities to be included to build a community. There might also be differences in the way that parents interact with social media if the parents themselves are neurodivergent.
Q3 Do you have any comments on the research itself, e.g. limitations or things you would have liked to see?#
The motivation for research was questionable; it seemed that perhaps the author felt quite strongly that sharing images of children online was a negative thing, which influenced the way the article was written. This impacts the kinds of assumptions that readers approach the paper with. There was no reflexive statement about the author’s viewpoints, which would have improved the piece.
The terminology used in the piece was problematic. The term “influencer” was not really defined; the number of followers an influencer has – thousands, hundreds of thousands, or millions – could have an impact on their behaviour and interaction with social media. “FGM” was used in reference to a focus group in the paper, however in the world of working with children this means something very different. We should be careful about the terminology and abbreviations we use, especially surrounding children. We didn’t like the division between “influencer” and “everyday” mums. Parts of the article positioned “everyday mums” as being very unsophisticated. We thought this was a big limitation of the methodology, as it is not necessarily a representation of the population as a whole.
Terminology was not the only problematic aspect of how the different groups of participants were separated. Interview techniques were very different for the different groups; whilst “everyday mums” were put in focus groups of 10 people, “influencer mums” had individual interviews. This doesn’t seem like a fair comparison, as relational dynamics may have an effect on the way conversations happen and the answers people provide. There were noticeably fewer “influencer” mums compared to “everyday” mums, and we wondered why this was the case. The diversity of participants was not discussed, and we wondered how results would change if the paper accounted for diversity aspects including separated parents and dads. This hugely limits the scope of participants.
For the participants that were included, we thought that more instigative questions could have been asked. It seemed that the interviews just cracked the can of the topic, and then left it there without delving very deep. There could have been more critical provocation, especially towards the end of the interviews.
This highlights a missed opportunity for intervention. Asking questions about privacy encourages parents to reflect – for perhaps the first time. Including external resources in interviews, such as explanations about regulation, could have been a great learning experience. An interesting follow up could have been to examine if such an intervention did affect the behaviour of participants.
Future work could expand to different social media platforms, like Reddit for instance. The paper is four years old, and a lot has changed in the realm of social media. We wondered if the findings are outdated now, and if there are differences in how people consume social media. The author has since moved into other areas, so we couldn’t find any further studies.
Bonus Question: What change would you like to see on the basis of this piece? Who has the power to make that change?#
Currently, the minimum age to hold an account on social media is about 13. However, the system allows us to represent children much younger than that. We wondered if this is fair, as young children do not have autonomy over the images shared of them. Without a flat ban, which seems unrealistic, this is difficult to navigate. The line denoting where posting is going beyond sharing a photo with friends is unclear, and perhaps should be further discussed.
Attendees#
Noshin Mohamed, Quality Assurance in Children’s Services
Natalie Zelenka, Senior Research Fellow in Health Data science, @NatalieZelenka GitHub,
Lucy Bowles, Data Scientist at Brandwatch
Jessica Woodgate, PhD student at University of Bristol
Nina Di Cara, Research Associate at University of Bristol
and others!