Sand tray interviews: Developing a method to explore the grief and support needs of 4- to 8-year-old parentally bereaved children

Martin Lytje

OmSorg, Danish Cancer Society; Center for crisis psychology, University of Bergen

Carol Holliday

Faculty of Education, University of Cambridge

Abstract

This article presents the ‘sand tray interview’ as a method developed to help researchers interview young children on sensitive topics. Earlier studies have often refrained from interviewing young children on sensitive matters due to the difficulties of conducting such research. This article explores how sand tray interviews can serve as a safe and engaging way for young children to explore painful topics, such as parental bereavement. Based on our application of this technique to a sample of children (4–8 years old), we discuss the development of the method and its applications, strengths, and limitations.Bottom of Form

Practice points

1. Recognise that young children are able to express valuable views about their own experiences and support needs.

2. Acknowledge that using the right methods it is possible to interview young children on difficult topics, in an ethical sound way.

3. Highlight that sand trays have the potential to be a powerful interview tool.

4. Encourage future research projects to utilise creative approaches when designing methods aimed at young children.

Keywords

young children, grief, research methods, interviews

Introduction

The previous decade has seen a significant increase in the number of studies exploring children’s grief following bereavement (Gray et al, 2011; Bylund-Grenklo et al, 2016). However, substantial gaps remain in the literature. One particular challenge is that many published studies (eg Raveis, Siegel and Karus, 1999; Bylund-Grenklo et al, 2016) have neglected to include young children. This is probably due to the significant methodological and ethical difficulties associated with undertaking such studies (Morgan et al, 2002; Clark, 2010). Studies (Brent et al, 2009; Gray et al, 2011) that have included young children exhibit another shortcoming, as they have often treated children as a homogenous group and failed to consider the specific conditions that distinguish conducting research with young children from research with older children and adolescents. This is problematic, as researchers have noted that young children are generally less resilient, more prone to feel distressed compared with their older counterparts, and require methods specifically designed for their developmental level (Kitson et al, 1996; Punch, 2002; Clark, 2010). In studies on childhood bereavement, researchers (Slaughter, 2005; Hunter and Smith, 2008) have found that young children have an evolving understanding of death that does not stabilise or become similar to that of an adult until the age of 9 or 10. Before this point, children may not fully understand what death means. As an example, children aged six to seven often struggle to understand that people cannot return from death, while it is not until the age of seven to eight that they understand that a deceased person does not need to eat and is unable to feel cold (Slaughter, 2007).

The abovementioned research challenges mean that research has generated fragmentary and limited knowledge concerning the grief and support needs of young children. To address this gap, we have undertaken a study which aims to enhance our understanding of grief in children aged up to six years and explore their support needs by interviewing bereaved children. The discoveries we glean from these data are to be published in two articles, of which this is the first. Here, we explore the application of the ‘sand tray interview’ method with young children aged four to eight.

Sand trays, being boxes filled with sand, act as a frame for telling stories through the use of figurines. They have primarily been used in therapy, to help children talk about difficult experiences. While sand trays have occasionally been used in research, authors (eg Mannay et al, 2017; Rees et al, 2020) have seldom described how they adapted this method from a therapeutic tool to a research instrument. In this article, we therefore focus on presenting the framework we developed for the sand tray method while also debating its strengths and weaknesses. We developed this method by merging central ideas from sand-based therapy with the core concepts applied in semi-structured interviews.

This article limits itself to describing the creation of the method, evaluating its success, and discussing areas for improvement. The method was developed in order to provide an interview method for young children (4–8) that could help them open up about difficult experiences, in this case the loss of a parent.

Challenges in research with young children

As several authors have noted (eg Clark, 2010; Eckhoff, 2019), undertaking research with young children is challenging. Should researchers fail to consider how such research differs from conducting studies with older participants, such projects may risk being unethical from their inception (Clark, 2010). In the literature, it is well established that the younger a child is at the time of data generation, the more challenging it will be to produce data that is of high quality while also having been obtained ethically (Vaughn et al, 1996; Krueger, 2009). The developmental levels of children have a particular influence on determining when they can be meaningfully interviewed. Before the age of four, few children have the communicative vocabulary or social understanding required to partake in conversation-based interviews (eg Astington, 1993; Happé & Frith, 2014).

An issue that can make interviews with young children immensely challenging is that the method might lead to the activation of defence mechanisms. These are processes by which a child unconsciously activates strategies to defend against painful thoughts, feelings, or memories (Freud, 1937). According to Cullberg (2006), defence mechanisms help the bereaved slowly come to terms with the experienced loss, as opposed to being overwhelmed by the cumulative consequences of the new reality. Defence mechanisms include regression, denial, avoidance, and repression. Of these, through our own clinical and research practice, we have found avoidance to be particularly prevalent and difficult to circumvent when interviewing younger children. An example occurred when asking a difficult question during a data generation session with a nine-year-old girl:

Interviewer: So, you were missing your mother . . . how did that feel?

Girl (9): I . . . I . . . (Silence) . . . A ball once hit me on my head!

The above example could be an example of one of two potential reactions: It might have been a way for the child to avoid painful memories, or it may be indicative of an unconscious link between the pain of loss and that associated with being struck by a ball. Whereas adults may be able to prepare themselves for a difficult discussion, young children seldom have the capability or experience required to do so. As a consequence, when young children experience difficult emotions, these emotions may surprise them, which in turn can activate avoidance and thereby hinder any progress during an interview. However, should a child continually change the topic when painful memories and sensations arise during an interview focused on a loss experience, it can become impossible to conduct the interview while ensuring adherence to the ethical principles of non-maleficence and beneficence (Bond & Griffin, 2013).

With the above challenges in mind, we required a methodological approach that was both ethically sound and capable of addressing these challenges. As the existing literature offers few good solutions when it comes to overcoming defence mechanisms during interviews, we looked to the field of therapy for inspiration. Child psychotherapy employs the medium of play to understand children’s experiences. Play is understood as the natural language of young children, who are less able to express themselves verbally. It is considered to be an externalisation of a child’s inner world achieved through the use of symbols (Blake, 2011; Sunderland, 2015). Thus, play is meaningful and may be considered a means by which children can convey their experiences. Therapists use play to both engage children and understand their experiences. Of particular importance is the capacity of play to provide the sense of safety that metaphor provides when addressing difficult topics. Thoughts, feelings, and memories do not have to be experienced directly but can be indirectly expressed through play. For example, if playing with a doll when speaking with the child, the pain of the loss may be contained within the doll and not the child itself. In this approach, the child will not be overwhelmed by their feelings but can still find expression and healing. Defences are necessary to protect against something vulnerable, and they need to be respected, not challenged.

When the protection offered by metaphor is intact (ie it is not translated directly into everyday life), play is a safe and protective medium for the exploration of experience and emotions in therapy. As an example, a child might play out the dynamic of the shark and the swan and gain relief and comfort from the empathy of the therapist, whereas talking directly about an experience involving abuse from, for example, one’s uncle would be too difficult. The metaphor of the play is protective, and the therapist does not make explicit its underlying meaning, as doing so would be exposing and shameful for the child. Play can be a useful medium for the researcher as well as the therapist, as exemplified by Koller & San Juan (2015), who use play-based interviews to explore young children’s perspectives on inclusion. In evaluating their research, the authors conclude that a symbolic place approach that features enjoyable storytelling and does not require the child to have any special skill is desirable. Working with a sand tray, as described in this paper, meets these criteria.

Developing a sand tray-based interview method

Since their conception by Margaret Lowenfeld (1979) under the name of ‘wonderbox’, sand tray techniques have been adopted in therapy approaches from a range of perspectives. These approaches include humanistic (Armstrong, 2008), Gestalt (Stevens, 2004) integrative (Holliday, 2013), and Jungian sandplay (Turner, 2017b). Dora Kalf, a prominent Jungian, coined the term sandplay, and other theoretical modalities tend to use the term sand tray. Sand tray approaches are ethically sound therapeutic methods that can provide children with a means of dealing with mental trauma. These approaches seek to create a space in which unconscious worries can manifest (Armstrong, 2008; Turner, 2017a). Telling a story about a character using the sand tray approach arguably helps children process difficult emotions in a non-confrontational manner. The projection offered by a story allows children to convey painful experiences in a manner that is more gentle than directly discussing them. The tray itself thereby serves as a form of containment that supports and eases the processing of experience (Miller & Boe, 1990; Carey, 2006).

Sand tray therapy has the potential to provide a solution to the challenge of young children activating defence mechanisms during interviews. However, since it is a form of therapy, significant issues arise when applying the method for research interviews. A central challenge associated with the technique when applied for research purposes is that the therapist does not select the topic for discussion but instead allows the child to decide which story to tell. There is a general thesis within sand-based therapies (Armstrong, 2008; Turner, 2017a) that if a child has undergone a challenging experience, such as bereavement, they will always include this experience in the story being told. However, short of directly asking a child to narrate a story about a loss, a researcher will have no way of verifying whether a story is indeed related to such an experience. Additionally, if a child has had more than one painful experience, the story they tell might relate to any of these experiences and may not necessarily concern a loss.

To address these issues inherent in the sand therapy method, we decided to merge the technique with some of the core tenets of semi-structured interviews. Semi-structured interviews are an efficient means of gaining knowledge about a specific area of interest while leaving room for participants to determine the direction and essential topics within the interview themes (Galetta, 2013; Brinkmann & Kvale, 2014). The method allows the researcher to define a topic of investigation, whereas the participant is allowed to determine what is important within this area of exploration. To exploit these advantages during our interviews, we decided to merge the open world of the sand tray method with the tenets of semi-structured interviews. We have termed this method the ‘sand tray interview’.

Sand tray interviews

We believe that using a sand tray during interviews with bereaved children may help them discuss difficult feelings and avoid the activation of unconscious avoidance mechanisms. However, the traditional sand tray approach cannot be utilised in interviews without raising ethical concerns.

Due to their ability to help create a protective barrier between a child and an experience, sand trays create distance from the complicated feelings that may be associated with a particular event (Rae, 2015; Pearson & Wilson, 2019). The sand tray thereby becomes a safe environment in which a child can discuss traumatic experiences that they may be unable to address in a more direct and traditional conversation. It is a method that does not rely on the participant having strong verbal skills and has been characterised as minimally invasive (Dobretsova & Wiese, 2019). In addition, the method has the advantage that the interviewer and the participant can focus on the sand tray rather than each other, which reduces the amount of time they have to directly look at each other. In our experience, it is often easier to discuss a difficult experience when directing one’s gaze elsewhere rather than making eye contact with an interviewer.

All modern ethical research frameworks consider explaining the nature of a study and the reasons for an interview as ethical conduct (Bond & Griffin, 2013). Adhering to these guidelines leads to a peculiar situation: without alterations to the sand tray method, the interviewer would have to begin by stating the purpose of the interview and research project and then ask the child to tell a story of their own devising. This could lead to two potential issues: first, as mentioned previously, the researcher will probably be unable to verify whether the story has anything to do with a bereavement.

Second, telling a child that the interviewers wish to learn about how their family dealt with the loss of a parent and then asking the child to tell a ‘random’ story could be considered deceptive. While young children may not establish a link between the purpose of the visit and the assigned task, older participants are very likely to do so. It would also be somewhat disrespectful and against student voice principles were children to automatically be considered as incapable of expressing difficult feelings and desires for support (Fielding, 2001). The adopted method would consequently need to establish a balance between respecting children who wanted to share their experiences and providing them with the support required to do so more indirectly should their emotions prevent them from discussing their experiences directly.

To address the above concerns, we opted for an approach in which the child would start by telling their own story but could switch to discussing a fictive character when necessary. How this was done is detailed later in this article. Our approach has the advantages that the child is able to discuss their experiences and desire for support and that the interviewer can help to shift the focus of the interview to a fictional character when doing so seems appropriate. In practice, this related to how a friend in the same situation might feel or what a friend might find difficult or need.

The above approach seems to strike a balance between ethics on the one hand and the purpose and strengths of the sand tray interview method on the other. It is more sincere to let children decide whether they want to discuss themselves using the sand tray as a medium or to tell a story about a friend. However, it also makes for a method that requires highly trained interviewers who can determine when a story needs to shift from being centred on the child to being about someone else. With the method decided, the process can move on to the second stage: the interview itself.

Interview preparation

To conduct sand tray interviews, a container for the sand is needed. Recommendations regarding sand tray size vary somewhat within the literature (Homeyer & Sweeney, 2016; Turner, 2017a). We chose a middle ground and created wood trays measuring 71 cm x 50 cm x 8 cm. Following general recommendations in the literature, the interiors of the boxes were painted blue, as this helps children to depict rivers, oceans, and lakes while recounting their experiences using the sand trays. No sand tray is complete without a set of figurines. We assembled a collection that included approximately 60 figurines (including people, ghosts, coffins, wheelchairs, and animals) and 40 wooden building bricks.

To provide the best foundation for success, four experienced interviewers were chosen to conduct our interviews. All held university degrees in child support and had previous experience interacting with bereaved children. As per the recommendations of Cameron (2005), interviewers were provided with readings on the sand tray therapy method and were invited to attend a training course. The curriculum centred on offering practical training and helping the interviewers fine-tune the method for use during the interview sessions. Subsequently, the interviewers practised the technique, initially with adult volunteers and subsequently with children, with whom they discussed non-sensitive matters.

Participant recruitment

While the perspectives gained from this study will be addressed in a separate article, this section briefly summarises the sample of children who participated. Participants were recruited through the local offices and networks of the Danish Cancer Society, as well as Facebook groups. In both approaches, families received a letter explaining how the interviews would be conducted, criteria for participation, participant rights, and how the collected data would be used. Participants could indicate their interest in participating by responding either via mail or phone. If a family chose mail, the interviewer would subsequently phone them and discuss the project in more detail. To qualify for participation, children had to satisfy the following criteria:

1. be four to eight years old and have experienced the death of a parent at the age of three to six years

2. have attended a daycare institution at the time

3. of loss

4. not have experienced the loss due to an extreme event (eg terrorism, murder, war)

5. have experienced a loss between six months and four years before the interview.

Parents were asked to discuss the study with their children and ask them whether they would like to participate. As recommended by Flewitt (2005), this helps reduce potential power issues that may arise should children have to accept participation while both the researcher and parent are there and encouraging them to do so.

Interviews were undertaken as family visits where two interviewers visited each family and interviewed both the remaining parent(s) and the child. Cameron (2005) notes the importance of the interview scene being well-known to and secure for young children. We deemed that the children’s homes would be most likely to represent such secure environments. The sand tray was only used with the children.

Children who had experienced loss due to an extreme event were disqualified, as researchers (Kaltman & Bonanno, 2003; Nakajima et al, 2012) have argued that such experiences often necessitate different forms of support than do less extreme bereavements. That six months should have passed since the loss was determined because this meant that a family would probably have overcome the most immediate shock and have gained experience in receiving support following the loss. The limit was set to four years as bereavements that occurred beyond this length of time may have been challenging to recall in detail. Overall, 12 children were recruited for this study from across Denmark. This number was slightly lower than anticipated, but recruitment was hampered by the Covid-19 pandemic. Interviews were audio-recorded and lasted between 25 and 50 minutes, with the majority being around 35 minutes in length.

Ethical considerations and post-interview support

The ethical framework for good practice in counselling and psychotherapy (Bond &Griffin, 2013) was followed in this study. This required ensuring the safety of the children taking part, informing the families that participation was voluntary and that the participants could leave at any time, and checking the participants understood what it meant to give consent. Such explanations were provided in child-friendly language. The parents subsequently provided written consent on behalf of their children.

The study further followed the European General Data Protection Regulation (General Data Protection Regulation (GDPR) – Official Legal Text, undated). Being done by a non-governmental organisation, the study underwent an internal ethics review before it was undertaken. In Denmark, projects do not need to be approved by an external ethic committee if they will not involve the collection of biological material (National Scientific Committee, 2020).

Extensive post-interview measures were made available if concerns were raised about the wellbeing of the participant post-interview. These measures included psychological or therapeutic consultation or assistance from a social worker employed by the Danish Cancer Society. In cases where such efforts were established, there would be an additional follow-up phone call approximately 14 days after the session had concluded. Parents were also able to make use of such support.

The personal qualities and professional skills of the interviewers probably served to significantly mitigate the possibility of distress or harm. We argue that projects of this nature demand interviewers who can establish and maintain a research environment in which children feel safe enough to engage in the process. Doing so requires interviewers who can model being empathic, accepting, and authentic (Rogers, 1957). These qualities are of particular importance when addressing emotionally sensitive skills, and doing so requires skills, knowledge, and personal integrity. This mitigation contributes to the ethical robustness of the project in that such interviewers are unlikely to cause harm and highly likely to promote beneficence.

Procedure

Sessions began with the interviewers arriving at the homes of the interviewees. They brought breakfast or a cake, depending on the time of the day. This choice was made to ensure that each session would begin with everyone sitting together and sharing a meal. During this meal, the interviewers had time to introduce themselves and explain why they were there. This included explaining the purpose of the study and providing information on participant rights. All information was provided in language that was comprehensible to the children.

The meal also allowed the remaining parent to summarise what had befallen the family. This was important for two reasons: First, it provided the interviewer a better understanding of the events that had occurred and thereby helped to set the stage for the upcoming conversation. Second, hearing the parent discuss the experience was intended to indicate to the child that this was an environment in which it was safe to discuss the loss.

Following the meal, one interviewer would ask the child if they wanted to go somewhere else (eg the living room) and try out the sand tray. At this point, children were nearly always enthusiastic to remove the lid of the large, mysterious box and begin. This reaction is in line with Turner’s (2017a) suggestion that, in contrast to adults, children seldom consciously question sand-based therapy. They seem to have a natural sense ‘that what they have built is important, and it is right’ (Turner, 2017a , pp 8) and are happy to engage with the sand tray. Meeting the interviewers approximately half an hour before the sand tray experiments began meant that most children were able to overcome any initial shyness they may have felt. Furthermore, most of the children seemed to find the sand box mysterious and intriguing. They all seemed to have an immediate desire to remove the lid and discover what lay beneath. This shows how a playful approach can promote engagement and help to build rapport and trust.

As per the recommendations of Turner (2017a), sessions began by having the child help open the sand tray box and then feel the sand with their fingers to stimulate creative thinking and encourage originality of expression. The child was subsequently asked to help remove the figurines from their box. In addition to making the figurines available during the interview, doing so provided the child with an opportunity to become familiar with them.

Following the unpacking of the figurines, the child was encouraged to tell a story about what they remembered from when their parent had been ill or died. This and subsequent questions were built around an interview guide with four parts. These parts are presented in Table 1 and were designed to mirror the previously presented research questions.

The young ages of the children and the intricacies of the method meant that the interviewers were asked to use an interview guide. However, they were also given leeway to follow the story in the direction the child took it. Given the view that a child will generally focus a story on the most important events in their memory, the interview guide was intended not as a rigid guideline but, similarly to semi-structured interviews, allowed for a degree of freedom. This is in line with the recommendation by Cameron (2005) that questions posed to young children should be open and reflective rather than interrogative in nature.

If, at any stage of the interview, the interviewer felt the child was beginning to struggle and become uncertain as to what to say, the interviewer would ask that the child shift the focus of the narrative to a friend who had had a similar experience and how that friend might have felt. Signs of distress could be the child looking away from the sand tray, becoming silent, or changing the topic of discussion. During the interviews, this approach proved effective in moving the conversation forward should difficulties arise.

Part four of the session was designed for two purposes: the first was to allow the child to discuss what they considered to be good support; the second was to conclude the interview on a positive and empowering note by allowing the child to help a fictive friend who had experienced a similar situation. We understand the children’s responses in this regard to communicate both their experiences and what they perceive to be helpful to a grieving child.

At the end of each session, time was allocated to providing an appropriate closure to the interview and discussing the child’s feelings on participation. Thereafter, the interviewer and child returned to the parent. At this point, parents would sometimes be intrigued to learn what had occurred. Occasionally, children would not want to share what they had discussed; if this were the case, the researcher would help convey this to the parent in an ethical manner.

Data analysis

The generated data was collected as audio. While video would have added another source of data for analysis, it was deemed that a camera may have been excessively intrusive in a context in which the interviewer had only one session in which to gain the trust of the child. The recorded audio data was analysed using thematic analysis. Howitt (2010) proposed that thematic analysis can be divided into three distinct steps: 1) transcription, 2) analysis, and 3) theme identification. In order to strengthen the analytical process, the guidelines laid out by Braun & Clark (2006) on how to conduct thematic analysis were followed. As the results of the data generation are covered in a separate article (Lytje, Dyregov & Holliday, 2022) the following sections focus on covering the analytical approach in greater depth.

Evaluation of sand tray interviews

In psychotherapy, sand tray work and play fall under the umbrella of arts-based methods. Therefore, criteria for evaluating arts-based research (ABR) are pertinent when evaluating the use of the sand tray method in research (eg Barone & Eisner, 2012; Sullivan, 2010; Leavy, 2019). Arts-based research practices are ‘methodological tools used by researchers across the disciplines during any or all phases of research’ (Leavy, 2019, p4). There is a consensus that such criteria are to be used judiciously and that they be project-specific. However, when considering the works of all of the abovementioned authors, there is an agreement that the general criteria of aesthetics, ethics, and usefulness are relevant.

Aesthetics

Aesthetics is a branch of philosophy that is concerned with the non-verbal, the senses, and emotions and draws on embodied and imaginal ways of knowing (Chaplin-Dengerink, 2019). Epistemologically, ABR assumes that art can create and convey meaning in and of itself. The art, story-making, and play of human beings can be seen as symbolic expressions of their life experiences. Barone Eisner (2012) coined the term ‘epistemic seeing’ to refer to knowledge secured through sight. We might expand this to ‘epistemic sensing’, referring to knowledge secured through the senses, as working in the sand tray involves senses beyond the visual, such as the kinaesthetic (movement) and haptic (touch). Therefore, we can surmise that there is meaning in the images that children create and the play that they engage in while working with sand trays and that this meaning relates to their life experiences. A potential benefit of ABR is that it can go beyond what is known verbally. Given that the present study involved conducting research with children whose developmental stage meant that their vocabulary was limited, this was advantageous.

Furthermore, playing with sand and working through sensation have the potential to facilitate self-soothing (Armstrong et al, 2017). Examples of such self-soothing were evident, as some children would allow the sand to flow through their fingers during difficult parts of the conversation.

Safeguarding

Creating a trusting atmosphere and being able to make significant methodological adjustments based on the subtle, ongoing feedback received during the sessions were challenging. Trust is established by exhibiting the core conditions of empathy, acceptance, and authenticity (Rogers, 1957). Hence, the relational skills of the researcher and their capacity to establish these conditions are crucial to the quality of the data that can be gathered. Simultaneously, however, the power of the method meant that the participating children would occasionally disclose highly concerning information, which the interviewer would need to process immediately. An example of this occurred when, during a session, a child said, ‘Since my father died, my mum has become evil’. Such comments necessitated impromptu decisions concerning whether an interview could proceed or if the focus should shift to an assessment of the child’s support needs.

Visiting families at home, we encountered some who were living under conditions that may have indicated that they were struggling. During such visits, the interviewers were required to disregard any negative stimuli and maintain focus on the family. In all cases where concerning findings were uncovered, the family was provided with post-interview support. This was done both to assist the family and satisfy our ethical obligations.

However, encountering such situations highlights the importance that untrained and inexperienced students do not use this method. Rather, it should be used by practitioners who have experience in working with children, exhibit a strong relational capacity, and have adequate support. In addition, when engaging in research as sensitive as this, a study must have access to the resources required to support participants post-interview.

In this study, we believe that the researchers demonstrated trustworthiness and integrity, that the participating children had sufficient understanding to give informed consent to participate, that the children were treated justly and not harmed, and that the interviewers strove to make participation as beneficial to the children as possible. Thus, we believe the method, as we conducted it, was ethically sound.

Usefulness

If a method is useful, it should offer insight into the research question(s) by offering answers, revealing that which may not have been known previously, or identify new challenges. We generally found sand tray interviews to be a powerful and flexible method that offers a range of tools with which the participating children could express their views and support needs, thus contributing to answering the research questions. The mysterious nature of the box proved intriguing to the children, and in line with what Turner (2017a) proposed, few participants questioned the logic or usefulness of the method. In two cases where participants had just reached the age of nine, there was some scepticism related to engaging in ‘doll play’. However, this scepticism vanished when the interviewer explained how the sand tray was not about playing with dolls but rather about considering each figurine as symbolic of something else.

Given that the present study involved interviewing children as young as four, an important question is whether the method is effective with this young an age group. This is a difficult question to answer. We would not recommend the use of this technique with children younger than four, as a degree of verbal ability is still required. Among the children we interviewed, some of the four-year-olds struggled to understand the dynamic of the method. They would often forget a question and become fully absorbed with playing with the figurines in the sand tray. At this point, their activity would take the form of more ‘normal’ play rather than a response to a question the interviewer posed and thus required a different analytical approach. An example of this came in the form of a four-year-old disregarding the initial questions and becoming fully absorbed by playing ‘mom and dad’ in the sand tray. However, this was not the case for all of the four-year-olds, as some did engage well with the method. In general, such challenges did not seem to arise during interviews with children of five and above. Most of these children were keen to depict scenes from their experience of loss in the sand tray.

Nevertheless, while age did account for some challenges, we generally found it to be a poor indicator of whether a child could engage with the method. The same has been noted by other researchers, who have also emphasised that children’s ability to engage depends on their developmental levels and personalities.

In conclusion, all the children we interviewed engaged with the sand tray. However, the level of engagement varied significantly. We generally noted two ways in which the children would use the tray: some became absorbed in the world they created in the tray and used this as the primary foundation for sharing their experience, which they communicated through playing in the sand tray; others used the tray as more of a support tool in which they could set up figurines to help depict what they were sharing verbally. In two interviews, one with a five-year-old boy and another with a six-year-old boy, we also witnessed the sand tray not being used excessively. These two boys were already eager to discuss their experiences from the beginning of the interviews and had to be encouraged to use the tray while doing so.

Regarding the sand tray interviews themselves, they worked well in terms of proposing themes that children would subsequently be encouraged to discuss. Nonetheless, particularly with regard to the younger children, the prepared questions occasionally had to be abandoned in favour of a trial-and-error approach to offering alternative phrasings with which the children could better connect. Such efforts could include spending more time discussing a fictional person and what they might feel. In particular, the final question, which involved an eagle asking the child to help a friend, worked well. Even for the younger children, this proved an effective approach to having them explain the friend’s needs and, by extension, their own desires.

After each session, the interviewer evaluated the session with the child. The participants generally provided positive feedback, as they found the method ‘intriguing’ and ‘fun’. However, one boy indicated that the interview had been ‘a bit boring’. In contrast, one girl asked if the interviewers could visit her grief group so that all the children there could try the sand tray. Another girl, aged five, reflected on how ‘putting’ her memories into the sand tray had helped her recall experiences from the time of her bereavement. For her, discussing her loss in the sand tray represented a means by which to investigate and explore her previous experiences with the interviewer.

Applications and results

Having utilised sand tray interviews for the first time, it is important to reflect on the lessons learned. In addition, it is worth discussing the circumstances in which using the method might prove advantageous. Sand tray interviews seem ideal when the goal is to explore sensitive topics with young children. Turner (2017a) suggested that children seem to be comfortable with the method and seldom need to question its purpose or how it works; we found this to be true. In most of the interviews, the children were so intrigued by the box that had been brought into their homes by the researchers and the different toys it contained that they were excited to explore these items further. In addition, being able to discuss another character or friend experiencing an event similar to that which they had undergone, helped provide distance and safety via the use of metaphor and moved the conversation forward when children found it difficult to express their feelings.

There are a few areas where we would recommend altering the approach we used. Compelling results may have resulted from visiting and interviewing children more than once. More time would have allowed additional trust to be built between the children and the interviewers. Such trust is, of course, vital when exploring a sensitive topic. Particularly regarding the four-year-olds, with whom we had mixed results, more time may have helped them to them better understand the method and open up. It may also have provided the interviewer with a better understanding of both the individual child and how to adapt the interview to the child’s specific needs.

While we did not do this in our study, it may be advisable to make an alternative approach available during the data generation so that children who prefer not to engage with the sand tray have an alternative option. This approach could be an activity such as drawing, working with clay, creating a collage, or using another art medium. We did not experience any need to switch to a different data generation technique during our interviews. However, particularly when applying the method to a larger sample, this might offer flexibility should researchers encounter children who are sceptical about using the sand tray. For now, this approach would benefit from further testing and fine-tuning with larger samples.

Concerning when sand tray interviews should be used, Turner (2017a) proposed that the unconscious mind will always focus on the problems that have previously or are currently occupying it. We saw evidence of this when interviewing children, as well as during our test sessions with adults. Here, sessions that were designed to explore non-sensitive topics would often ultimately focus on these topics regardless. The adults in particular would often discover something interesting in their own stories and indicate a desire to explore what this meant using the tray. Whereas children unconsciously engaged with the sand tray and their stories, adults seemed to be fixated on understanding what their creation meant and where their own story was going. This process helped them gain a new understanding of what was taking place in their lives and, often, a deeper comprehension of their current challenges. This kindling of feelings and understanding is precisely why sand tray work is used as a therapeutic medium (Mitchell & Friedman 1994).

Whereas the adults would be consciously aware that they were on this journey, this was a rare occurrence among the children. With this in mind, the technique seems most suitable when the goal is to explore the challenges participants experience in their current lives. It may be less effective at investigating less sensitive topics, as the sand tray has a habit of bringing challenges that participants are currently experiencing to the fore, regardless of the initially posed question (Armstrong, 2008; Turner, 2017a). With the above perspectives in mind, sand tray interviews may be even more effective when used with adults than when interviewing children. Thus, it seems that sand tray interviews work with both children and adults, as is the case with sand-based therapies ( Armstrong, 2008; Garrett, 2014).

Conclusion

Sand tray techniques have a proven record of helping children open up about traumatic experiences (Rae, 2015; Pearson & Wilson, 2019). While effective in therapy, the method cannot be used in research without the potential of encountering significant ethical and methodological risks. These include concealing the real purpose of a sand tray session and being able to link children’s stories to specific experiences. When sand trays are combined with the methodology used in semi-structured interviews, many of the above issues can be avoided, and a compelling new research technique emerges. We have termed this approach the ‘sand tray interview’. This method features the strengths embedded in the original therapeutic approach while allowing an interviewer to steer the conversation in directions that are important to the topic of investigation while also heeding Flewitt’s (2005) observation ‘that researchers have a responsibility towards participants of all ages not only to establish a robust and negotiated ethical framework for their research, but also to ensure that these ethical principles are applied throughout all stages of the research process’ (pp 564). While, in our case, the method was used with children who had experienced the loss of a parent, it seems that this tool could help researchers explore many areas of children’s lives where the latter may have experienced something that might be difficult to discuss and can activate defense mechanisms.

Sand tray interviews also seem a powerful method in which the experiences and challenges currently occupying the participant always seem to be brought to the fore. This makes them effective at uncovering the challenges that exist in a participant’s life. However, it might make sand tray interviews less effective when exploring areas of people’s lives that are not related to challenges participants experience as most prominent.

When considering the use of the method, researchers should be aware that it is challenging to master, as it demands that the interviewer both be experienced and have specialised knowledge regarding children. Furthermore, the power of the technique can lead to unexpected revelations during interview sessions that the interviewer will have to react to.

As proposed by Turner (2017a), the unconscious mind will always focus on those issues that are of most significance at time of the conversation. This poses additional ethical demands on the researcher and requires significant post-support initiatives to be made available to participants before such interviews can be justified. We also recommend that researchers using the method have expertise with young children through either training (eg being a social worker or psychologist) or research experience.

If used appropriately, the technique creates an environment where the participant and researcher can explore sensitive matters safely and positively. Even young children can benefit from this approach, and this study had significant success in interviewing children as young as the age of five. With further adjustments and more time, we believe good results could also be achieved with some four-year-olds. Nonetheless, using sand tray interviews has been an experimental process. For the method to reach its full potential, it will have to be further evaluated and developed as experiences are gained in subsequent studies.

References

Armstrong SA (2008) Sandtray therapy: A humanistic approach. Ludic Press.

Armstrong SA, Foster RD, Brown T & Davis J (2017) Humanistic sandtray therapy with children and adults. In: ES Leggett & JN Boswell (eds) Directive play therapy: theories and techniques, 217–253. Springer.

Astington JW (1993) The child’s discovery of the mind. Harvard University Press.

Barone T & Eisner E (2012) Arts based research. SAGE Publications Inc.

Blake P (2011) Child and adolescent psychotherapy. Karnac.

Bond T & Griffin G (2013) Ethical framework for good practice in counselling and psychotherapy (5th ed). British Association for Counselling and Psychotherapy.

Braun V & Clarke V (2006) Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2) 77–101. https://doi.10.1191/1478088706qp063oa.

Brent D, Melhem N, Donohoe MB & Walker M (2009) The incidence and course of depression in bereaved youth 21 months after the loss of a parent to suicide, accident, or sudden natural death. The American Journal of Psychiatry, 166(7) 786–794. https://doi.10.1176/appi.ajp.2009.08081244.

Brinkmann S & Kvale S (2014) InterViews: Learning the craft of qualitative research interviewing (3rd ed). SAGE Publications Inc.

Bylund-Grenklo T, Fürst CJ, Nyberg T, Steineck G & Kreicbergs U (2016) Unresolved grief and its consequences. A nationwide follow-up of teenage loss of a parent to cancer 6–9 years earlier. Supportive Care in Cancer, 24(7) 3095–3103.

Cameron H (2005) Asking the tough questions: a guide to ethical practices in interviewing young children. Early Child Development and Care ,175(6) 597–610.

Carey L (ed) (2006) Expressive and creative arts methods for trauma survivors. Jessica Kingsley Publishers.

Chaplin-Dengerink A (2019) The philosophy of Susanne Langer: Embodied meaning in logic, art and feeling. Bloomsbury Academic.

Clark CD (2010) In a younger voice: Doing child-centered qualitative research. Oxford University Press.

Cullberg J (2006) Crisis and development. Natur och Kultur.

Dobretsova A & Wiese EBP (2019) Sandplay: Psychotrauma signs in asylum seeking adolescents. Journal of Infant, Child and Adolescent Psychotherapy, 18(4) 403–416.

Eckhoff A (2019) Participatory research with young children. Springer.

Fielding M (2001) Students as radical agents of change. Journal of Educational Change, 2(2) 123–141.

Flewitt R (2005) Conducting research with young children: Some ethical considerations. Early child development and care, 175(6) 553–565.

Freud A (1937) The ego and the mechanisms of defense. The International Psycho-analytical Library, 30. Hogarth Press.

Galetta A (2013) Mastering the semi-structured interview and beyond: From research design to analysis and publication. New York University Press.

Garrett M (2014) Beyond play therapy: Using the sand tray as an expressive arts intervention in counselling adult clients. Asia Pacific Journal of Counselling and Psychotherapy, 5(1) 99–105.

General Data Protection Regulation (GDPR) (undated) General data protection regulation (GDPR) – official legal text. Available at: https://gdpr-info.eu/ [accessed 4 September 2020].

Gray LB, Weller RA, Fristad M & Weller EB (2011) Depression in children and adolescents two months after the death of a parent. Journal of Affective Disorders, 135(1) 277–283.

Happé F & Frith U (2014) Annual research review: Towards a developmental neuroscience of atypical social cognition. Journal of Child Psychology & Psychiatry, 55(6) 553–577.

Holliday C (2013) Creative therapeutic interventions. In: C McLaughlin & C Holliday (eds). Therapy with children and young people: Integrative counselling in schools and other settings. Sage

Homeyer LE & Sweeney DS (2016) Sandtray therapy: a practical manual. Routledge.

Howitt D (2010) Introduction to qualitative methods in psychology (1st ed). Prentice Hall.

Hunter SB & Smith DE (2008) Predictors of children’s understandings of death: age, cognitive ability, death experience and maternal communicative competence. OMEGA – Journal of Death and Dying, 57(2) 143–162.

Kaltman S & Bonanno GA (2003) Trauma and bereavement: Examining the impact of sudden and violent deaths. Journal of Anxiety Disorders, 17(2) 131–147.

Kitson GC, Clark RD, Rushforth NB, Brinich PM, Sudak HS & Zyzanski, SJ (1996) Research on difficult family topics: Helping new and experienced researchers cope with research on loss. Family Relations 45(2) 183–188. https://doi.org/10.2307/585289.

Koller D & San Juan V (2015) Play-based interview methods for exploring young children’s perspectives on inclusion. International Journal of Qualitative Studies in Education, 28(5) 610–631.

Krueger RA (2009) Focus groups. A practical guide for applied research. SAGE Publications Inc.

Leavy P (ed) (2019) Handbook of arts-based research. Guilford Publications.

Lowenfeld M (1979) The world technique. Allen & Unwin.

Lytje M, Dyregrov A & Holiday C (2022) When young children grieve: Daycare children’s experiences when encountering illness and loss in parents. International Journal of Early Years Education. Advance online publication. https://doi. 10.1080/09669760.2022.2025581.

Mannay D, Staples E and Edwards V (2017) Visual methodologies, sand and psychoanalysis: employing creative participatory techniques to explore the educational experiences of mature students and children in care. Visual studies, 32(4) 345–358.

Miller C & Boe J (1990) Tears into diamonds: Transformation of child psychic trauma through sandplay and storytelling. The Arts in Psychotherapy, 17(3) 247–257.

Mitchell R & Friedman H (1994) Sandplay: Past, present and future. Routledge.

Morgan M, Gibbs S, Maxwell K & Britten N (2002) Hearing children’s voices: Methodological issues in conducting focus groups with children aged 7–11 years. Qualitative Research, 2(1) 5–20.

National Scientific Committee (2020) Hvad skal jeg anmelde? [What do I need permission for?] Available at: www.nvk.dk/forsker/naar-du-anmelder/hvilke-projekter-skal-jeg-anmelde [accessed 27 October 2020].

Nakajima S, Masaya I, Akemi S & Takako K (2012) Complicated grief in those bereaved by violent death: The effects of post-traumatic stress disorder on complicated grief. Dialogues in Clinical Neuroscience, 14(2) 210–214.

Pearson M & Wilson H (2019) Sandplay Ttherapy: A safe, creative space for trauma recovery. Australian Counselling Research Journal, 13(1) 20–24.

Punch S (2002) Research with children the same or different from research with adults? Childhood, 9(3) 321–341. https//doi:10.1177/0907568202009003005

Rae R (2015) Sandtray: Playing to heal, recover, and grow. Rowman & Littlefield Publishers.

Raveis VH, Siegel K & Karus D (1999) Children’s psychological distress following the death of a parent. Journal of Youth and Adolescence, 28(2) 165–180. https//doi:10.1023/A:1021697230387.

Rees A, Maxwell N, Lyttleton-Smith J & Staples E (2020) Visiting mum: children’s perspectives on a supported scheme when visiting their mother in prison. Child Care in Practice, 28(3) 247–262. https://doi:10.1080/13575279.2020.1769025

Rogers C (1957) The necessary and sufficient conditions of therapeutic personality change. Journal of Consulting and Clinical Psychology, 21(2) 95–103.

Slaughter V (2005) Young children’s understanding of death. Australian Psychologist, 40(3) 179–186.

Slaughter V (2007) Death understanding and fear of death in young children. Clinical child psychology and psychiatry, 12(4) 525. https://doi.org/10.1177/1359104507080980.

Stevens C (2004) Playing in the sand. The British Gestalt Journal, 13 (1) 18–23.

Sullivan G (2010) Art based practice as research: Inquiry in the visual arts (2nd ed). SAGE Publications Inc.

Sunderland M (2015) Conversations that matter: talking with children and teenagers in ways that help. Worth Publishing Ltd.

Turner BA (2017a) Introduction – What is sandplay therapy? In: BA Turner (ed). The Routledge International Handbook of Sandplay Therapy (1st ed). pp 1–32). Routledge.

Turner BA (ed) (2017b) The Routledge International Handbook of Sandplay Therapy (1st ed). Routledge.

Vaughn S, Schumm JS & Sinagub J (1996) Focus group interviews in education and psychology. SAGE Publications Inc.