This study included 99 shelter professionals who worked in child and family social work in a shelter for homeless families or in women’s shelters in the Netherlands during the study period (October 2018 until February 2019). The shelter professionals were selected from 20 shelters that participated in a Dutch nationwide project aimed to enhance the wellbeing of families in shelters through the development and use of natural environments. One year prior to data collection, each shelter had received funding for the design and landscaping of natural places. Each shelter developed a restorative garden, a natural play area, a children’s farm, and/or a vegetable garden.

Shelter managers were asked to include team members for participation in this study on the basis of being a professionally educated child and family social worker, being motivated to use nature in shelter social work, and being motivated to participate in research. Shelter managers provided the researchers with a list of team members who fitted the inclusion criteria, after which researchers contacted the professionals to inform them about the goal of the study and their rights as participants. Participating professionals signed for informed consent. For the participant flow, see Table 1.

All participating professionals were educated in child and family social work in secondary vocational training, bachelor education, or master education. Their training includes assessing the needs of their clients by listening, questioning, observing, and professionally weighing these sources of information to form a professional judgment, as well as in facilitating activities in the support of their clients, and in critically reflecting on the effects. For participant characteristics, see Table 2. Professionals participated during their regular and paid working hours. Shelters could claim the expenses for professionals’ time spent on participation in the research, with a maximum of 32 h per professional at their hourly rate.

All participating professionals took four training sessions in which they shared ideas for and experiences with nature activities for families under their care and reflected on their practice in conversations with colleagues, with the aim to develop, maintain, and share professional insights on nature activities for parents in shelters. After the second and third training session, data collection took place.

Professionals facilitated a nature activity for a family under their care, either on the shelter property or in natural environments near the shelter. Nature activities were offered according to the family’s possibilities, preferences, and needs. Immediately after the nature activity, professionals wrote a case description with the date and time, the name of the shelter, a written observation of the parent’s parental needs based on the question “What needs did the parent have at this moment (based on your professional judgment)?”, a written description of the nature activity based on the question: “You facilitated a nature activity. What exactly happened? Describe the activity,” and a written observation based on the questions: “What did you notice in the parent? And what else? And what else?”. Parents filled out an online questionnaire about their parental need satisfaction and need frustration and their connectedness to nature, of which the results have been published (Peters et al., 2020b).

The Scientific and Ethical Review Board of the Faculty of Behavioral and Movement Sciences of the VU Amsterdam approved of the study protocol (VCWE-2018-0138).

Data consisted of 160 case descriptions. Eleven cases were removed because they had missing data or because the case described no nature activity with a family (for example, cases that described a professional with a child during a nature activity, without its parents). This resulted in a total of 149 cases for analysis. The data was thematically analysed using Atlas.ti 9.0.5 for Mac, using a Grounded Theory approach (Glaser and Strauss, 2017). The analysis was aimed to identify professional decisions in choosing nature activities for families in shelters One researcher conducted the analyses, supported by two researchers who parallel coded parts of the data. Researchers discussed the codes and importance of the codes until consensus was obtained. Memoing was used throughout the process of analysis. The analysis followed six steps:

Case descriptions portrayed nature activities with parents (age M = 31.95, SD = 6.79) and their children (age M = 5.2, SD = 3.69). Nature activities were conducted with one child (n = 97), two children (n = 34), three children (n = 10), four children (n = 3), or five children (n = 3) and one parent (n = 143) or two parents (n = 4; missing data n = 2).

The process of axial coding resulted in a series of codes. Supplementary Material 1 shows the codes and the number of occurrences of each code. The codes from axial coding were summarized and combined, resulting in three major themes.

Two cases were in contrast with the rest of the data and were identified as negative cases. In these cases, professionals aimed for a different pattern for interaction between parent and child compared to the rest of the cases in the data set. Both negative cases described an activity in which professionals gave strict behavioral instructions to both parents and children with the intention to practice a new skill (psycho-physical resilience in one case, collaborative skills in the other case). In these cases, professionals left limited degrees of freedom to the child’s exploration. In the rest of the data professionals may also have described assignments and directions for behavior in parents in children, but still left room for initiatives from the parent or the child which allowed them to balance between proximity seeking and exploration.

We made a network of codes to identify structures in the data set, with the aim to understand professionals’ choices when facilitating nature activities for the support of parents. The model presented in Figure 1 describes professionals’ choices as immerged from the data.

For validation, results were discussed in a focus group with six professionals who had participated in the study, and with four researchers in the field of environmental psychology. The professionals recognized that they used nature as a setting with stress reducing and strengthening capacities for parents and exploration opportunities for children, and that they made choices on the eight dimensions to make the nature activity fitting for a specific family (Quote respondent 103: “I do indeed recognize that our social workers consciously think about relaxation moments for parents and play opportunities for children. As a team, we certainly think that nature is helpful in this”). All professionals in the focus group recognized that they chose activities that promoted a back-and-forth between children’s exploration away from the parent and return. Some professionals recognized their own role in this (Quote respondent 101: “I do try to sometimes involve a parent in the child’s activities: proximity; and sometimes stimulate a parent to let a child do something independently: distance. Sometimes you give some guidance, sometimes you just let things happen between parent and child”). Other professionals saw it as a side effect (Quote respondent 103: “I think that our family social workers are to a lesser extent consciously engaged in creating a balance in distance and proximity, but rather experience that this is a favorable side effect of working with nature,” Quote respondent 104: “At present the issue of distance/proximity and mutual contact between parent and child remains consciously observed by our colleagues but is not yet really pursued on purpose. It just happens, outside”). The professionals and the psychological researchers stressed that the choices for practical dimensions of the nature activities were partly influenced by practical considerations (Quote respondent 101: “But practical considerations, for example close to the shelter, free of charge, etcetera, are also decisive for the choice of activity”) and that decisions were not always planned prior to the activity but partially made in response to what happened during the activity.

When facilitating nature activities for the support of parents in shelters, professionals chose to facilitate nature activities that promoted a back-and-forth between children’s exploration away from the parent, and return. These interactions followed the pattern of Secure Base Phenomenon (Ainsworth, 1967; Posada et al., 2013). Professionals facilitated such interactions, and only by exception opted for a different pattern for interaction (the “negative cases” in the Results). Other patterns may have occurred if professionals made different choices in the design of nature activities, for example by choosing activities that left little room for exploration, or by choosing activities in which parents could not be available for the child such as activities that caused high levels of arousal, or activities that required directed attention on something else than the child. Social workers’ primary focus on creating interactions according to secure base phenomenon is in line with a recent finding that child care professionals rely on Attachment Theory most often for their child supportive work (Department of Education, 2018).

In facilitating secure base interactions, professionals used nature’s capacities for supporting children’s exploration, and nature’s capacities for facilitating stress reduction and strengthening experiences. Regarding children’s exploration, professionals chose activities in which nature functioned as an interesting play environment that allowed children’s freedom in play behavior and stimulated their involvement in play activities. Theories on play suggest that natural environments can function as a setting for rich explorations (Nicholson, 1972; Heft, 1988). Natural environments are described as a setting that fits with children’s needs and desires for exploration and play (Spencer et al., 2019), where children play long, involved and diverse (Luchs and Fikus, 2013; Zamani and Moore, 2013). Professionals’ choices revealed that they set the scene for secure base behavior by choosing the environment so that it offered exploration opportunities for the child.

Regarding parents’ stress reduction and strengthening experiences, professionals chose activities in which nature offered opportunities for restoring energy, reducing feelings of anxiety, and experiencing positive interactions. Experts in the field of environmental studies have also described these capabilities of nature with the terms restoration and building (Markevych et al., 2017), with restoration referring to the ability of nature to restore resources that have been depleted in efforts to cope with stressors, and building referring to the deepening or strengthening of capabilities for meeting everyday demands (Marselle et al., 2021). Several theories were aimed to explain why natural environments can be experienced as non-threatening and stress reducing (Wilson, 1984; Kaplan and Kaplan, 1989; Ulrich et al., 1991; Kaplan, 1995; Kuo, 2015). Interestingly, recovery from stressors can been linked to higher psychological availability of parents, and more autonomy supportive and less controlling parenting behavior (Van Der Kaap-Deeder et al., 2019; Robichaud et al., 2020). Professionals’ choices revealed that professionals set the scene for secure base behavior by choosing the environment so that it facilitates parents to be available to their child and supportive of the child’s autonomy.

To facilitate exploration in children and stress reduction and strengthening experiences in parents, professionals made practical choices on eight dimensions. Professionals made unique choices on these dimensions for each family. According to Theory of Affordances (Gibson, 2014) every physical setting has unique properties for every individual. This means that, when aiming to facilitate certain behavior through engagement with an environment, one must consider the physical aspects of the environment as well as the characteristics of the family. Professionals’ choices show that they selected unique characteristics of a place to make its affordances fitting for a family.

Professionals chose to facilitate interactions according to secure base phenomenon in their nature activities. This does not mean that we expect nature activities to be uniquely suitable for secure base interactions, nor that we expect professionals to use their knowledge on secure base phenomenon only during activities in nature. It may equally well be expected that professionals facilitate possibilities for secure base interactions on other moments in their professional practice, such as during indoor play moments or dinner time. This research shows that professionals also included secure base knowledge in their choices for facilitating nature activities.

The current study uncovers professionals integrated knowledge that is compatible with attachment theory in their choices for nature activities. Professionals might have well used other knowledge aspects, such as knowledge on physical fitness (which could have shown if professionals chose activities focused on building physical strength or getting vitamin D), or knowledge on social connectedness (which could have shown if professionals chose activities focused on connecting to the neighborhood or building friendships), or knowledge on self-connectedness (which could have shown if professional chose activities such as forest bathing, mindfulness, or yoga). It is of interest that professionals chose a social interaction perspective because it adds a new perspective to existing literature that has mainly focused on nature activities for physical health and mental wellbeing (Van den Bosch and Sang, 2017; Twohig-Bennett and Jones, 2018; Bratman et al., 2019; Lackey et al., 2021).

As professionals in the focus group highlighted, some aspects of the activity ‘occurred’ without professionals’ intentional guidance. Each activity reflected input from the professional, the parent, the child, and characteristics of the natural environment, that all interacted with each other, which makes the activity not only a result of the choices of professionals. The professional choices that were the focus of this study referred therefore to the input that professionals chose to provide in the decision process with the family about the nature activity.

The study was conducted in the setting of a shelter, during regular working practice, with regular clients. This contributed to the ecological validity of the study findings. The study was conducted among a selected group of professionals. All professionals were educated in child and family social work, were selected by their team manager, worked in a shelter that had implemented nature to enhance the wellbeing of families, and were trained in the implementation of nature for parents. This allowed us to analyze data from professionals who we expected to be skilled and knowledgeable on the subject. Their choices were the basis on which a practice-based model was made.

No professionals dropped out during the informed consent procedure, but 23% of professional participants dropped out during data collection (see Table 1). Natural turnover of staff and clients, which is to be expected with a study period of 12 months, explained almost half of the dropout. The remaining dropout was 12%, which is acceptable (Furlan et al., 2009).

The data in this study consisted of case descriptions in which professionals described moments in their own practice. As Ryle (2009) argued, the mind of professionals is revealed in their doings, and explainable by the doers’ aims. This makes case descriptions suitable material for analyzing professional choices. We recognize three limitations in the way we collected case descriptions. Firstly, the case descriptions were limited in richness. Case descriptions were based on a predefined set of questions, which limited the options for rich elaborations by professionals, and limited opportunities for researchers to ask further questions. The written accounts by professionals were a representation of their actions in practice, but a first-hand involvement from the researchers, using their reflective and interpretive stance in interaction with the professionals, could have deepened our understanding. Secondly, professionals’ desired behavior, intentions, and world views could have interfered with their descriptions of their actual behavior. Professionals wrote about their own practice, which is already a reflection on their behavior (even though it was written immediately after the nature activity) and not necessarily an actual representation of the behavior itself. Professionals may have filtered their actual actions through the lens of their desired behavior, to make it sensible, logical, and concurrent with their values. A relational approach in data collection with researchers closer to the professionals, e.g., by actively observing professionals in action, could have strengthened the study. Thirdly, data collection and analysis were performed separately, which prevented theoretical sampling (Glaser and Strauss, 2017). Theoretical sampling could have made further examination of the categories and their relationships possible, such as an examination of the conditions under which certain characteristics of nature activities were chosen.

This study describes the daily practical choices that professionals make when facilitating a nature intervention for a family in their care with the intention to support parents. The results of the study can be used by professionals who aim to implement nature to support parents in their practice. The results can make professionals aware of the choices that other professionals make. On the one hand, such insight may be interesting for professionals who are beginners in the use of nature activities in their parenting supportive practice. For them, the results can function as a “cheat sheet” on which they can see how more experienced colleagues make their choices, and so inform their own choices for facilitating nature engagement among their clients. On the other hand, the insights from this study may be used by professionals to reflect on practice. The results can function as a ‘mirror’ that reflects current choices. Such a ‘mirror’ can stimulate reflective conversation, for example by discussing if the choices reflect what professionals consider good practice. To aid this practical use of the study results, we made a printable poster that can function as a reflective tool for practice (Supplementary Material 3).

The data of this study have been collected on shelters that had some form of nature on their own property, safe and open to the families. Shelter professionals who do not have a natural environment at their disposal may be limited in the choices available to them. Especially for clients for whom safety is at risk or who face financial barriers, professionals will be very limited in choosing to go outdoors, to go further away from the shelter, or to facilitate autonomous family time in nature, if no safe and affordable natural environment of their own choice is available. It may well be that a safe garden on the shelter’s property is an important prerequisite for being able to make apt professional choices. If professionals are limited in making the choices from the model due to practical constraints, the model cannot function as intended. We invite professionals to make an analysis of the natural environments they have available at work, to determine whether using the model fits to their practice.