Social interactions among nonparental adults and parents in the neighbourhood: a mixed-method vignette study

ABSTRACT

Supportive social interactions between nonparental adults (i.e. social work professionals, volunteers, and other parents that have contact with children but are not the primary caregiver), parents, and children are important for children’s well-being and development. Parenting styles, types of child behaviour, and location in the neighbourhood may influence these interactions. The aim of the present study was to identify when and how nonparental adults respond in interactions with other adults and children in the neighbourhood. A mixed-method study with vignettes and interviews (N = 114) was conducted to gain insight into which factors (parenting style, child behaviour, location in the neighbourhood) influence the nonparental adults’ intention to respond to children and/or parents. Nonparental adults indicated they were most likely to respond in the context of a permissive parenting style or a child’s externalising behaviour. Professionals more often felt responsible than parents and volunteers, although they did not respond more often. All three factors were related to the participants’ willingness to respond and promote a supportive social structure in the neighbourhood. Social work professionals and their organisations can use this study to identify social support interactions and to discuss their responsibilities.

SAMENVATTING

Sociaal ondersteunende interacties tussen mede-opvoeders (zoals sociaal werk-professionals, vrijwilligers en andere ouders die contact met kinderen hebben, maar niet primair verantwoordelijk zijn) zijn belangrijk voor het welzijn en een positieve ontwikkeling van kinderen. Het doel van deze studie was inzicht geven in hoe mede-opvoeders reageren in interacties met andere opvoeders en kinderen in de buurt. Een mixed-method design met vignetten en interviews is toegepast om inzicht te krijgen in welke factoren (opvoedstijl, gedrag van een kind en locatie in de buurt) de reactie van mede-opvoeders beïnvloeden. Mede-opvoeders gaven aan dat ze het meest reageren in situaties waar sprake is van een permissieve opvoedingsstijl of externaliserend gedrag van een kind. Professionals voelen zich meer verantwoordelijk dan ouders en vrijwilligers, maar reageren niet vaker. Opvoedstijl, gedrag van het kind en locatie in de buurt hangen samen met de mate waarin respondenten reageren en om een ondersteunende sociale structuur in de buurt te bevorderen. Sociaal werkers en hun organisaties kunnen deze studie gebruiken om sociaal ondersteunende interacties te identificeren and over hun verantwoordelijkheid te discussiëren.

KEYWORDS:

• Nonparental adults
• community
• interpersonal relations
• neighbourhood
• social interaction

KERNWOORDEN:

• mede-opvoeders
• community
• interpersoonlijke relaties
• buurt
• sociale interactie

Nowadays, many families feel disconnected from their community (OECD, 2016; Ulferts, 2020; Zahran, 2011), which may affect social support for parents (Pimentel, 2016) and can result in negative health outcomes for families (Ozbay et al., 2007). Social support of families refers to the relational dimension with reciprocal interactions among community members (Gottlieb & Bergen, 2010), including community members who act as role models (Elshater, 2018; Kegler et al., 2005). This social support reinforces effective parenting (Caughy et al., 2012; Kesselring et al., 2013) and is associated with higher self-esteem and better social skills of children (Scales et al., 2004). Therefore, the importance of social support and social networks with structural and functional connections with other community members (Gottlieb & Bergen, 2010) resonates in many Western-European and other Western countries, including the Netherlands (Baker & Daro, 2005; Kesselring et al., 2013; Pastor Seller, 2015). Relatedly, there is increasing attention for the role of social work professionals and other nonparental adults (e.g. volunteers, other parents) that have contact with children, in creating supportive and inclusive communities for parents and children, although they are not the primary caregiver (Doornenbal et al., 2015; Geens et al., 2019; Kesselring et al., 2013).

Supportive social structures in communities and neighbourhoods (Jenks & Dempsey, 2007) are important for creating social support (Holt-Lundstad & Uchino, 2015), and refer to a mix of bridging and bonding meaningful relationships for families (Fram, 2005; Geens et al., 2019) with other nonparental adults. Conversely, unsupportive interactions, deviant behaviour, or negative modelling can result in a lack of social support (Holt-Lundstad & Uchino, 2015). Professionals, volunteers, and parents can positively influence supportive structures (Brown et al., 2020; Hill et al., 2014). For example, in areas with supportive structures present, children experience fewer internalising problems (Ma & Grogan-Kaylor, 2017). Little is known, however, about building supportive relations in neighbourhoods with low levels of reciprocity (Baker & Daro, 2005), such as neighbourhoods with ethnically diverse networks (Barwick, 2015). Explicit examples and in-depth understanding of community are lacking, as are professional efforts to create supportive relationships for parents (Geens et al., 2019; Kesselring et al., 2012; Scales et al., 2004). Much research attention is paid to the effect of violence and socioeconomic status of children and parents (Bradley & Corwyn, 2002; Oravecz et al., 2012) and the substantial role of parents and peers in the socialisation process of youth (Fitzgerald et al., 2012; Owen et al., 2012). Research into community efforts is underrepresented, even though these efforts are important for children’s socialisation and development, because nonparental adults might contribute to supportive relations when they consider daily interactions with others important and opportunities to discuss social norms (Holt-Lundstad & Uchino, 2015; Scales et al., 2004). Exploring nonparental adults’ supportive relations with and responses to parent–child interactions in the neighbourhood may help to gain insight into individual community efforts and supportive social structures for children and parents.

Examining supportive dynamics in neighbourhoods

When examining supportive relations in urban neighbourhood interactions, parenting style is an important factor (Goldfeld et al., 2015; Miller & Tolan, 2019). Parenting styles can be influenced by the cultural and ethnic background of community members (Goldfeld et al., 2015; Liu & Guo, 2010). For example, Chinese parents can be more controlling than American parents (Liu & Guo, 2010). A complicating factor is that interactions between members of the community may not occur naturally in urban neighbourhoods with an ethnically diverse population, because parents might experience a barrier to discuss norms and values, which may vary widely between adults in superdiverse neighbourhoods (Crul & Lelie, 2019; Gadsden & Dixon-Román, 2017; Van Beurden & De Haan, 2019).

Another important factor is child behaviour (Beyers et al., 2003), as this is related to internalising (emotional problems, for example: withdrawn behaviour) and externalising problems (anti-social behaviour, for example: hitting someone) (Ma & Grogan-Kaylor, 2017; Oravecz et al., 2012). Social interactions between nonparental adults, parents, and children can be influenced by the type of behaviour, because emotional problem behaviour is often not recognised by nonparental adults (O’Brien et al., 2016) whereas externalising behaviour is more visible.

Third, social interactions in informal meeting areas in urban neighbourhoods (i.e. third place) are crucial for children’s development (Elshater, 2018), because they learn through various interactions with peers and nonparental adults (Bennet et al., 2012; Caroll et al., 2015). Social interactions can be stimulated or hindered by the location where they take place (Bennet et al., 2012; Geens et al., 2019; Goldfeld et al., 2015). The presence of institutions such as a (pre)school may facilitate positive interactions between parents (Geens et al., 2019). This suggests the importance of other meeting places in urban neighbourhoods for social interactions, such as preschools, playgrounds, and the street (Bennet et al., 2012; Elshater, 2018; Geens et al., 2019). An important element of the third place is the fact that different types of nonparental adults are usually present at different locations (Elshater, 2018). However, relatively little is known about whether and how nonparental adults contribute to social support in various locations (Ulferts, 2020). Attention for the role of professionals in neighbourhoods is increasing (Doornenbal et al., 2015; Geens et al., 2019). Current research has a strong focus on the childrearing responsibility shared between parents and pays less attention to the role of volunteers in neighbourhood interactions (Kesselring et al., 2012; 2016; Scales et al., 2001). Further examination of the role of nonparental adults and other factors influencing neighbourhood interactions is important for increasing insight into social support in the neighbourhood. This requires new methods to systematically question nonparental adults about their role in this understudied area.

The present study

To stimulate supportive social structures in the public space, a positive attitude of nonparental adults towards adult–child interactions is important (Kesselring et al., 2013). However, a knowledge gap exists about which factors (i.e. parenting style, child behaviour, location) influence the attitude of nonparental adults (i.e. professionals, volunteers, other parents) in parenting situations and how this may influence social support. In order to identify predictors and describe the supportive social structures within neighbourhoods, deeper insight is required into whether and how nonparental adults respond to adult–child interactions (Caughy et al., 2012; Goldfeld et al., 2015; O’Brien et al., 2016). Identifying the relation between social factors and the response to parenting situations may help further the understanding and development of social support and social networks in the neighbourhood.

This study addresses the following question: What is the influence of social factors (i.e. parenting style, child behaviour, location in the neighbourhood) on how nonparental adults (professionals, volunteers, other parents) respond to parenting situations in the neighbourhood? A study was conducted using a vignette design in an ethnically diverse group of stakeholders, this yielded contextualised responses and complements existing research, which relies on questionnaires (Hughes, 1998; Kesselring et al., 2012).

Method

A mixed-method design with vignettes and interviews was used. The study was conducted using an exploratory concurrent mixed-method design, because quantitative data (i.e. answers to closed questions from the vignettes) and qualitative interview data (answers to open questions) were collected simultaneously.

Participants

A heterogeneous group of 114 nonparental adults participated, all living or working in Amsterdam, the Netherlands. The Southeast city district was selected, because of its diverse population. Participants were classified by the function they held at the time of the interview: 47 professionals working in primary education (22) and preschools (25), 24 volunteers and 43 other parents. The mean age was 39.2 years (SD = 11.35) and a majority (87%) of participants was female. Less than half (42%) had a Dutch background, more than a quarter (29%) had a South-American background (e.g. Surinam, Dominique Republic), 14% an African background (e.g. Ghanaian, Congolese) and 6% a Middle-Eastern background (e.g. Turkey). 5% of participants had other migration backgrounds and of 4% their backgrounds were unknown.

A sample size of 60–120 is advised in the literature for qualitative data saturation (Guest et al., 2006). Further, an a priori analysis indicated that a sample size of 100 participants allows for a test of medium-sized differences at the conventional alpha level of.05 with adequate statistical power (β = .80).

Procedure

Participants were recruited between October 2018 and September 2019, through four organisations in Amsterdam: a local child welfare organisation, two primary schools, and one division of a national welfare organisation through which volunteers provide parenting support at home. Together, these organisations cover large parts of the institutional contexts. The inclusion criteria were that participants needed to have contact with (other) parents and children, and live or work in Amsterdam-Southeast. Organisations used flyers in Dutch and English (e.g. with the goal, duration, executive research institute, and researcher) to recruit and inform participants. Prior to the interview, the researcher informed the participants by an information letter about the study (e.g. about the anonymity statement, data storage, and that participants were allowed to refuse participation at any time). All participants gave informed consent.

In an attempt to reduce socially desirable answers, vignettes were used in a random order in an individual face-to-face interview with open questions rather than in a focus group. After each vignette, 5 questions were asked in Dutch or English, about what the participant would do in this situation (i.e. self-report measure) and how they assessed the described behaviour. Each interview lasted approximately 40 min. The interviews were recorded and transcribed. If the participant did not give permission to have the interview recorded, a student assistant wrote along. After the interview, parents received a 15 euro gift voucher.

Measures

The vignettes were developed by experts and the project research team, supervised by the first author, and tested by 12 student assistants. For their construction the steps of Atzmuller and Steiner (2010) were used. Parenting experts (N = 10; school professionals, volunteers, parents) were consulted, after which three important vignette factors were determined and theoretically examined: parenting style, child behaviour, and location.

Parenting style

Commonly, three different parenting styles are distinguished (Robinson et al., 1995; Weis & Toolis, 2010): authoritarian, authoritative, and permissive (Baumrind, 1971). Although the uninvolved parenting style (Collins et al., 2000) is often distinguished as well, this style was not included because it proved unsuitable for the vignette construction. An authoritarian parenting style comprises strict rules and little support for children, while an authoritative parenting style gives children the opportunity and support to develop a sense of responsibility. In a permissive parenting style there is little parental control: few rules are set and/or enforced. The validated Parenting Practices Questionnaire Constructs (Robinson et al., 2016) is based on Baumrind’s parenting styles and consists of three scales, each measuring a parenting style. The contents of items of each of these scales were used in the vignettes. Only well-suited items with factor loads >.50 were included.

Behaviour of the child

To operationalise this factor, the Dutch qualification scheme CAP-J for children was used. The CAP-J is a classification system focused on the nature of youth’s problems. The CAP-J is applied in addition to the commonly used DSM-5 screening method for behavioural problems and is intended to classify less severe problems. The factor ‘child behaviour’ incorporates the description of internalising and externalising behaviour from the CAP-J. These descriptions score well / reasonably on interrater reliability (Konijn et al., 2009).

Locations in the neighbourhood

Based on literature (Bennet et al., 2012; Caroll et al., 2015; Elshater, 2018; Geens et al., 2019) and expert interviews two types of location were determined: around school and within the neighbourhood. These were then further specified, for example: the schoolyard (around school) or the neighbourhood playground (within the neighbourhood).

A total of 120 vignettes were developed, systematically combining three factors. Preschool professionals, school directors, and volunteers (N = 8) assessed their usability and relevance, which makes responses closely related to reality (Hughes, 1998; Rahman, 1996), after which 24 vignettes were chosen. Four sets of six vignettes were developed, with an equal distribution of the different factors within each set. Examples of vignettes were: Parent and child are standing in the schoolyard after school. The child punches their parent in the stomach and kicks them in the leg. The parent sits down next to the child and says: ‘You are punching me in the stomach and kicking me in the leg. Why are you doing that?’ or: Parent and child are standing in the neighbourhood playground after school. Other children are playing together. The child is hiding behind the parent. The parent is watching a video on YouTube. Each vignette contains an operationalisation of three factors: child behaviour (underlined); parenting style (italics); location (bold). The child’s age ranged between 2 and 8 and was described in each introduction. The vignettes were translated into English and Arabic for the interviews with non-Dutch speaking participants. Each participant received a subset of the entire vignette population (Atzmuller & Steiner, 2010) and on average responded to 4.6 vignettes.

Data analysis

The quantitative and qualitative results of the vignette study were analysed. The quantitative data were used as a framework to determine relevant factor combinations and explore relationships between responses (Creswell, 2013). For the quantitative analysis, interview segments were converted into numeric data based on the qualitative answer categories. For example: do you respond to the vignette situation? 1 = not responding; 2 = responding, 3 = doubtful. HOMALS, also known as multiple correspondence analysis, was used as scaling technique (Bijleveld & Commandeur, 2009) for nominal data. This provided insight into relationships between vignette measures and the extent to which nonparental adults would (feel a responsibility to) respond. In HOMALS, similarities and differences between persons and categories are represented by distances between points in a plot (Figure 1), as are the relationships between variables on two dimensions. Responses with variables that often occur together are close to each other, large differences are reflected by a large distance (Bijleveld & Commandeur, 2009; De Heus et al., 1995).

Figure 1. Positions of variables in two-dimensional space.

The analysis resulted in an overview of the discrimination measures of variables (Table 1). The size of the discrimination measures determines the importance of the variables and can be used to interpret the dimensions. Eigenvalues of the two dimensions and a two-dimensional plot (see Figure 1) with background variables (i.e. nonparental adults, ethnic background, sex) show relationships between variables (i.e. child behaviour, response or lack thereof, location, parenting style, responsibility). The background characteristics of the participants negatively influenced the eigenvalues on both dimensions and, hence, the fit of the solution. The characteristics were added as passive variables in the HOMALS analysis for illustration and included in the plot.

Table 1. Discrimination measures per dimension.

Variables	Dimension
	1	2
Response	−.785	−.038
No response	.860	.037
Authoritarian	−.192	.109
Authoritative	1.036	−.251
Permissive	−.756	.111
Internalising	.197	.434
Externalising	−.212	−.437
Around school	.005	−.744
Neighbourhood	−.016	.543
Responsibility	−.770	−.749
No responsibility	1.104	−.313
Doubtful responsibility	−.216	1.054

The qualitative data were used to explore both the participants’ motivations for their responses to the closed questions and any further context highlighted by their answers to the open questions. For the analysis of qualitative data, thematic analysis was used (Braun & Clark, 2006). The interviews were coded by two independent coders in MAXQDA18 (Kuckartz & Rädiker, 2019) with a data- and theory-driven coding scheme with 23 main codes. Examples of codes with closed answers categories are: ‘respond: yes/no’, ‘responsibility: yes/no/doubtful’. An example of codes with open answers is: ‘how to respond and why (not) to respond’. In the first round of coding a data-driven coding scheme was used based on the aforementioned 5 interview questions. Subsequently, a theory-driven coding scheme was used (e.g. Bennet et al., 2012; Geens et al., 2019; Kesselring et al., 2016). The kappa coefficient, which was calculated using a random selection (5%) of interviews and 150 fragments with the main codes, was good (.815) (Viera & Garrett, 2005). From the analysis, overarching themes emerged on how, why and when participants would (not) respond to vignettes with specific factors. The themes were discussed in group meetings with parents, volunteers, and preschool and primary school professionals (N = 23) as a member check.

Results

Multiple correspondence analysis

The HOMALS analysis model was applied to analyse the similarities and differences in participants’ feelings of responsibility and willingness to respond, and various factors (i.e. parenting style, child behaviour, and location). Table 1 shows for both dimensions in the plot discrimination measures of variables with relatively high loadings of ±.60 (i.e. response, parenting style, responsibility, location).

Figure 1 presents a two-dimensional plot with both background variables (i.e. nonparental adults, ethnic background, sex) and relationships between variables (i.e. child behaviour, response, location, parenting style, responsibility) and summarises the relationships between the different variables. On the left side of the plot of dimension 1, the variables ‘no response’, ‘authoritative style’, and ‘no responsibility’ occur together. This indicates that when vignettes included an authoritative parenting style, participants were less likely to respond and felt less responsible. On the right side of the plot of dimension 1, the variables ‘response’, ‘permissive parenting style’, and ‘responsibility’ are clustered, indicating that participants responded more often and felt more responsibility when vignettes included a permissive parenting style. Dimension 1 has an eigenvalue of.347, which indicates a reasonable fit (Bijleveld & Commandeur, 2009).

On the upper side of the plot of dimension 2, the variables ‘neighbourhood’ and ‘doubtful responsibility’ appear together. This shows that the variable ‘neighbourhood’ caused some doubts among participants whether they felt a responsibility to respond. On the lower side of the plot of dimension 2, the variables ‘around school’ and ‘responsibility’ are clustered. This indicates that participants more often felt responsible when vignettes contained a location around school. No relationship was found between response and location. Dimension 2 had an eigenvalue of.228, which is relatively low (Bijleveld & Commandeur, 2005) and indicates that the variables fit slightly less compared to dimension 1.

In sum, dimension 1 revealed that participants’ willingness to respond and feelings of responsibility were dependent on the described parenting style. Parenting style proved a significant variable in whether participants would respond (when confronted with a permissive parenting style) or not (when confronted with an authoritative or authoritarian parenting style), and significant differences between parenting styles and response levels were found, χ² = 73.73, df = 2, p = .000. The location mainly affected the perceived responsibility of participants, but not whether they responded. Analysis did not show a significant difference in response levels between different locations, χ² = 0.20, df = 1, p=.654. To better understand dimension 1 and the significant differences between parenting styles and response levels, a follow-up analysis was conducted.

Follow-up analysis

Permissive parenting style

In vignettes where the permissive parenting style was combined with internalising child behaviour, more than half (57%) of participants considered the parenting style as negative. Participants would therefore respond by starting a conversation with the child or parent, aimed at stimulating the child to go play. ‘I would ask the child (…) “don’t you want to play?” and “shall we do something?”. Try to involve the child a little with the other children. (…) “do you want to play with her?”’ (professional). Participants also said that they would talk briefly to the parent, for example, about the striking behaviour of the child. As one participant said: ‘Because I think: you are staring at the ground, normally all children go completely crazy when they are at the playground [the behaviour of the child is therefore striking]’ (professional). The goal was to positively influence the child’s emotional well-being and support them. Some professionals said they would respond in the schoolyard as they felt they had a responsibility to support parents.

In vignettes which described externalising child behaviour, 59% of participants considered the parenting style as negative. Participants would respond significantly more often in vignettes describing externalising behaviour taking place around school than in vignettes describing internalising behaviour, χ² = 8.95, df = 1. p = .003. Participants mentioned two main reasons for responding. First, if they assessed the child’s behaviour as being inappropriate. Some participants would then discuss the child’s behaviour with the parent and/or child: ‘That bag is not yours, you are not supposed to take it’ (professional). Second, if the child was in danger or was a nuisance to their surroundings. ‘(..) it is dangerous because the parent is not focusing on what the kid is doing and there are other people and maybe they don’t like it’ (parent). When responding to the parent, participants liked to respond in a friendly manner, by having a short conversation about the behaviour of the child, or by expressing support towards the parent. Contact varied from a simple ‘hello’ to short questions.

Authoritative parenting style

In vignettes where the authoritative parenting style was combined with internalising child behaviour, 5% of participants considered the parenting style as negative and 65% as positive. Participants would respond to support child and/or parent, for example, by having a short conversation about the child’s behaviour or by supporting the parent:

I have a very social girl, so I would say to her that maybe the other child would like to play. But sometimes the parent doesn’t want that, then I wouldn’t interfere. But if she wants to, you can lend a hand, right? (parent)

In vignettes describing externalising child behaviour, 17% of participants considered the parenting style as negative and 62% as positive. Participants liked to respond to discuss the behaviour of child and/or parent with the parent, in order to positively influence the situation. This happened, for example, when the child behaviour was considered disruptive, and the parent’s response insufficient: ‘You don’t have to ask the child: “why did you kick me?”, you don’t have to [because it’s too kind]’ (parent).

Authoritarian parenting style

In vignettes where the authoritarian parenting style was combined with internalising child behaviour, almost half (46%) of participants considered the parenting style as negative. In vignettes describing a combination of an authoritarian parenting style with externalising child behaviour, 24% of participants considered the parenting style as negative and 28% as positive. Participants said that they would respond significantly more often to internalising behaviour in the neighbourhood than to externalising behaviour, χ² = 6.78, df = 1, p = .009.

Participants responded in vignettes combining an authoritarian parenting style with internalising child behaviour, because they wanted to clarify the situation and/or because they were worried about the child’s well-being (e.g. ‘I am going to ask her or him (…) what is going on with the child. Does the child not feel like playing?’) (professional). Some participants wanted to support parent and/or child by motivating the child to play and/or by involving the parent. Another important reason for participants to respond was being a role model. When vignettes contained externalising behaviour, participants would respond because they assessed the child’s behaviour as negative. For example, when the child kicked or criticised the parent, this was considered transgressive behaviour. If the parent only responded non-verbally, participants thought the parent did not set enough limits: ‘The child has to learn that there are limits, that she may not say certain things, (…) “you are fat and ugly”, that is very embarrassing.’ (volunteer). Some participants mentioned that responding could provide support for the parent and positively influence the child’s behaviour.

Not responding

Participants would refrain from responding when the parent responded and the child’s behaviour was assessed as normal or positive. Participants would not respond in vignettes describing a combination of authoritative parenting style and internalising/externalising behaviour, because the parent’s response was considered positive. Some participants mentioned that parenting styles could be related to migration backgrounds. For example, an authoritarian parenting style can be considered normal for people with a migration background, while others prefer an authoritative parenting style. As long as there were no unsafe situations, participants did not respond.

Well if there was a reaction from the mother, I would leave it, yes. (…) parents are going to solve that in their own way. I tend to disagree [with their parenting style], but that’s how they do this. That it is not ignored, I think that’s the most important thing. (professional)

Furthermore, participants mentioned that they did not always have enough information about the situation and thought that their reaction could negatively influence the situation or undermine parental authority. As one participant put it: ‘Maybe (…) she has set certain boundaries or there is another reason why the child is angry. Then I don’t want to respond right away, that’s not good for the mother’ (parent). Some participants also found it difficult to assess behaviour. For example, an authoritarian parenting style raises questions: to what extent is the parent approachable? Does the parent always respond like this? These questions could impede contact.

Feelings of responsibility in interactions

Whether participants felt a responsibility in the described situation, and what role they might have, influenced to some extent whether they responded. Participants mentioned two reasons for this. Parents and volunteers said that they felt responsible for observing parenting situations in the neighbourhood. ‘I would observe the mother so she would think: oh, I will be watched’. ‘Being an example’ and ‘creating a safe environment’ were mentioned less often as reasons. Professionals said that they felt a responsibility in situations taking place around school. A majority stated that their responsibility is to observe the situation and set an example. ‘Working together with parents’ and ‘ensuring a safe parenting environment’ were mentioned less often as reasons. ‘I think you should not keep pointing your finger too much after school (…) during school we have rules about safety because we have the 30 children (…) after school you cannot keep an eye on everything’ (professional). At the same time, participants doubted whether they had a responsibility. For example, parents questioned to what extent parents are responsible for bringing up other people’s children after school. The extent to which parents know each other matters. For example, if they are friends or acquaintances, this might be a reason to respond.

Furthermore, sometimes participants had difficulty deciding if and how they would respond, and would therefore not respond, even though they felt a responsibility. ‘I don’t know, I find it difficult (…) I’m not sure what to say to the parent’ (volunteer). Some participants mentioned that it can be difficult to discuss parenting situations when transgressive behaviour (such as shouting or screaming) is being displayed. Others said that the intensity of a situation can make it difficult to make contact with other nonparental adults: they do not always know how to constructively discuss the situation.

How can you explain to parents that certain behaviour is not accepted? We’ve had a situation where a mom was screaming in front of all the kids. You can usually not reach such parents, not at that moment, because then they feel attacked. (parent)

Discussion

This study examined how parenting style, type of child behaviour, and location in the neighbourhood influence whether and how nonparental adults (i.e. professionals, volunteers, and other parents) respond to parenting situations in the form of vignettes. In about half of these situations, they would respond. It is interesting to note that all three factors influence whether and how participants are willing to respond and, as such, may influence social support in the neighbourhood.

In the case of an authoritative parenting style, nonparental adults would refrain from responding, because this style was seen as positive. This is in line with the literature which shows that many adults adopt an authoritative parenting style (Liu & Guo, 2010) and may also reflect the social norms of the participants. Participants felt more responsible and were more likely to respond when confronted with a permissive parenting style, which may be different from their own style or clash with their perceived social norms, particularly in combination with externalising child behaviour and the location ‘around school’. In more than half of the cases participants assessed externalising child behaviour as negative, leading them to respond, while internalising behaviour was not viewed as such. This might indicate that there are higher levels of informal social support around school when externalising behaviour is displayed, which may provide more benefit for children who exhibit this behaviour. In locations where clear regulations and a low threshold are already in place (Doornenbal et al., 2015), such as a school, it might be easier for participants to respond. This finding suggests it is important to align with and strengthen existing structures in schools and other neighbourhood institutions to foster social support (Geens et al., 2019; Holt-Lundstad & Uchino, 2015).

It was found that there is not always sufficient awareness of children’s internalising behaviour (particularly in a school context). A likely explanation is that emotional problem behaviour is not always recognised by nonparental adults (O’Brien et al., 2016). Therefore, more attention and alertness of professionals at (pre)schools may be required when it comes to children’s internalising (problem) behaviour, for example, by reflecting on the influence of different types of behaviour, or by supporting parents and nonparental adults in discussing internalising child behaviour. This finding is in line with the importance of the broader role expected of professionals in strengthening supportive structures in the local community (Doornenbal et al., 2015). On the other hand, when location in the neighbourhood was combined with internalising behaviour, nonparental adults responded significantly more often than when it was combined with externalising behaviour. A possible explanation is that participants seemed to be more motivated to support other nonparental adults in neighbourhood playgrounds. This is consistent with findings that playgrounds can be facilitators for parent interactions (Bennet et al., 2012), which is a precondition for supportive structures (Kesselring et al., 2013) and underlines the importance of community social work in the neighbourhood (Pastor Seller, 2015; Weil, 2004).

This study further supports the idea that the presence of a parent may keep bystanders from responding (Kesselring et al., 2016; Scales et al., 2004) and thus can affect social support. Participants found it difficult to assess these situations and feared a negative reaction from the primary caregiver. Background characteristics were unrelated to nonparental adults’ response, which is not in line with literature about diverse neighbourhoods (Gadsden & Dixon-Román, 2017). More research into the influence of superdiversity (Crul & Lelie, 2019) on social support may help to increase understanding of the context of diverse urban neighbourhoods.

Finally, these findings affirm the importance of professionals, community social work, and organisations in strengthening supportive structures in the community (Geens et al., 2019; Mini & Sathyamurthi, 2017). In interactions that took place around school, professionals felt more responsibility than parents and volunteers. This suggests that in social work organisations more attention for professionals’ feelings of responsibility and for perceived barriers to intervene is urgently needed.

These results emphasise the need for discussion about the response and responsibility of nonparental adults (Kesselring et al., 2016; Scales et al., 2001) to strengthen social support in the neighbourhood. In addition, the results show that nonparental adults do not always know how to address (negative) behaviour of parent and/or child in neighbourhoods, although this is considered important. The developed vignettes provide an opportunity to further examine, reflect on, and discuss (Bernabeo et al., 2013) the connections and differences between the perceptions and responses of professionals and other nonparental adults, and may be a basis to further develop community interventions (Pastor Seller, 2015). Strengthening nonparental efforts in the community seems a promising intervention for increasing social support and children’s well-being (Holt-Lundstad & Uchino, 2015; Scales et al., 2004).

Limitations and future research

This study is not without limitations. A first limitation is that participants might respond differently to vignettes than they would in reality and may also have given socially desirable answers as vignettes are obviously self-report measures. It should be noted, however, that analysis of their responses showed that participants did not give uniform answers, which suggests that the responses have not been affected by a strong bias to give socially desirable answers. Acknowledging the limitations of self-report measures, this study provided insight into important factors influencing nonparental adults’ attitudes towards neighbourhood interactions.

A second limitation is that the vignettes provided only limited information. For example, personal characteristics of the main characters (e.g. parent, child) were only briefly described. In order to overcome these limitations, a qualitative method was used to gain insight into participants’ underlying feelings and thoughts, and to better understand how their experiences related to their responses.

Finally, the study included a convenience sample from a specific urban neighbourhood. Examining a diverse sample from more urban neighbourhoods and rural areas would complement the current findings.

Conclusion

This study shows that the type of parenting style, child behaviour, and location influence neighbourhood interactions and whether and how social work professionals, volunteers, and parents (feel a responsibility to) respond to parenting situations. By systematically combining these different factors it was found that supportive interactions of nonparental adults are more likely to take place in case of a permissive parenting style and externalising child behaviour around school. The newly developed mixed-method approach of this study with vignettes and interviews can be used to discuss, and reflect on, social neighbourhood interactions, and may contribute to the development of social support interventions aimed at the positive involvement of stakeholders in a community and social work setting.

Acknowledgements

We thank Esmee Verheem, Linda Kroeze, the participating organisations and students who assisted in the data collection for their contributions to this project.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Research data are not shared.

Additional information

Funding

This work was supported by Netherlands Organisation for Scientific Research [grant number 40017603]; and the Amsterdam University for Applied Sciences, Centre of Expertise on Urban Education.

Notes on contributors

S. M. Rumping

Sanne Rumping, MSc, is researcher at the Amsterdam University of Applied Sciences and PhD student by the Research Institute for Child Development and Education at the University of Amsterdam, the Netherlands. Themes in her work are, intervention development for youth social work, implementation, collaboration between professionals, parents and volunteers and urban education.

L. Boendermaker

Leonieke Boendermaker, PhD, is professor implementation and effectiveness in youth care services at the Amsterdam University of Applied Sciences. Her work focusses on the development, implementation, quality and effectiveness of interventions in youth care services, with special attention for residential care, interventions for young people with serious behavioural problems and urban education.

R. G. Fukkink

Ruben Fukkink, PhD, is professor of Pedagogy at the Amsterdam University of Applied Sciences and is also professor by Special Appointment of Childcare and Education Services for Young Children in the Faculty of Social and Behavioural Sciences at the University of Amsterdam, the Netherlands. His work has appeared in various peer-reviewed journals. Special themes in his work are: early childhood education and care, parental support, urban education.