The Murky Ethics of Visual Qualitative Methods: Picturing a Clear Path Forward

As a visual qualitative researcher, I believe in the value of methods that include images, such as photos, in the production and interpretation of knowledge (Glaw, Inder, Kable, & Hazelton, 2017). Photovoice in particular—a participatory visual qualitative method in which people take pictures to identify, discuss, and address their challenges (Wang & Burris, 1997)—changed my own observational lens and allowed participant-generated ideas to emerge in my inquiries. As a young HIV researcher, I became discouraged by my own well-intentioned work. I feared that the questions I was asking women about their HIV-related risk behaviors were demoralizing them by assuming solutions (e.g., condom use) that were incompatible with their problems (e.g., relationship violence). Photovoice gave me another perspective. Ten years later, I still use the method to integrate the visions of people living with HIV into the interventions and policies designed to help them (e.g., Teti, French, Bonney, & Lightfoot, 2015).

Many other qualitative researchers have also described the benefits of Photovoice (e.g., Carlson, Engebretson, & Chamberlain, 2006; Chilton, Rabinowich, Council, & Breaux, 2009; Jurkowski, 2008; Wang, Morrel-Samuels, Hutchison, Bell, & Pestronk, 2004). Photos facilitate expression across varied levels of literacy, capture experiences difficult to articulate through language alone, and inspire creativity and solutions among participants who have often previously lacked opportunities to discuss their experiences (Wang, 1999). As a participatory action research (PAR) method, Photovoice privileges the well-being of participants and the translation of research to action as much as the discovery of information (Cornwall & Jewkes, 1995). Researchers and participants are knowledgeable partners in Photovoice and collaborate to enact research, analyses, and solutions. Participant knowledge is located specifically in the photos they take, interpret, and share (Evans-Agnew & Rosemberg, 2016). Thus, other potential benefits of Photovoice are the dismantling of negative social stereotypes and the ability to educate and influence policy and decision makers via photos that act as communication bridges (Foster-Fishman, Nowell, Deacon, Nievar, & McCann, 2005).

As the use of visual methods and Photovoice in particular has grown, critics have rightfully pointed out, however, that Photovoice researchers tend to present a romanticized view of participatory photography (Prins, 2010). Similarly, many visual researchers, myself included, have warned of the ethical challenges of using images. I read, often nervously, the visual ethics literature, asking myself whether and how I contribute to the problems and solutions. Recently, my concerns about the growing list of ethical dilemmas were leaving me overwhelmed and inactive, versus responsive. Although I was reading numerous articles about the ethics of visual methods, it was common to see statements that indicated few people were talking about ethics or that we needed more research to better understand ethics. I agreed with Boxall and Ralph’s (2009) warnings that if we, as qualitative visual researchers failed to grapple through difficult ethical scenarios, the use of creative visual methods could diminish, in a world where the public’s use of photos to create and make sense of information is increasing rapidly. In this editorial, I propose that the existing literature on visual ethics is robust and abundant, versus lacking, and that we should act on this knowledge in at least three key areas, to better inform and use institutional review boards (IRBs), choose the right visual methods to protect participants, and establish ways to stay informed on ever-evolving visual ethics challenges.

The Current Visual Methods Ethics Knowledge Base is Significant

Despite claims that information on visual ethics is lacking (e.g., Evans-Agnew & Rosemberg, 2016), there are dozens of articles that address ethics and provide visual ethics guidelines (e.g., Cox et al., 2014; Papademas & International Visual Sociology Association, 2009). This list is brief, and nonexhaustive, but existing ethical analyses address:

using visual methods with communities who experience multiple forms of oppressions and/or who may need special protections for visual research (e.g., refugees, sexual minorities, people with HIV, people with disabilities, youth, and substance users; e.g., Boxall & Ralph, 2009; Drew, Duncan, & Sawyer, 2010; Pittaway, Bartolomei, & Hugman, 2010; Teti, Murray, Johnson, & Binson, 2012),

potential for visuals to compromise people’s safety via increased visibility (e.g., Black, Alun, Dalis, & Chambers, 2018; Harley, 2012; Holtby, Klein, Cook, & Travers, 2015; Joanou, 2009),

risks posed by online dissemination (e.g., uncontrolled, permanent) of photos (e.g., Creighton et al., 2018),

rights of those photographed by participants (e.g., Teti et al., 2012; Wang & Redwood-Jones, 2001),

challenges of using IRBs in visual research (e.g., Flicker, Travers, Guta, McDonald, & Meagher, 2007; Khanlou & Peter, 2005; Mok, Flora, & Tarr, 2015),

problematic misinterpretation of images by the public including reinforcement of negative stereotypes about marginalized groups (e.g., Langmann & David, 2014; Quaylan, 2012),

lack of social justice or social change via PAR visual methods like Photovoice that promise those outcomes or undue pressure on participants to make broader changes happen (e.g., Evans-Agnew & Rosemberg, 2016; Johnston, 2016; Liebenberg, 2018; Mitchell, 2011; Prins, 2010),

too much researcher control over analysis and interpretation of images (e.g., Guillemin & Drew, 2010; Liebenberg, 2018), and

confusion over who owns the images (e.g., Gubrium, Hill, & Flicker, 2010; Guillemin & Drew, 2010).

Below, I choose three areas from this list and provide recommendations on moving from knowledge to action to help us think about how to continuously use and improve visual methods despite ethical challenges.

Most IRBs Are Not Able to Provide All Needed Guidance on Visual Research

IRBs are well-meaning and use the principles of autonomy, nonmaleficence, beneficence, and justice to protect research participants. The relevance and application of these principles to visual research and PAR research overall is not well understood, especially since IRBs generally assess risks to individuals and not communities (Flicker, Travers, Guta, McDonald, & Meagher, 2007). For example, consents or protocols may meet IRB approval even if they fail to address how particular images may reinforce negative stereotypes about a community or how PAR visual methods like Photovoice include the community in all stages of the project, enhance the well-being of a community in addition to the production of new knowledge, or influence social change in a community. In addition, visual methods are commonly employed with communities that are systemically marginalized with limited opportunities to inform the policies and resource allocation related to their contexts. Yet, a consent’s usefulness hinges on the assumption that participants have power, however, at least to speak up and take action if they become uncomfortable with the research process. Such entitlement may not exist among groups that are recruited specifically because they “lack voice.”

If the Research Question Drives the Choice of Methods, Ethics Wins

Qualitative research students frequently ask me if they should use qualitative methods. I tell them that the research question and project goal should determine method. Not all questions are suitable to visual methods or PAR visual methods like Photovoice. As Liebenberg (2018) points out, Photovoice is intended to give participants collective inspiration, voice, and plans for action. If we do not believe we can facilitate those outcomes, photo-elicitation may be more appropriate and ethical than Photovoice. Holtby, Klein, Cook, and Travers (2015) also remind us that even the correct use of a methodology will not erase the historical or present harms that exist in communities, that researchers need to understand their work amid broader social processes, and that researchers should work with participants to identify and address concerns. If we choose Photovoice, therefore, we should decide how to incorporate the community to enhance the ethics of the project. Participants can engage in complex ethical discussions about which methods are most appropriate, how their visual stories may be interpreted by others, the pitfalls of Internet photo dissemination, and how to ensure that their photos do not reinforce negative stereotypes, for example. Pittaway, Bartolomei, and Hugman (2010) described a visual research project with refugees who asked to be trained to tell their own stories without researcher involvement because previous interactions with researchers traumatized participants and left them without support. Participants can also advise about safe forms of social action. In a Photovoice project I did with brain injury patients, they did not want to talk to decision makers about their photos for fear it would be too inconvenient and overwhelming. They did, however, want the research team to present their images at a local brain injury conference, to help clinicians understand the role of their pre-injury lives in their injury management goals (Teti, Hart, & Reid-Arndt, 2015).

Ethics Are Fluid and Constantly Evolving

Holtby, Klein, Cook, and Travers (2015) points out that the question about how to conduct Photovoice ethically is unanswerable. Ethics are constantly evolving, and a one-size-fits-all approach will not work. Camera technologies change daily. Smartphones enable everyone to take and share photos widely. Internet platforms like Instagram and Facebook allow pictures to be shared, copied, and reused for years (Creighton et al., 2018). Recent commotion over FaceApp, which allowed Facebook users to see what they would look like when the aged, caused alarm when a Tweeter warned that the app could be taking photos from users’ phones and uploading them to servers without user permission or knowledge. Although shown to be just a scare, the incident does exemplify the ever-changing concerns about digital images (Brewster, 2019). Also, ethical issues arise based on study sample and context. For example, while working with Salvadorian adults, Prins (2010) found that fear and mistrust in the community and country created safety issues for photographers who were assumed to be stealing or causing trouble while taking photos. In one of my projects, I learned through discussions with the participants that one woman wanted to take “Wanted” style photos of the man who infected her with HIV and publically post them. These examples, and many others, indicate that new and specific ethics issues are always emerging.

Recommendations

The seriousness of the ethical challenges of visual research notwithstanding, we do have the knowledge base and ability to move forward in the following ways to continue advancing both visual research and visual research ethics. Examples of key action steps include the follwing:

Perform a scoping review of existing ethics analyses summarizing findings to date.

Develop additional and/or disseminate (e.g., Cox et al., 2014) protocols to appropriately address ethics of visual methods.

Inform IRBs of the protocols that we develop and encourage IRBs to be trained on visual research and PAR visual research methods. We can do this by becoming formal members of our institutional IRBs or by meeting with IRBs to discuss our research protocols.

Advocate for the inclusion of community members on IRBs, so that they can help guide discussions about community needs and safety concerns for visual PAR projects.

Learn about multiple visual methods and their differences; use Photovoice only if social action is appropriate and possible.

Identify and plan for ways to include participants in complex ethics conversations.

Train our research teams about ethics and encourage them to share concerns as and if they arise throughout the course of projects. We can also include ethics discussions, problem-solving, or trainings in ongoing research project meetings.

Write about and disseminate these experiences for others versus only writing about our research outcomes and successes.

Images can be used in ways that facilitate rich research information and participant well-being and empowerment. They can also be tools of surveillance, punishment, or voyeurism. It is important to hold all of these options as possibilities as we continue to discuss visual ethics. The body of information about the ethics of visual methods is growing quickly, however. The time has come to assess what we know and move forward by addressing the gaps in our IRBs, improving our knowledge and application about visual method options, and developing better protocols to stay abreast of ongoing ethical challenges.