ARTICLES

 

Exploring Ethical Complexities in Social Media Data Mining: Questions From a Researcher's Journey Investigating Depression

 

 

Lesedi Kgatla

University of South Africa ORCID No: 0000-0002-8110-7867 brightnesslesedi77@gmail.com

 

 

---

ABSTRACT

Mining data on mental health topics such as depression on social media platforms presents a myriad of ethical challenges. This paper delves into the nuanced ethical considerations encountered during the collection of data on depression from social media platforms, highlighting the blurred and complex nature of research ethics in this domain. The paper investigates the ethical dilemmas inherent in mining data in the public domain of social media. It addresses concerns related to privacy, consent, and the potential for harm to participants, emphasising the difficulty in distinguishing between public expression and private disclosure in online spaces. Furthermore, the paper explores the challenges of ensuring the confidentiality and anonymity of participants in studies involving mental health issues. It discusses the tension between the need to protect individuals' identities and the imperative to analyse data in a meaningful way, while also acknowledging the risk of re-identification. Moreover, the paper reflects on the ethical implications of engaging with vulnerable populations and the responsibility of researchers to minimise harm and prioritise the well-being of participants. It discusses the importance of employing sensitive and ethical research methods when studying individuals' experiences with depression on social media platforms. Throughout, the paper underscores the importance of ongoing ethical reflection in navigating the complex landscape of social media research. By sharing personal experiences and insights, it aims to contribute to a deeper understanding of the ethical challenges inherent in mining data on social media, ultimately advocating for ethical research practices that prioritise participant well-being and uphold research integrity.

Keywords: research ethics, online research, data mining, mental health

---

 

 

Introduction

Allow me to take you through the journey of my data collection method that took place on social media for my doctoral study (Kgatla, 2023). The paper will focus on the journey from the beginning until the end, highlighting different ethical challenges I encountered while mining data on social media. Data mining involves extracting valuable, complex information from large datasets (Hashmi & Ahmad, 2016). For instance, I gathered information from participants' social media pages and dedicated part of the interview to analysing that data with the participants. Prior to starting my data mining on social media, I had preconceived assumptions about how the journey would unfold. My assumption was based on the idea that the study's criteria involved people who were vocal about their experiences with depression on social media. Thus, I assumed that people already sharing their experiences with depression online would not have an issue with participating in the study. I also assumed that it would be easy to identify such people on social media.

Ethical requirements are very important in qualitative research or any other research to ensure that participants are respected and not taken advantage of. This is to ensure that people's human rights are not violated by the researcher (Mirza et al., 2023). Therefore, I needed to ensure that my participants were comfortable about participating in the study because it focused on a sensitive topic. Their comfort was a priority. Additionally, I needed to ensure that participants understood that their participation was voluntary and there were no penalties for withdrawing or refusing to participate. Furthermore, because the study was partly based on mining data on social media, I needed to ensure confidentiality was upheld by protecting participants' identities. Another important ethical requirement was ensuring that participants were protected and safe. These are all ethical requirements expected in any research undertaken, whether collecting data traditionally or in digital spaces (British Educational Research Association, 2018). These spaces are very different, and I realised that data collection in digital spaces also presented different challenges to those when collecting data in traditional face-to-face interactions. These challenges will be discussed in detail in subsequent sections.

In this paper, I explore the differences in expectations about mining data on social media in order to investigate the challenges and opportunities from data collection in digital spaces from vulnerable people. Thus, the paper will focus on various ethical requirements that need to be carefully followed when gathering data in digital spaces about vulnerable individuals. Reflecting on my experience was essential for gaining a deeper understanding of the ethical challenges inherent in mining data from social media.

Ultimately, I advocate for ethical research practices that prioritise participant well-being and uphold research integrity.

 

Informed Consent? Easier Said Than Done on Social Media

Before beginning with data mining, ethical clearance must be sought so that the researcher can proceed with their study (Mirza et al., 2023). An important process, after identifying potential participants, is ensuring that consent is sought from the participants to take part in the study. Usually in a traditional face-to-face data collection, participants sign an informed consent form agreeing to participate in the study, that they understand what the study is about, and that there is no compensation for participating or penalty for withdrawal.

Thus, because the study had two data collection methods-data mining and interviews-I sought consent from participants for firstly mining data from their social media profiles. Participants had to sign an informed consent form agreeing that they understood that I would go through their social media pages and use their content for the study. Additionally, participants gave consent to be interviewed and to be recorded for the purposes of the study. However, there were few challenges that I encountered when mining data from participants' social media platforms. These led to complexities and questions that I needed answers for in order to ensure that ethical requirements were upheld in my study and in future studies.

I now provide a brief description of various ethical challenges I encountered in terms of informed consent on social media. Protecting social media users is very important, and users have various options to ensure that they can protect their content and choose with whom they share their content. An option usually invoked by celebrities or social media influencers is a disclaimer in their profile description (bio) that states that if anyone intends to use their content, they need to first seek permission from them. Another option is for users to set their social media profiles to private so that others cannot access their content without requesting to follow the said user first. Lastly, users can simply set their social media profile to public, meaning it is accessible to everyone and their content is readily available to anyone.

Each of the aforementioned options presented various complexities when mining data from participants' social media platforms. Fortunately, no participants had a disclaimer on their bio, thus I did not need permission or consent to use their content. However, the point still evoked questions, for instance, if a potential participant has a disclaimer on their bio and you seek consent before using their content, does that count as informed consent? Or do they need to sign a separate document agreeing that you may use the content? Given that the use of data on social media is acceptable in research, should the disclaimer not be acceptable and sufficient as consent? Can that disclaimer in the bio and subsequent conversation with the potential participant be produced as evidence that consent was conferred?

Many people employ the option of making their social media accounts private. Some of my potential participants had such private accounts, so I had to request to follow them before I could access their content. The question that arose from this option was, given that I had had a conversation with participants prior to accessing their social media account, in which I explained that their content would be used in the study, did I still need consent from them in written form in order to access and use their social media content? In other words, if they know the reason behind me requesting to follow them and acquire access to their content, do I still need a written consent? Can a screen shot of their account (showing that it is a private account) serve as enough evidence to indicate that consent was clearly provided given that I was able to access their information?

The third option is usually employed by many social media users-leaving their social media pages public, meaning their content is accessible to anyone who visits their social media account. Thus, because their information is on a public platform and they are consciously choosing to let it remain public, knowing that it will be accessible to anyone, is permission still needed to use their content? Thus an important question arose, namely, because social media can be public and if a user does not use the options of a disclaimer or to privatise their account, does that mean they have it given other people permission to do as they please with their content? The question arose because even though an account can be public, the user nevertheless still has to provide their personal details such as email address or private mobile number and a password to access that account. This led to me to wonder whether it was invading people's privacy if consent was not sought? Because, as much as other people can see their content, it still requires a password for them to log into their profile, which is meant to serve as protection of their information. Additionally, if in the duration of data collection for the study, the participant had their account set to public and I accessed their data without consent and later, they changed their setting to private or added a disclaimer, could this impact the ethical compliance of the study given that there would not be any evidence of seeking consent to use their content?

Thus, the main question for all these complexities was: "Because social media users need to provide personal details to be able to access their account (regardless of whether they add a disclaimer or set the account as private or public), do researchers still need consent to use content on participants' social media pages?" The answer is simple. It is not about simply setting accounts as public or private or having a disclaimer; most important, is to go to the core of ethical adherence and understanding, which is that in any research, there should be human decency and people's human rights should be protected and upheld.

Therefore, regardless of how a user chooses to set their social media account, the fact that they need to provide personal details before accessing that account means that whatever content is there is private, and researchers need to seek consent to use it. This helps in protecting your participants and ensuring that you respect them and make them comfortable in participating in the study, knowing that they are being treated as humans (Mirza et al., 2023).

Navigating ethical considerations when mining data from social media platforms presents complex challenges for researchers. While some users may have disclaimers or set their profiles to private, and others maintain public profiles, the fundamental principle remains the same: respect for individuals' privacy and autonomy. Despite the visibility of content on public platforms, the underlying personal details required for access highlight the need for explicit consent. Whether through formal agreements or implicit understanding, researchers must prioritise the ethical treatment of participants, safeguard their rights, and ensure their comfort and dignity throughout the research process. Ultimately, ethical compliance in data collection transcends mere categorisations of online visibility, and emphasises the paramount importance of human decency and the protection of individuals' rights in all research endeavours.

 

Are You Comfortable?

My doctoral study (Kgatla, 2023) was titled, How and Why do Young People Use Social Media to Cope With Depression. Therefore, I decided to first mine data on participants' social media platforms. Then I would do interviews based on the data mined, and other pre-structured interview questions. The criteria for the study included using social media to cope with depression-either by posting on social media about their experiences with depression or, if not posting, using social media to follow people or pages that focused on mental health awareness, how to cope with depression, or psychoeducation. Thus, I expected people to be willing to participate in the study because they were already on social media and using it as a coping tool for their experiences with depression.

Firstly, I approached people I knew who posted on social media about their experiences with depression and used social media as a coping mechanism. Some potential participants I approached agreed to participate in the study because they were mental health activists on social media, which made it easy for them to participate in the study. I therefore assumed that if people were already being vocal on social media about depression, it would be easier for them to participate in the study. But, to my surprise, some were not keen to participate in the study even though they were mental health activists and vocal about their experiences on social media.

I started to visit various social media platforms such as Facebook, X (previously, Twitter), Instagram, and TikTok to find people who shared their experiences of depression online. I would first visit a platform and search for #depression or #mentalhealth, and go through each profile to see if it fit the study's criteria. I had to take note that some people might post about depression or mental health as a passing message for awareness, and that they were not necessarily using social media to cope with depression or experiencing depression themself. I had to navigate various accounts and focus exclusively on those that were clearly aimed at creating mental health awareness or using the platform to cope with depression. This included viewing their posts, captions, statuses, and checking whether they were constantly posting about their experiences. It is important to note that I only viewed their profiles to check if they used social media as a coping tool, and I did not use any of their content before I had gained their consent. Therefore, those who fit the study's criteria were messaged. Some did not respond to my message, others agreed to participate, and yet others were not comfortable to participate.

I recall an instance where a woman was hosting a live chat on Instagram, sharing about her experiences with depression. I joined the live chat until it ended, and then messaged a request for her to participate in my study. I shared an information sheet about the study with her, and waited for her response. She responded that she was uncomfortable with doing the interview. At the time, I could not comprehend that statement because she was already sharing her experiences online, doing live chats, and answering people's comments and questions. However, I had to understand that from her perspective, her live chats and sharing about her experiences with depression, gave her a sense of control over her social media presence and her experiences of depression. She had control over who to answer and which issues to address-and that was part of her coping with her depression. In my doctoral research, participants highlighted that the most important part of using social media as a coping mechanism is the ability to control your platform to ensure that it does not affect you negatively (Kgatla, 2023). The sense of control acquired on social media helps them to cope with depression because they are actively involved in managing their depression. Thus, I had to understand that her doing live chats on social media conferred more control than being interviewed by someone else; and she was not ready to relinquish control. I had to come to that realisation in order to understand things from her perspective.

I also posted an information sheet about the study on my social media pages (including Instagram, Facebook, X, and TikTok) and requested people to message me if they were interested in participating. Through that strategy, I was mostly approached by people who did not actively post about their experiences with depression, but did use social media as a coping tool. For instance, they followed certain people or pages that posted about mental health or depression with suggestions on coping strategies to manage the condition. Because they could get information about mental health from those pages, social media was the source of information that helped them to cope with depression. However, the pages and people they followed were not exclusively those that posted about mental health or depression. They also followed pages that posted any type of content that helped lift their mood. These included people who posted funny skits, cooking, or makeup videos. All these helped participants to feel at ease and comfortable at times when they felt depressed, thus helping them to cope with depression.

Some of the potential participants I approached were concerned about comfortability, and I received various rejections for participation because they were not ready. Another instance occurred when a participant approached me after viewing my information sheet on one of my social media platforms-she was interested in participating. We then had a conversation and I briefed her more on the study, what was expected, and how I would be collecting my data. (This was information that was on the information sheet, nevertheless, I always had a brief discussion with potential participants before they consented in order to ensure that they were ready to participate.) We set a date for the interview and on the morning of the interview, she messaged to say she was not comfortable doing the interview, and we postponed it. We postponed twice before she completely withdrew from the interview. She withdrew because she was not comfortable and was not ready to converse about her experiences with depression. She had thought she was ready, which is why she approached me, but as the interview date approached, she always panicked and became anxious, and therefore withdrew from the study.

As a researcher, I had to navigate such incidents and understand that when researching such a sensitive topic, there might be people who wished to participate but were not mentally or emotionally ready to take part. This was especially true when they learned that their content from social media would be used in the study. For them, it was like the invasion of a safe space they had created for themselves where they could share their personal experiences, and they were not ready to be public in a more formal sense. They were aware that people could view their content, but that was at a distance because they did not always see who viewed the content or when there was someone actually viewing their content. Knowing that someone was particularly viewing their content could present as a challenge for them and make it feel as if it were an invasion of the safe space they had created.

This multifaceted recruitment strategy enriched the diversity of perspectives within the study cohort, capturing the various ways in which social media serves as a coping tool for depression. Overall, the approach to participant recruitment and engagement highlighted the importance of flexibility, empathy, and adaptability in qualitative research, particularly in the context of sensitive topics such as mental health. Prioritising participants' agency and experiences not only enriched the depth of the study findings but also demonstrated a profound commitment to ethical research practices.

 

Influencers vs. Nobodies

When I started searching for potential participants on social media, I also searched for known social media influencers and celebrities who create mental health awareness. I messaged a few of them but my messages were left on read. This means that the receiver opened the message, possibly read it, and then decided not to respond. It made me realise two things. Firstly, you may have the same goals as other people, but who you are matters on social media. For people to work or collaborate with you, or even respond to you, you need to be well known as well. I am aware that people are not obliged to answer their messages on social media however, I assumed that because my study aligned with one of their interests (mental health activism, creating awareness), it would at least grab their attention.

The second thing I learned is that followers are very important on social media, regardless of what you are trying to accomplish. As stated above, having the same goal does not matter; what matters, is who you are to people. That said, I am not very popular on social media, meaning I do not have a lot of followers. When I approached those well-known personalities, my messages were ignored because they would not benefit from participating in the study. Because they are more famous and well known than I am, from their perspective, I would benefit more than they would. What matters, is social media presence and if you cannot offer something different or more than what they already have, people do not entertain you. This is why most of the people I interviewed, even though they were mental health activists, were not yet well known or social media influencers. They were still at a beginner level of influencing and activism.

After realising the importance of having followers on social media, in order to find potential participants, I outsourced the activity to one of my friends who has more followers than I do. This friend had more than 10,000 followers on X and upwards of 5,000 on their other social media platforms. Using their profile to post my information sheet, repost my posts on the study, and reshare my status on the study, helped with reaching more potential participants. This cemented my initial thought that having a lot of followers helps in reaching certain people because some of those who then reached out to participate had more than 10,000 social media followers.

Now to consider the interviewing of influencers and nobodies (people who are not well known or famous). One of the ethical requirements of research is confidentiality when the researcher needs to conceal participants' identities. Thus, interviewing participants who are well known and famous presents a challenge for the researcher to conceal their identity because followers or fans are very good at identifying celebrities or their beloved social media influencers. As I have said, social media influencers or celebrities usually collaborate to brand their names or elevate their name and therefore, might not opt for using a pseudonym or hiding their identity. So, in my study, participants who were mental health activists preferred me to use their real identity and name. They were okay with revealing their identity because they were trying to create mental health awareness and thus saw fit to be associated with the study (Kgatla, 2023). Nevertheless, as a researcher, I chose to conceal their identity by giving them pseudonyms and blurring their faces in any content I used from their social media pages.

Therefore, as part of the requirement to conceal participants' identities when using data from social media platforms, does it negate or overrule the requirement of concealing identity if the participant prefers using their real identity? This made me question whether the ethical requirement of confidentiality should be expanded or changed to accommodate research done on well-known personalities on social media. Do you still need to conceal the identity of a well-known participant? Here, I mean when you use content from their social media pages because if it is just an interview, you can easily conceal their identity. Furthermore, if participants are well known, it is highly plausible that their fans will be able to decode the images or videos because you cannot blur out the whole picture to conceal their identity. So, is confidentially then disregarded automatically?

Another issue was that I had a video from one of the participants in the study that I wanted to reference. The video was already posted on social media by the participant, and everyone who participated in the video had consented to it being public on social media. I therefore only inserted a link to the video in my paper, which ensured the participants' anonymity. In other words, I posted the video onto my social media platforms, and then used the link on my TikTok account in the paper. When someone clicks the link, it takes them to the video, however, it is in my profile and not in the participant's profile. I am sharing this incident because of my question earlier about confidentially. Can confidentiality be disregarded when using content from social media given that the link takes you to my profile and from there, you could decode who is who because the people in the video are not blurred out? Earlier, I discussed consent and respecting participants' privacy. Whether the accounts are private, public, or have a disclaimer, researchers still need to treat it as invading others' privacy if they do not ask for consent. In this instance, consent was given by the people in the video to the participant to post it online, and I received consent from that participant to use anything from their social media account, which means that whoever reads the study has access to the link, the video, and everyone who participated in the video. The question is, is the consent from the participant enough to use that video? Or do all people in the video need to be consulted by the researcher regardless of their initial consent to the study participant? The issue of confidentiality is very complex when using data such as images and pictures from social media pages because even though the pictures are on a public platform, they are still on a personal account. It becomes even more difficult when researching well-known people because it can be challenging to ensure confidentiality. Researchers can try to minimise exposure of participants' identities but if they are decoded, does it affect the study? The researcher would have complied with the ethical requirement of concealing participants' identity.

This led to another issue. Given that depression is not always acknowledged as an illness, especially in the Black community (Kgatla, 2019), if people were able to identify some participants who were then cyberbullied because they had participated in the study, how does a researcher navigate that challenge to help their participants? Or does their responsibility as researcher end with the interview and briefing session?

Interviewing well-known people presents various challenges versus interviewing nobodies because it is easier to comply with every ethical requirement with the latter. When using content from social media, it is very difficult to conceal the source of an image or video. It is easy for people to decode images and assume that they know the person; and it is easy to spread misinformation, which can go viral. Where does the role of the researcher end? Is it when the interview ends? Is it after the briefing session ends? Or, are you still responsible for your participants even after participating in your study? Is it a lifetime role as long as your study can be accessed by the public? Are you responsible for the after-effects of the study on participants?

The issue of post-study responsibilities and participant well-being remains complex. As researchers, our ethical obligations extend beyond the confines of data collection and analysis. We must consider the potential impact of our research on participants, especially in sensitive topics like mental health, where stigma and discrimination are prevalent. Evan though we may not have direct control over participants' experiences after their involvement in the study, we still bear a moral responsibility to mitigate any harm that may arise, and provide support where possible. Ultimately, the role of the researcher extends beyond the confines of academic inquiry to encompass a broader ethical commitment to participant well-being and integrity. By grappling with these ethical dilemmas and seeking proactive solutions, we demonstrate a dedication to upholding the highest standards of ethical conduct in research.

 

Protecting Participants: Are We Allowed To Criticise?

I received an interesting comment from one of the reviewers of a paper that is in the process of being published. In my doctoral thesis, I made a researcher's note in which I analysed the findings from a different perspective, different from the objectives of the study. It was a point made by one of the participants. The study focused on how and why young people use social media to cope with depression and for some of them, a part of creating mental health illness awareness (which also helps them to cope) is displaying depressive episodes. These include times when they are feeling depressed-they film themselves crying or overwhelmed by emotion and post these on social media. The idea is to show people that depression is real, and that it can also affect Black people. This was a response to the criticism and stereotypes that refuse to acknowledge the authenticity of depression, and that it is a "White people" disease. In the said paper, I argued that those people are trying to create mental health awareness, and how they portray or display it, is not misleading young people. There is literature that indicates that people question the authenticity of displaying emotions such as crying on social media because the idea of a person crying and deciding to record themselves does not seem real to some people (Dunn, 2017).

The reviewer's comment that left me shocked and astonished was that "participants might feel offended by your interpretation of the findings." It made me think about protecting participants and ensuring that they are not harmed. This includes ensuring that their stories and experiences are portrayed in a positive light or, at the very least, from their perception and truth. I realised that I had done that in my findings- focused on the experiences of participants from their perspectives and provided a detailed analysis of their perception, understanding, and truth. However, in doing that I also highlighted things that were different from their perceptions, or understanding. In a nutshell, I was being critical. That was questioned by the reviewer because it might make the participants feel uncomfortable or offended.

I understand that as researchers, we need to protect our participants and ensure that they are not harmed and violated but adding to that, we also need to be able to be critical and understand that we not always going to share their perceptions and ideas. As researchers, we report and investigate the truth of what our participants have said and present their perceptions of the investigated phenomena-however, still allowing space and room for criticism, different perceptions, and ideas. That is not to say the participants' perceptions are not true, authentic, or valid. As sociologists and researchers, we understand that there are various ways to interpret a phenomenon, which does not imply that other interpretations are inferior or superior. Sociology allows you to see or perceive things from multiple eyes and not be stagnant in one way of thinking or understanding. So, as researchers, it is our duty to protect our participants however, does that mean we are not allowed to question them? Or be critical or offer a different perspective? Are we only allowed to report what was said without analysing it from various perspectives? The problem is not being critical or offering a different perspective; the problem arises when participants share their perceptions, and we change that to suit our own interpretation and perceptions. The acceptable way is to report their perception, their truths, and stories whilst also maintaining our sense of agency.

The reviewer's comment raises a crucial debate about the balance between protecting participants and maintaining critical analysis in research. While it is imperative to ensure that participants' experiences are portrayed accurately and respectfully, it is also essential to engage in critical reflection and analysis of the data. Research inherently involves interpreting and analysing participants' narratives within the broader context of the study. This does not diminish the validity of participants' perspectives but enriches the discussion by offering different interpretations and insights. As researchers, it is our responsibility to navigate this balance delicately, acknowledging participants' truths while also exercising academic rigour and critical inquiry. The key lies in transparently presenting participants' perspectives while offering nuanced analysis and interpretations, fostering a dialogue that respects both the participants' voices and the academic integrity of the research.

 

Researcher vs. Participants: She is Watching

When conducting data mining on participants' social media pages, I analysed posts and statuses already shared by them. Specifically, I focused on content from 2021 to 2022, although most of the data was from earlier periods. This led me to wonder about the potential impact on participants' behaviour if they were aware that a researcher was monitoring their social media activity. If potential participants know a researcher is viewing their content, would it influence the quality and authenticity of their posts? Would they modify their behaviour to create a certain impression, thereby aligning with the findings of Knudson et al. (2008) that people on social media may then perform rather than show their authentic selves?

My study employed Goffman's (1959) theoretical framework of dramaturgy in the context of online interactions. According to Goffman, individuals in the "front stage" are performing for an audience, and this can be applied to social media where young people may be performing for their audience/followers (Goffman, 1959; Knudson et al., 2008). They might not show their real selves, emotions, or feelings, aiming instead to be accepted by their audience. Therefore, applying this logic, if participants are aware that a researcher/audience is observing them, will they also put on a performance to present an image they believe the researcher would accept? For instance, mental health activists online might hope for engagement with their content. However, if they know a researcher is watching for a particular reason, could this awareness skew the data collected? This raises questions about the integrity and quality of data mined from social media when participants are aware of being observed. Can researchers still obtain effective and reliable data under these conditions? Alternatively, would it be ethical for a researcher to conduct data mining without participants' initial knowledge, subsequently seeking their consent to continue using their content? This approach may help preserve the authenticity of the data while addressing ethical considerations.

Balancing the need for authentic data with ethical considerations remains a complex challenge in social media research. While informed consent is crucial, researchers must be aware of its potential impact on participants' behaviour. Employing strategies to mitigate this impact, adhering to ethical guidelines, and considering the context of the research are essential to conducting effective and responsible social media research. Ongoing dialogue about these issues is vital as the landscape of social media and digital ethics continues to evolve.

 

Conclusion

Reflecting on the data collection journey for my doctoral study on social media, it is clear that although digital spaces offer unprecedented opportunities for gathering rich, contextual data, they also present unique ethical challenges that must be navigated with care. The process of mining data from social media and conducting interviews revealed complexities in informed consent, participant comfort, confidentiality, and the potential for performance bias when participants are aware they are being observed. The first major challenge was securing informed consent in a way that aligns with traditional ethical standards while accommodating the nuances of social media. In traditional research settings, consent is typically obtained through signed documents after participants are fully briefed on the study. However, in the digital realm where content can be public, private, or conditionally accessible, the lines of consent become blurred. The ethical principle of respecting participants' autonomy and privacy necessitates explicit consent, regardless of the public or private status of their social media profiles. This ensures that participants are aware of how their data will be used, and feel comfortable and respected throughout the research process. Ensuring participants' comfort is paramount, particularly when dealing with sensitive topics like depression. The reluctance of some individuals to participate, despite their public advocacy for mental health, underscores the importance of respecting personal boundaries and the need for researchers to approach such topics with empathy and sensitivity.

Confidentiality remains a cornerstone of ethical research, yet it is especially challenging when dealing with public figures or social media influencers whose identities are easily recognisable. The ethical dilemma of whether to conceal identities when participants are well known highlights the need for nuanced guidelines that balance transparency with respect for privacy. Additionally, the act of linking social media content, even with participant consent, raises questions about the extent to which researchers can ensure anonymity. This complexity is further compounded when considering the digital footprint left by participants, which can potentially expose them to unwanted attention or cyberbullying.

The question of the researcher's ongoing responsibility to participants post study, remains an open ethical issue. The potential for participants to experience negative consequences as a result of their involvement in the research, particularly in the public domain of social media, suggests that researchers may have a duty to offer support or intervention if adverse effects arise. This extends the ethical responsibility of researchers beyond the confines of the study, emphasising the need for a compassionate and participant-centred approach to research.

The tension between protecting participants and maintaining the critical integrity of research also emerged as a key issue. Although it is crucial to report participants' experiences faithfully, researchers must also engage critically with the data to provide a thorough analysis. This balance requires sensitivity to how participants' narratives are presented, ensuring that their perspectives are honoured without compromising the analytical depth of the study. As sociologists, it is essential to navigate these dual responsibilities, ensuring that participants' voices are authentically represented while also contributing to broader scholarly debates.

Finally, the issue of participants potentially altering their behaviour if aware of being observed by a researcher is another significant concern. According to Goffman's (1959) dramaturgical theory, individuals perform differently when they know they are being watched. This could impact the authenticity of the data collected because participants might tailor their online behaviour to fit what they perceive to be the researcher's expectations. This performance bias poses a threat to the integrity of the research because the data may not accurately reflect participants' genuine experiences and behaviours. To mitigate this, researchers must carefully consider the timing and method of obtaining consent, potentially opting for retrospective consent to preserve the naturalistic quality of the data.

In conclusion, the journey of collecting data from social media for my doctoral study highlighted the evolving landscape of ethical research practices in the digital age. It underscored the importance of maintaining participant respect, consent, and confidentiality while navigating the unique challenges posed by digital data collection. By critically reflecting on these experiences, researchers can advocate for ethical standards that uphold participant well-being and research integrity, ensuring that the benefits of digital research are realised without compromising ethical principles.

 

References

British Educational Research Association. (2018). Ethical guidelines for educational research (4th ed.). London.

Dunn, E. (2017). Blue is the new black: How popular culture is romanticizing mental illness [Unpublished honour's thesis]. Texas State University.         [ Links ]

Goffman, E. (1959). The presentation of self in everyday Life. Random House.

Hashmi, A. S., & Ahmad, T. (2016). Big data mining: Tools & algorithms. International Journal of Recent Contributions from Engineering, Science & IT (iJES), 4(1), 36-40. http://dx.doi.org/10.3991/ijes.v4i1.5350        [ Links ]

Kgatla, B. L. (2019). How do Black South African youth understand and cope with depression? [Master's dissertation, University of Johannesburg]. UJContent. https://hdl.handle.net/10210/421238        [ Links ]

Kgatla, B. L. (2023). How and why do young people use social media to cope with depression? [Doctoral dissertation, University of Johannesburg]. UJContent. https://hdl.handle.net/10210/511350        [ Links ]

Knudson, M., Svanoe, S., & Tappe, M. (2008). The social construction of depression at St. Olaf College (Ethnographic Research Methods 373). St. Olaf College. https://tinyurl.com/3ecud89c

Mirza, H., Bellalem, F., & Mirza, C. (2023). Ethical considerations in qualitative research: Summary guidelines for novice social science researchers. Social Studies and Research Journal, 11(1), 441-449. https://www.researchgate.net/publication/370838199_Ethical_Considerations_in_ Qualitative_Research_Summary_Guidelines_for_Novice_Social_Science_Researchers        [ Links ]