Picture This . . . Safety,Dignity,and Voice—Ethical Research With Children

Assent or Willingness to Participate

In order to uphold respect for a child participant’s dignity the child’s assent or willingness to participate is sought in addition to informed consent by the parents (Dockett, Einarsdottir, & Perry, 2009; Dockett & Perry, 2011; Mishna, Antle, & Regehr, 2004). Assent is sought after written consent from a parent or legal guardian is obtained. This occurs by explaining to the child the purpose of the research project and the child’s role in the process (Dockett et al., 2009; Dockett & Perry, 2011; Mishna et al., 2004). A number of scholars point out that the purpose of the study and the process of assent is to be explained in a way that a child can understand and that using pictures and/or age-appropriate language is helpful to foster the child’s comprehension (Helseth & Slettebø, 2004; Lambert & Glacken, 2011; Mishna et al., 2004). Acquiring assent reinforces the child’s right to refuse to participate or to dissent, even though consent by the parent has already been given (Dockett & Perry, 2011; Mishna et al., 2004).

Most IRBs will require a written assent form to be used in studies involving school-aged child participants. Some of the questions that informed the design of our assent process are as follows: Why are you [the research participant] here? Why are we [the researchers] doing this study? Will there be any tests? What will happen to you? Will this study help you? What will happen to your pictures? What if you have any questions? Do you have to be in the study? It was helpful to read the assent form with the child, or when feasible to have the child read the question and the researcher read the answer back to them. Creating opportunities for children to ask many questions is another consideration in obtaining assent (Lambert & Glacken, 2011). We discovered that when working with children with disabilities, this frequently required additional time and patience especially when children were using augmentative communication devices and suggest that this is worthy of consideration in the research design process. Most important, for purposes of assent, the researcher must be clear that the child is not required to participate in the research project if she or he does not want to, that the child is not expected to participate by the researchers or anyone else (including parents, teachers, etc.), and that the child will not face negative consequences if she or he withdraws from the study at any point which can be a child’s biggest fear Bruzzese & Fisher, 2003; Dockett & Perry, 2011; Lambert & Glacken, 2011).

Assent forms are also signed by the child, which can potentially be a fun activity, particularly if children enjoy printing their names or “signing autographs.” Helseth and Slettebø (2004) contend that written assent “can often give children a feeling of significance in the situation and empowers the feeling that their consent to participation really counts” (p. 303). However, in our experience, for some children the act of written consent may feel like a school task, and be time consuming, especially if printing is not easy for them (which was often the case when working with children with physical disabilities). For example, in our research a child with cerebral palsy wished to sign her name on the assent form. She was able to sign her own name; however, it was somewhat difficult for her. She was determined to complete the task to show what she could do, as this was something she was proud of. For this reason, it was important to allow her to do so; however, completing her signature took approximately 45 minutes, which was exhausting for the child and almost the length of the planned session. This not only took a lot of time, but a lot of the child’s energy and attention, which may have been better utilized for the actual interview. In such cases, it may be beneficial to work into the ethics protocol the option to obtain recorded verbal assent, along with assent by marking an “X” or placing a sticker on the signature box. Another possibility may be to ensure that the option for additional visits to complete data collection is built into the research design, should delays of this nature arise.

An additional caution is to consider whether children are signing their names simply to show they can, or to please their parents or the researcher, without truly understanding why they are signing and what it entails. For example, in our study some children became anxious to sign or print their names to show their skills with respect to this task or to get started quickly without going through the assent information and questions that were perceived as “boring.” Reflexivity in this regard may help the researcher stay attuned to the situation and ensure the assent process is enacted and revisited without moving on prematurely.

Obtaining assent can be easier said than done. It requires patience on the part of the researcher and sensitivity to ensure that the child truly understands what they are about to participate in and why, not only when completing the formalities of the assent form but throughout the research process. Although the completion of a formal assent document is in most cases required by the IRB, it does not mean assent is finite and should not be revisited throughout the research process (Dockett & Perry, 2011; Fine, Weis, Weseen, & Wong, 2000; Helseth & Slettebø, 2004). Cocks recommends that researchers make a conscious effort to remain vigilant to the child’s responses throughout the entire research process, becoming attuned to children’s ways of communicating in order to recognize moments when children are uncomfortable with their participation. Children can assent in different ways, including inviting you, as the adult/researcher, into their worlds (Cocks, 2006). In this study, the researcher began each session with time for “rapport building.” Spending time simply talking about the child’s day created opportunities for them to ask questions such as: “Why are you asking kids questions?” “What is your project for?” “When is your project going to be done?” Often in this time children invited the researcher to see their pets, bedrooms, poster collections, latest craft project, or computer game, which could be interpreted as accepting the presence of the researcher in their activities (Cocks, 2006).

Cocks (2006) contends:

It is the beliefs of the researcher that influence the ethical nature of a piece of work and impact the final product. Therefore, flowing through each of the core issues of the ethical framework, particularly through the application of “assent,” is the reflexivity of the researcher. (p. 261)

We propose that one reflexive practice of an ethical researcher is to ask yourself if you believe the child has understood and truly assented to participation in the study. This requires sensitivity, awareness, and acute observational skills. At times, the child’s expression or body language may reveal that she or he is uncomfortable or not sure of what is going on (Dockett et al., 2009). We observed that often children just want to move on to the “doing” of the research, the “fun stuff.” If you sense this might be the case, one option is to reiterate some of the information in a playful way, while also offering opportunities for children to ask questions and potentially stop participation if they no longer are interested. Consent and assent with children is better approached as a process versus a single event (Renold, Holland, Ross, & Hillman, 2008; Warin, 2011). Revisiting a child’s choice to participate at each stage of the research process is critical (Etherington, 2007; Warin, 2011). Such attention to the child during the research process requires ongoing researcher reflexivity and ethical mindfulness (Cocks, 2006; Etherington, 2007; Warin, 2011).

Informed Consent and Assent Using Visual Methods

When using visual methods with children, there is another level of informed consent required, and additional informed consent and assent procedures take place. Informed consent is also required from others who may appear in the pictures taken by the participants (Wang & Redwood-Jones, 2001). This can be challenging for the participants to manage, especially if a child wants to take a picture of another child, as informed consent must be obtained from the other child’s parents and assent must be obtained from the child appearing in the picture. For example, in our study the following questions were used to design the assent process for children: Why are you being asked to be in a picture? What will happen to you? What will happen to the pictures?

In addition to all consent documents, Wang and Redwood-Jones (2001) contend, it is beneficial to offer the option for participants to choose if they would like their pictures to appear in research publications, presentations, or other appropriate venues in order to protect their privacy and ensure their confidentiality. Although participants may consent and assent for their pictures to be displayed in different media arenas or texts, there are a number of ethical implications that arise surrounding representation of the child. These are discussed in more depth later in this article.

When approaching others who may be photographed by participants, the consent and assent procedures are out of the researchers’ hands. We found it necessary to provide really clear and detailed instructions; it was helpful to provide participants with a folder, clearly marking the documents and highlighting places for signatures. When it is out of the researchers’ hands, it is difficult to know how the procedures took place and if other participants were clearly informed; therefore, we suggest emphasizing the importance of the consent and assent information and procedures with participants and their legal guardians.

This raises another question about the demands of the research that some may view as ethical: Is this process too much to ask of participants? Certainly reflexivity about the pros and cons of the demands on participants of consent and assent, especially when others may be involved, is a consideration worthy of attention during the design of the research.

Issues of Disclosure

Given that children may not have the capacity to anticipate the information they may be asked to disclose, children may not understand the potential risks and benefits to participation in research. Visual methods can be particularly problematic in connection with anticipating potential risks associated with disclosure, as images allow the researcher access to dimensions of an individual’s experience, in more delicate and intimate ways than are possible when the researcher is physically present (Clark-Ibáñez, 2004). Given the potential of the images to reveal aspects of the individual’s experience beyond their conscious control, we found that researcher reflexivity was helpful in negotiating the potential risks of unintentional disclosure.Some scholars contend that these risks may be more pertinent in research with children (Mishna et al., 2004; Punch, 2002).

We suggest that researchers attempt to anticipate potential risks and explain them in a way that is meaningful to the child, recognizing that we can never anticipate them all. Reflexivity may increase researchers’ awareness of their own preconceived assumptions about children and childhood (Davis, 1998; Finlay & Gough, 2003) and heighten sensitivity to potentially harmful situations when anticipating ethical challenges and when actually faced with “ethically important moments” (Guillemin & Gillam, 2004, p. 262).

Based on the experience of interviewing children, we have found it best to let the child direct the interview to the extent possible, thus offering the child a chance to choose what she or he feels comfortable sharing. The PEI method is helpful, as children can choose the pictures they would like to discuss (or choose not to discuss) and share these in any order they wish (Dell Clark, 1999; Dockett et al., 2009). Letting the child direct the interview can also help the researcher be more sensitive to moments where the child may feel vulnerable after a response, influencing the researcher’s decisions whether or not to probe difficult topics. For example, an excerpt from one of the transcripts illustrates a moment where Shanon felt that the child began to reveal more than she initially intended. Shanon trusted her gut feeling and did not probe the issue, letting the child direct the next step.

In text alone, it may be hard to imagine the moment in context. However, in this moment there was definitely some tension and discomfort experienced by both the child and the researcher. This discomfort was observed in Sarah’s nonverbal cues; Sarah avoided eye contact, looked away from the researcher and focused her gaze on the ground, her voice became low and quiet, and she began to speak quickly with hesitation at the end of each sentence.

Following this event we (the researchers) engaged in reflexive dialogue to explore if Shanon’s decision not to probe further was the “right” thing to do. This was an “ethically important moment” for both researchers, as the tension between collecting important data and respecting the child’s body language regarding her discomfort with disclosure came to a point of tension. This was the first time the child had talked about her peers, which was something the researcher was looking to explore. The question of pursuing the collection of additional data was weighed in light of the ethical cost at hand. Glenn (2004) suggests that in many instances silence can be just as powerful as speech. We recognized that interpretive insight into Sarah’s experience could be gained by observing the silences, not necessarily requiring a probe for more data that could be “heard.” According to Macklin and Whiteford (2012), although informed consent meets the demands of IRBs, researchers may still face a moral conundrum in the field: “Clarity versus care; which to choose”? (p. 97). In this case, Shanon opted for care. This is not to say that care supersedes clarity in all cases, but we suggest that reflexivity on the part of the researcher may help to negotiate the moral conundrums researchers face.

In such situations the environment may play an important role. Mishna et al. (2004) suggest that creating an environment that is too comfortable might affect the ability of the child to protect him or herself, leading to the unintentional disclosure of information. Jokinen, Lappalainen, Meriläinen, and Pelkonen (2002) contend that over time the participant–researcher relationship can transform from stranger to friend, “making it easier to acquire knowledge from the informants’ point of view” (p. 166). This can be beneficial from a research point of view; however, if a child sees a researcher as a friend, she or he could in turn become increasingly vulnerable. Reflexivity may assist researchers in treading these very fine lines.

Power Imbalances

Issues of power and power imbalances in relationships with researchers are also of concern when conducting research with children (Phelan & Kinsella, 2011). Although a relationship that creates a false sense of “friendship” may raise ethical issues, so does a relationship that minimizes the child’s agency, creating another fine line that researchers may find it helpful to negotiate through reflexivity.

The way in which a researcher presents oneself, the language used, the clothes worn, the body language adopted, how one interacts with the child, and the context of the interview can each influence the balance of power in both negative and positive ways (Phelan & Kinsella, 2011). For example, dressing casually, using fun “child-friendly” language, and conducting interviews on the playroom floor may work to decrease power imbalances. Whereas wearing formal attire, conducting interviews at the dining room table where the child usually completes his or her homework, using language that the child may not understand or language that sounds like something a teacher might say may act to create power imbalances. Such approaches might favor the researcher as the powerful adult and the child as less powerful and expected to comply with whatever the adult asks or expects of them. The child’s perception of the adult researcher as unduly powerful is potentially dangerous. Ethical researchers are called to be reflexive about how to create conditions where children have agency and share power to the extent possible (Punch, 2002). We deliberated about how to present ourselves as researchers, not only on the first visit to the home but also during the following visits. Questions considered included (a) how the parents might perceive the researcher dressed professionally or casually, (b) how comfortable the parent would be inviting the researcher into their home based on first impressions, (c) how the children might perceive the researcher, (d) how important it was to appear approachable upon first impression from a child’s perspective, and (e) how to talk to both parents and children in ways they could understand, even if that meant explaining things two different ways. In the end, the choice was made to wear clothes that were plain, casual, and comfortable in order to facilitate getting down on the floor if the children wanted to work in different places. Each child was asked where she or he would like to work that day, and typical places like kitchen tables were not assumed. Interestingly, most children chose the kitchen table on the first session; however, second and third sessions were often conducted in different places like the playroom, bedroom, or a place that was less frequented by other family members (we assume for privacy).

One challenge is that children may fall into the familiar teacher–student role, where they feel they must perform to their best, be on their best behavior, and provide information that the “teacher” wants to hear (Burke, 2005; Cappello, 2005). Therefore, the establishment of a relationship that is distinct from the teacher–student relationship and one that elicits the child’s own perspectives is critical (Phelan & Kinsella, 2011). During this research, some children assumed that Shanon was meeting with them in their home as a “helper” (support staff) who would be coming to visit them on a regular basis like other care workers who come for respite or therapy. This positioned Shanon in a particular role with a significant degree of power, despite efforts to avoid a situation such as this.

Another issue related to power is the importance of ensuring that the child understands that he or she can choose to participate or withdraw at any time. Often children do not understand that they have a right to withdraw from the research project unless asked directly and reassured that this is okay (Mishna et al., 2004). For this reason it is beneficial to create a safe and reassuring environment from the beginning and to offer the child opportunities to withdraw in ways that she or he feels it is safe to do so. One strategy that we adopted was to begin each session with an overview of what we were going to do that day, followed by asking the child if that sounded okay, if she or he had any other questions, and if she or he still wanted to participate. In one instance, on the third session, one child participant asked if she had to have her pictures shown in magazines, books, and presentations (assent and consent was formally obtained in the first session prior to taking the pictures).

This afforded the opportunity to go over the assent form again and clearly let her know that she did not have to say yes, even if her mother did say yes. In this case the child decided she would rather not have her pictures published. Her mother overheard our conversation and came over to ask why, in a way that would be encouraging the child to change her mind. This provided an opportunity to reiterate that it was the child’s choice and that it was okay not to use the pictures. Warin (2011) encourages researchers working with children to be ethically mindful of the potential disparities between child and parent consent and assent, recognizing that often there can be a mismatch, explicit or implicit.

Nutbrown (2010) contends “children’s words, drawings, and images as well as the children themselves can become the objects of research if dynamics of power are not recognized, acknowledged, and addressed” (p. 7). Reflexivity has been noted as a vehicle for helping researchers to bring their awareness to power dynamics that might arise in the research process (Barker & Weller, 2003). This includes reflexivity about the power dynamics between parents and their children as these may also influence the research project with respect to data collection and fear of withdrawal (Barker & Weller, 2003). Balancing concerns about safety and protection of the child with issues of dignity and voice are an ongoing challenge for reflexive researchers working with children (Danby & Farrell, 2004).

The choice of research methods may also help to minimize power imbalances. An overarching assumption of the PEI method is that it has the potential to minimize power imbalances between the researcher (adult) and the researched (child) (Dell Clark, 1999; Phelan & Kinsella, 2011). Power may be shared with the child by asking the child to direct and take his or her own photographs making it possible for the child to have a sense of control over the research process (Dell Clark, 1999; Epstein, Stevens, McKeever, & Baruchel, 2006; Frohmann, 2005). However, despite the best intentions, this method can still be influenced by authority figures (i.e., parents or caregivers), which may limit the child’s control and level of participation in the research (Barker & Weller, 2003; Clark-Ibáñez, 2004). The study design and researcher’s decision regarding who will take the photographs inherently influences the distribution of power. For example, although it is possible to use researcher-produced photographs when conducting theory-driven research, one shortcoming is that this process reduces the child’s influence in the research process (Clark-Ibáñez, 2004). In more inductive research, photo elicitation “auto-driven” interviews (Dell Clark, 1999) allow participants to take their own pictures. This approach enables the child to “drive” the interview process so that what is discussed is more likely to be pertinent and relevant to the child and the child maintains a certain sense of shared control and power (Dell Clark, 1999). “Who” is generating the photographs and the rationale for anyone other than the child generating the photographs become significant considerations in photo elicitation research (Burke, 2005; Clark-Ibáñez, 2004; Dell Clark, 1999; Phelan & Kinsella, 2011). Such issues may have implications for determining whether or to what extent the “voice” of the child is represented and for contributing to contextualized interpretations of the data.

On one hand, the choice of methods, such as PEI, can help to balance power in the research process; however, one may argue that this too can be potentially harmful. Methods that are considered “fun,” novel, and more attuned to children’s competencies or interests may create an environment where the child is more comfortable or “at ease” with the adult researcher (Punch, 2002). This can be seen not only as an advantage but also as a danger in the sense that the child may become too comfortable, rendering them vulnerable to unintentional disclosure, among other risks. Punch suggests using both traditional and innovative research methods to balance and address both ethical and methodological issues associated with research with children.

Representations of the Child

Closely related to issues of power, the ethical implications of how the voices and images of children are represented in research are significant ethical issues. Wang and Redwood-Jones (2001) raise the issue of participants being “placed in a false light by images and by words” (p. 566). The ethical imperative is on the researcher to be sensitive to how participants are being represented in both dimensions. Attending to one’s own reflexivity in the writing and dissemination phases of research reporting, recognizing the ethical imperative to represent the child with dignity, and acknowledging that researcher interpretations and representations of a particular child are only one of many possible representations would seem to be ethically important.

When deciding whether to use photographs in publications or presentations, researchers might consider ways to hide the identity of the child, such as pixilation, blurring the face or image, or perhaps consider not to use the photographs but rather to use only text in representation of the findings (Nutbrown, 2010). This raises another ethical debate about voice—by deciding to blur faces or not use photographs, even though you may have consent and assent, might researchers be silencing children in attempts to protect them? Nutbrown challenges decisions to distort images of children or not to include images of children in research despite consent, suggesting that such decisions contribute to the “Othering” of children in research. Nutbrown wrestles with the notion of presenting distorted pictures to represent the experiences of participants and what the act of distortion might imply with respect to identity, voice, being true to participants, and integrity in research and as a researcher. She suggests that conducting research with an ethic of respect for persons, knowledge, democratic values, and quality of research might not be possible if researchers choose to manipulate images even though they have full consent to use them. This debate contributes to the representation conundrum. Barron (2000, cited in Nutbrown, 2010) cautions researchers to consider moments when children potentially become “vulnerable to representations that others impose on them” (p. 33).

In our work, we have wrestled with the tensions that have emerged surrounding decisions to use photographs in publications and presentations. In a recent publication we considered using a photograph and corresponding text to illustrate an example of photo elicitation and its ability to tap into social dimensions that may otherwise remain silent. The photograph depicted a very powerful image of disability, which could be interpreted as an example of resistance to dominant discourses of disability. However, it could also be interpreted to display a more negative view of a child’s experience. We considered the importance of including this representation, yet at the same time were unsure of how the child or parent might feel 5 or 10 years later, knowing that this picture was published in a book, where the child’s identity was clearly visible and potentially construed as vulnerable through the corresponding text. As we could not come to a decision we were comfortable with at the time, we decided not to use that particular picture as an example or to distort the image or add pixilation to hide the child’s identity. We are still wrestling with this tension, considering the arguments and counterarguments, drawing upon our own reflexivity, and engaging in reflexive dialogue to consider representation of this in future work. We recognize that the child has a right to her or his voice being heard (Lundy, 2007) and, however, that it takes time and reflexivity to consider how to represent her or his voice while attending to privacy, safety, and dignity.

In addition, with methods such as PEI the risks of invasion of privacy become amplified (Wang & Redwood-Jones, 2001). Invasion of privacy may occur when photographs of others are taken without their knowledge. Researchers can attempt to address this concern through the use of written consent forms and assent forms for participants to use with others appearing in their photographs. However, as previously discussed, these forms and procedures have limitations.

Representation that maintains the voice, safety, and dignity of the child is a fine line, and reflexivity on the part of the researcher may help to work through some of these ethical issues. Guillemin and Gillam (2004) suggest that reflexivity on the knowledge produced and how the knowledge is generated is also important. Asking questions such as “How do I know?” and “How do I know what I know?” (Hertz, 1997) may help researchers to better discern how to represent findings in an ethical manner. Considering how children or parents might feel about their photographs appearing in print 5 or 10 years from now and considering any nagging gut feelings lingering with the researcher may also assist in discerning issues of representation. Furthermore, researchers might consider what the images add to the discussion: Are they necessary? Do they supplement the text to provide a richer picture? Are they just an added novelty?

Using photographs and text together can be more revealing and identifying by nature. If photographic images and text are to be used together in a larger document such as a book or thesis dissertation, the implications of all of these details in one place may raise additional ethical concerns. Despite pseudonyms used, the clear use of photographs together with detailed text from interview transcripts can allow for identification of the child and has the potential to compromise the child’s privacy and confidentiality. This may be dangerous for the child, especially if what is represented is of a sensitive nature. Researcher reflexivity may assist in determining where, when, and how the data is used together.