Positive relationships: All about ... Childhood friendships

Children go through recognisable developmental stages in their ability to make friends, which nursery life can foster, writes Philip Erwin

Children's friendships, and peer relationships in general, do not exist in isolation or suddenly appear, but rather originate in and are built on the child's earlier relationships within the family.

Children's relationships with the primary caregiver and the rest of the immediate family provide an experiential foundation and set the tone and expectations children hold for their future peer relationships.

Research by Alan Sroufe and his co-workers has shown that infants with a secure attachment to their caregivers do subsequently tend to be more effective with their peers, an advantage carried forward throughout the childhood years.

These children appear confident and curious, skilful, socially oriented and empathic, and act in a co-operative, friendly and outgoing manner to other children. All of these characteristics can play an important role in promoting and sustaining children's friendships, and securely attached children consequently tend to be more popular and have more friends. But as we shall see, psychologists have come up with strategies that can be used to help children who are struggling to improve their friendship networks.

Functions and environment

Research suggests that friendships serve some important functions for the child. They offer the child the opportunity to hone peer relationship skills, exchange and test social knowledge as they compare themselves to others, promote confidence and comfort in intimacy, buffer and support the child from life's emotional up and downs, and generally stimulate social-cognitive development.

In the experience, the child receives positive stimulation and becomes a social star, the centre of positive attention in the microcosm that is their friendship. These experiences are important for increasing self-esteem.

It is worth stressing that although these are a general set of functions that relationships will serve for all children, specific functions may be more or less important for any given individual and this may be reflected in the way they relate to their peers.

It also follows from this that environmental conditions may also make some needs more salient than others for the child. For example, fear, anxiety or uncertainty in general may increase the desire to be with others, to try to understand a situation and work out what is an appropriate response.

These discussions about the relationship between the environment and the functions of children's relationships raise questions about how children's relationships change as they develop, and the role of social functions and environment in promoting the development of peer relationships.

Non-social activity

Peer interest is evident even in children only a few months old, though non-social activity occupies more of a child's time than social activity at this age. Despite the evident peer interest of young children, their social networks are limited by their dependency on their parents to arrange contacts, venues, times and transportation to meetings with peers.

Young children lack the social skills to sustain social interaction, particularly with equally unskilled peers who cannot accommodate and adapt to their shortcomings.

In children of six months to one year of age, slightly less than half of attempts at peer interaction succeed in achieving co-ordinated social behaviour, and even these generally consist of a single action-response sequence.

At this age, children look at each other, touch and imitate each other's activities, and vocalise and smile at each other.

Infants who have more turn-taking experience with skilled partners, such as older siblings, tend to show a more extensive repertoire of skills with peers. Peer experience increases the level and complexity of involvement that can be attained with familiar peers and does show some generalisation to unfamiliar peers.

More than half of all pre-school children do manage to establish stable, reciprocal relationships with peers, though at this age friendships are relatively unstable. They readily collapse but are even more easily formed. The defining characteristic of friendships at this age is that there is more positive behaviour to and from friends than non-friends. Children expect more positive, rewarding behaviour from their friends and regard peers that treat them this way as friends.

There is an extensive literature on the role that objects play in promoting and sustaining children's relationships. With those of six to 12 months of age, peer interactions without toys tend to be quite brief and often end with at least one of the children showing signs of distress.

With the availability of toys to play with, any interaction that does occur may be as the children play independently but are influenced by each other; they can be seen to exchange glances from time to time, watching and even imitating each other's actions and sometimes sharing and exchanging toys.

This is a much more harmonious and easily managed style of interaction for young children. For young infants, toys are a useful adjunct to social contact with peers which ease the strain of attempting to maintain complex, extended sequences of interaction with them.

Social interactions

In contrast to younger infants, toddlers may make much more direct use of a joint orientation to toys and play materials in their social interactions. From about one to one-and-a-half years of age, parallel play becomes a prominent and distinctive feature of children's peer interactions.

'Parallel play' is the term given to the situation in which children may play with toys near to and in parallel with other children, but independent of them. Initially, the level of co-operation may be relatively low and occasional, but this increases substantially throughout the succeeding years of childhood as genuine peer relationships begin to emerge and develop.

Parallel play serves an important function in children's relationships as it supports the development of more complex, sustained types of interaction, allowing children to gradually move from playing alone to joining in with another.

Although these early social contacts are relatively superficial and transient, they are an important precursor and building block for developing the more complex and sophisticated relationships of later childhood. Parents who provide their children with the opportunity for enriching experiences are giving them an important headstart on their less fortunate counterparts.

Social networks

With age, the child's social network includes proportionately more peers and there is more daily contact with their peers. Research by Candice Feiring and Michael Lewis examined the social networks of children aged three to nine years old and found that even children as young as three often have fairly extensive social networks - on average 22 social contacts - and this increases with age. About a third of a child's social network may be comprised of peers.

Children start to use the term 'best friend' at about the age of four. Robert Hinde suggests that at this age, most children will be involved in at least one close, reciprocated peer relationship - though this may be almost the full extent of the friendship network. Throughout the school years there is a greater opportunity and ability to establish and maintain friendships, and they increase in number from five or six at age six, to perhaps nine or ten by age nine. This does not, of course, mean that younger and older children attach the same meaning to the term 'best friend'!

Throughout middle childhood best friends may be chosen because of the demands of specific situations or occasions. As examples, children may have different friends at school and in the home neighbourhood, and specific individuals may be chosen when certain games are played. At this age many friendships are still relatively transient, but they are also formed easily, and overall children gain more friends than they lose.

Research using very different methodologies (for example, from Robert Selman, and Brian Bigelow and John La Gaipa) has come to encouragingly similar conclusions about the developmental stages of children's understanding of friendship. Unfortunately, though, there is little research examining how they affect the dynamics of actual relationships. As Brian Bigelow has noted, there is often little or no correspondence between why children say they behaved in a certain way and the actual factors that influenced their behaviour.

Major stages

Typically, three major stages are identified in the development of children's friendships, broadly mirroring those of cognitive development in general.

An initial 'egocentric' or 'situational' stage lasts until around seven or eight years old. The child at this stage is self-centred and concerned with their immediate situation. There is an emphasis on the external characteristics and behaviour of others.

Friendship expectations at this age emphasise rewards and costs, proximity, shared activities, physical appearance and possessions. Having an attractive toy or sweets is likely to be a great boost to peer popularity! The use of apparent trait terms (such as 'kind' or 'mean') are used as descriptions of recent behaviour, rather than representing insights into personality.

A second 'sociocentric' or 'normative' stage of friendship lasts until about 11 years old. The prime characteristic of this stage is an emphasis on sharing and the strict adherence to the rules and obligations of relationships. Friendship expectations are also starting to include a greater emphasis on psychological dimensions, on inner traits and qualities. Typical statements might include 'she shares more things' or 'he's big-headed'.

The final, 'empathic' or 'internal-psychological' stage is a continuum that stresses the increasing differentiation and organisation of children's friendship expectations, making them extremely complex.

There is an emphasis on intimacy and trust, self-disclosure and the abstract psychological characteristics of friends. The adolescent can comprehend and cope with inconsistencies and exceptions to the general characterisation of a person. Without too much effort, the adolescent can accept that an individual can be both 'kind- hearted' and 'catty sometimes'.

Making friends

In addition to their efforts to understand the pattern and bases of children's friendships, psychological researchers have also been keen to apply their knowledge to improve the lot of children who have problems relating to their peers and establishing or maintaining friendships.

One particular approach, Interpersonal Cognitive Problem Solving or ICPS (subsequently renamed 'I Can Problem Solve'), was developed in America by George Spivack and Myrna Shure.

This approach works on the assumption that many children fail to function effectively in social contexts not because they lack an appropriate behavioural repertoire, but because they lack the relevant social problem-solving skills to guide and direct their behaviour.

At its most basic level, this revolves around a child's ability to understand and anticipate the consequences of their actions in social encounters. Without this ability, the child is simply trusting to luck that the way they try to relate to others will be successful.

A major advantage of ICPS over some other programmes is that it can be incorporated easily and naturally into everyday interactions in the nursery or primary school classroom.

Although there are a wide variety of social problem-solving skills, among the most commonly researched have been 'Alternative Solutions Thinking' (the capacity to think of alternative solutions to resolving difficult social situations) and 'Consequential Thinking' (the ability to envisage the potential consequences of the various solution strategies).

Children generally enjoy working in groups to practise these skills, and research has shown that even relatively small numbers of sessions can produce significant improvements in children's ICPS skills and their levels and patterns of interaction with peers - an improvement most marked in children who initially seemed most deficient in these skills.

It is encouraging to note that although systematic research on children's friendship is still a relatively recent endeavour, dating mostly from the 1970s, it is already evident that it has many insights to offer people working with young children and much potential to improve children's lives.

- Professor Philip Erwin is programme leader for Psychology, Department of Social and Psychological Sciences, Edge Hill University, Ormskirk

References and further reading:

- George Spivack & Myrna Shure (1974), Social Adjustment of Young Children. San Franciso: Jossey-Bass

- Judy Dunn (2004), Children's Friendships: The beginnings of intimacy. Oxford: Blackwell


Within the broad pattern of development in friendship expectations, certain factors influence a child's specific friendship choices.

Children appear to prefer peers at a slightly more advanced developmental stage than themselves, though it is, of course, a logical impossibility that every child can be a friend with a more advanced peer. When left to their own devices most children seem to have their closest friendships with peers at a similar developmental level.


The impact of gender on children's friendship choices and patterns of relating has been well researched. There is a very visible sex segregation in peer relationships that increases through early childhood and is almost total by middle childhood. This schism is both in terms of actual friendship choices and in the style of relating that can be seen. Mary Waldrop and Charles Halverson have termed the different patterns of relating as 'extensive' or 'intensive'.

'Extensive' vs 'Intensive'

The intensive relationships of girls are likely to focus on a single best friend, and are characterised by intimate conversation, personal knowledge of the friend and emotional involvement. Boys' relationships are characterised as extensive, because of their tendency to spend time mostly with a larger group of peers which may be fairly heterogeneous in terms of age and gathered together to engage in a specific task, game or activity.

Evidence suggests that socialisation influences at home and at nursery and school can reinforce or diminish this sex segregation.

Nursery World Print & Website

  • Latest print issues
  • Latest online articles
  • Archive of more than 35,000 articles
  • Free monthly activity poster
  • Themed supplements

From £119 per year


Nursery World Digital Membership

  • Latest digital issues
  • Latest online articles
  • Archive of more than 35,000 articles
  • Themed supplements

From £119 per year


© MA Education 2020. Published by MA Education Limited, St Jude's Church, Dulwich Road, Herne Hill, London SE24 0PB, a company registered in England and Wales no. 04002826. MA Education is part of the Mark Allen Group. – All Rights Reserved