The Dance of Friendship

A paper by Penny Haworth and Joy Mepham, Massey University; Eileen Woodhead, Napier Kindergarten Association; Heather Simmons, Liz Schimanski and Pam McGarva, Wycliffe Nga Tamariki Kindergarten.  This report describes the first phase of their Round one Centres of Innovation (COI) project. The phase was undertaken in the second half of 2003 and focused on children's friendship interactions.

Background
Wycliffe Nga Tamariki Kindergarten provides five morning and four afternoon sessions a week, each attended by thirty children. During the afternoon sessions, which became the focus for the study, twenty-six children became participants: eleven Samoan children, nine Māori children, and six New Zealand European children. Only the Samoan children could be classified as bilingual speakers. In particular, the context for the afternoon sessions has been influenced by a collaborative relationship that has developed over a number of years between the kindergarten and a neighbouring unlicensed playgroup run by the Samoan Assembly of God church. Children from the Samoan playgroup came to the kindergarten during the afternoon sessions, resulting in increased participation and interest from Samoan parents who have become part of the Wycliffe Nga Tamariki (WNT) community of learners. The Ministry of Education (2002) affirms such collaborative relationships, in particular with Pasifika communities, noting that, "Children's learning and development are fostered if the well-being of their family and community is supported; if their family culture, knowledge and community are respected and if there is a strong connection and consistency among all aspects of the child's world" (p. 16). The current project therefore contributes to the knowledge about practices that build and maintain such links.

Over time, the teachers at the kindergarten have also become increasingly aware of how the partnership between the kindergarten and the Samoan community provides enhanced opportunities for the learning and development of all children. As the afternoon sessions in the kindergarten are rich in Samoan, Māori and English languages, all children in the WNT Kindergarten's learning context are exposed to, and gain from, the diversity of linguistic and cultural input. Teachers believe that Samoan children achieve greater educational outcomes when they learn in a setting that embraces their language and culture (Fleer, 2002) and can more easily acquire English as an additional language when actively engaged in a natural play-based programme (Baker, 1996; Krashen, 1982).

Teachers at WNT Kindergarten regularly use Information and Computer Technology (ICT) to involve children and their parents and families as part of the community of learners within the kindergarten. Each day's events are recorded on digital camera and as stories which are made available at the end of each session. Families are therefore able to share in their children's learning as they see the photographs and read the stories that have been recorded on the computer. This has been found to be an ideal way of communicating with parents for whom English is not their first language. Yong and McMullen (2003) also discovered that new technologies could provide an alternative communication channel for parents from non-English speaking backgrounds, thus connecting them to the early childhood learning community.

From its inception, the project has been guided by one over-arching question: 'In partnership with Upu Amata playgroup, how can learning and teaching be improved for all children in the intercultural setting of the afternoon sessions at Wycliffe Nga Tamariki Kindergarten?' Of special interest to teachers in the WNT context were the cultural and linguistic elements of play and the mix of languages that might occur as friendships developed and children became "master players" (Harley, 1999). Therefore, the initial phase of the project, which focused on children's friendship interactions, was especially valuable in contributing to teachers' understanding and knowledge of how a community of learners is created and enhanced within a multicultural setting.

Why focus on exploring children's friendships?
In children's terms, friendship is defined as "a need to belong, to feel attached to someone" but young children are most likely to see friends as "playmates" and this is especially associated with fantasy play (Haslett and Samter, 1997; Rizzo, 1989). Friendship interactions are vital for the healthy development of children (Haslett and Samter, 1997), and may result in heightened self-esteem, positive self-identity, effective communication, as well as successful cognitive, social and conceptual development. Furthermore, the opportunity for children to engage in cross-cultural friendships may enhance cultural tolerance and acceptance.

The promotion of friendships between children within Early Childhood Education (ECE) reflects an effective interweaving of the five strands of well-being, belonging, contribution, communication and exploration in Te Whāriki (Ministry of Education, 1996) with the key principles of the document: holistic development, empowerment, family and community and relationships. Through children's friendships in the kindergarten setting, positive outcomes emerge for all members of the community of learners, including teachers, caregivers, and children. Families within a community of learners can be pulled together through children's friendships, crossing cultural barriers, extending family support networks, and allowing children to explore, learn and grow in a caring, secure environment that promotes positive social values and skills. Learning is thus encouraged in an environment where friendships flourish. Furthermore, Vygotsky (1978) holds the view that effective interactions often assist in constructing meaning and can encourage children to be co-creators of knowledge.

Another effect of a friendship-filled environment is that it minimises anxiety, so opportunities for natural communication are increased. Krashen (1982) believes that bilingual development is promoted in such situations, where the affective filter is lowered. Research into the reasons for code-switching (changing languages while speaking) indicates that bilingual speakers often change languages when interacting in order to convey friendship (Baker, 1995, 1996). Despite some earlier research findings indicating that friends tend to share physical, attitudinal, and personal characteristics such as age, gender, race and IQ (Rizzo, 1989; Haslett & Samter, 1997), it is thought likely that, in a multicultural setting, young children's friendships may promote qualities such as loyalty, acceptance and sharing (Rizzo, 1989). It is also significant to the present study that the sharing of languages and the extension of worldviews resulting from these cross-cultural liaisons may be influential in long-term learning success (Baker, 1996).

A further dimension of friendship is its capacity to enhance the role that dramatic play has in the development of effective communicative strategies. Dramatic play may involve situations where role-playing becomes more complex, including dressing up, development of dialogue and the creation of environments whereby different roles can be taken on and familiar scenes and stories re-enacted with friends (Haslett & Samter, 1997; Bodrova & Leong, 2003). One special benefit of dramatic play in friendships is that it enhances the child's motivation; facilitates cognitive decentering, (i.e. it allows the child to take on another child's perspective, leading to the development of reflective meta-cognition); advances the ability to operate with symbolic representations for real objects and this contributes to the development of abstract thinking and imagination; and in this sort of play, partners (friends) monitor each other to make sure all players are following the informal rules they have set (Harley, 1999; Bodrova and Leong, 2003).

Research shows that even the interactions of very young children of two and a half years can often demonstrate sustained attention, an awareness of turn-taking, and reciprocity and that their interactions may be strung together into complex routines involving activities such as taking on other roles (Haslett & Samter, 1997). However, young children's friendships can also be fragile, fleeting encounters (Rizzo, 1989). Ongoing reflection on these issues and their impact on teaching and learning became part of the first cycle of the action research. In particular, this phase asked:

1. How are children moving in and out of groups?

2. Do non-verbal communications initiate and sustain friendships and if so, how (e.g. smiling, touching, gestures)?

3. Does code-switching (or changing between various languages) occur in the forming of friendships?

4. What other factors contribute to making friends?

How did we explore children's friendship interactions?
As with the other COI, this project takes a participatory action research approach that utilises contextualised problem-solving (Wadsworth, 1998) so it is seen as especially relevant to teachers in a practical setting (Bell, 1997; Burns, 1999). Such a process is a cyclical one which begins with a question about practice being posed; observation and reflection follow; actions are identified and implemented; and observation and reflection on these actions leads reiteratively into further investigative cycles (Wadsworth, 1998). Kemmis and McTaggart (1988) suggest that the distinguishing feature of action research, as opposed to the general on-the-run problem solving that teachers engage in their daily practice, is the formal recording of information, the reflection on and analysing of collected information and the subsequent writing and presenting of what happened as a result of the process. Action research therefore focuses on enhancing both children's learning and teachers' reflective practice. Eventually, a further outcome from the COI will be publications like this, which open new theories and knowledge up to wider scrutiny and make these available for others to use in, or adapt for, their settings.

The teaching team is regarded as the significant research partner; however, the teaching team work with support from the Education Manager of the Napier Kindergarten Association. Two research associates from Massey University also provide support and guidance for the project. Members of the research team meet at key points in the research process and collaboratively analyse data in order to co-construct theories and knowledge which ultimately influence practice. In addition, findings are reported on to a focus group from the wider educational community. Such an approach fits Donaldson and Sanderson's (1996) model in which "the work of collaboration differs from solo work because it is accomplished, not first in one person's mind, and then in the other's, but on the loom between them, in the centre of their joint space" (p. 44). Such a weaving image is reminiscent of the inter-relationship between the strands in Te Whāriki, the Early Childhood Curriculum (Ministry of Education, 1996).

Children's friendship interactions were documented during the final term of 2003. Over the course of the term, each teacher took a turn at being a researcher for a week. Each child was observed for a full afternoon session. These observations were recorded in a narrative style, similar to that used by Cullen and Allsop (1999). As recommended in Sharman, Cross and Vennis (2000), information was also included on the time frame, speakers, and contextual factors. Once interactions had been documented, teachers then highlighted incidents they regarded as critical for a child's friendships on the day of recording. To do this they initially drew on friendship indicators identified by Rizzo (1989), progressively refining these and making them relevant to the data they had gathered (see Table 1).

The importance of critical incidents in provoking changes to teachers' practices is widely supported in the literature on reflective teaching (Alton-Lee & Nuthall, 1992; Hargreaves, 1994; Richards and Lockhart, 1994; Wragg, 1994). The records of each child's interactions were thus entitled Friendship Maps; but it should be noted that these are just snapshots of each child's friendships located in a specific time and space. The children's Friendship Maps therefore do not provide information similar to a sociogram of friendship interactions within the kindergarten, nor do they provide a holistic picture of any individual child's friendships. As Sharman, Cross & Vennis (2000) note, it is not possible to provide a definitive description of the children's friendships in general from this sort of limited data, although increased insights can be gained into these areas. An example of a Friendship Map is provided in Appendix One.

A sensitive audio recording device was fitted to the child being tracked. This equipment had been used prior to the data gathering to familiarise the children and staff with the procedures. Some difficulties were encountered, however, with the transcription of the Samoan children's interactions. As Samoan is a highly context-embedded language, relying on the immediate to make sense of the verbal, it was often difficult for an interpreter to understand a child's meaning if he/she had not witnessed the event in context. The task of transcribing was further complicated by the amount of code-switching between English and Samoan languages that took place in the course of children's interactions. Nonetheless, some interesting conversational data was obtained through the personal microphones used to record the children's verbal interactions.

Learning stories were also gleaned from the teachers' observations of children's friendship encounters. These were included in children's portfolios and later supplemented by parents' voices. In addition, teacher-researchers kept personal reflective journals which provided an opportunity to note important points about the context on the particular day, to engage in some preliminary data analysis, and to record notes about any changes to the routine. These additional sources provided teachers with greater contextual information on the data, and were often referred to when discussing findings with the wider research team.

The researchers also kept in mind key ethical principles, including respect for personal dignity, individual beliefs, privacy and autonomy as well as the need to minimise any potential for physical, emotional, or spiritual harm to participants, researchers, communities or institutions¹.

Consideration of how to avoid possible conflicts of interest was also an important issue for teachers who take on an additional role as researcher. Teachers decided at the start of the project that, although their roles were to expand beyond teaching to research, the children's learning needs and welfare would remain paramount.

Samoan parents were provided with translated information and an interpreter supported the informed consent process. In particular, all parents and other participants were told of the voluntary nature of their participation, and that all data collected involving children or their families and any information included in children's portfolios remained the property of the child and their family. Therefore, information was only used with family consent and the names of children used in these reports have been changed to protect identities. Lists of ethnically suitable pseudonyms were provided by a Māori member of the research focus group and by members of the Samoan Upu Amata community.

What did we find out about children's friendships?
Initial discussions contributing to the data analysis were held over three team meetings involving teacher-researchers and the research associates between October and December of 2003.

As a result of reflecting on children's Friendship Maps, a number of verbal and non-verbal friendship indicators were identified (see below). Teachers decided that friendships could be subjectively recognised through the presence of one, or a combination of several of these indicators. This extends earlier work on friendship indicators by Rizzo (1989), and shows that friendships are certainly seen by children as 'the people I play with'.

Friendship indicators at Wycliffe Nga Tamariki Kindergarten

• the use of and calling out of names
• greetings
• smiles, laughter and fun sounds
• making jokes and silly faces
• inviting others to join play
• close proximity
• holding hands
• chants and singing
• touching (e.g. putting an arm around a child's shoulder)
• mirroring actions and following
• mirroring language
• talking to one self (rehearsing)
• socio-dramatic play (e.g. dressing up, using props)
• sharing active experiences
•  sharing or making space and time together
• asking questions
• accepting an offer or a friendly gesture
• gifting an object
• helping others
• looking, watching and waiting
• establishment of intimate zones (e.g. the barrel swing).

After further reflection, it became evident that children's friendship interactions occurred at different levels (see below). At one end of the spectrum, friendships ranged from passive non-verbal actions such as smiling and watching, to more active friendship overtures such as offering an item to someone. At the other end of the spectrum, friendships ranged between passive and pro-active collaboration. In the midst of this, some friendship indicators seemed to be imbued with a sense of familial belonging, such as laughter, singing, and increasing physical closeness. In contrast to the friendship overtures (which were often not reciprocated), friendship interactions involving collaboration (such as negotiating turn-taking, discussing concepts and helping others) were more typically reciprocal. Ultimately this detailed picture of children's friendship interactions was described as The Dance of Friendship, a title which vividly captures the essence of children's friendships in action. Just as in dance, so friendship interactions varied in tempo, rhythm, volume and harmony.

Indicators of different levels of friendship

Passive overtures
Indicators:

• Turning towards/looking /watching

Active overtures
Indicators:

• Smiling
• Calling a child's/adult's name
• Approaching/moving closer
• Greeting
• Offering an item
• Showing an interesting item (Hine: look at my plaster; Luka: look at my police jacket)
• Showing an impressive action (e.g. look at me - swinging upside down on bars)
• Dressing up (another persona perhaps gives a child more confidence in friendship overtures, e.g. Luka repeats, 'Leoleo' [police] showing off jacket)
• Introductions
• Copying/shadowing

Familial acceptance and belonging
Indicators:

• Playing/doing things alongside others
• Laughter
• Sitting close (on mat or during story reading)
• Responding to greetings and overtures
• Singing and chanting (either to oneself, or in concert with others)
• Holding hands/hugs/cuddles/sitting on lap

Passive collaboration
Indicators:

• Accepting help
• Accepting items offered
• Accepting rejection
• Automatic turn-taking
• Making space for others to join/come alongside

Pro-active collaboration
Indicators:

• Negotiating turn-taking (children need to be assertive within socially acceptable boundaries - perhaps also linked to leadership)
• Expressing concern for others
• Giving instructions to others
• Helping others
• Translating for others (e.g. Peteru provides Samoan words for song to a peer - 8, 19)
• Providing English words for others (eg Tama gives the word for head)

Teachers' initial findings about children's friendship indicators are summarised below.

Elements of friendship at Wycliffe Nga Tamariki Kindergarten
Friendships, in the context of Wycliffe Nga Tamariki Kindergarten, are:

1. Not time bound, vary in length, and may be transient

2. More likely to develop when children perceive the situation to be non-threatening

3. Often indicated by acceptance of and joining in with another person's play activities

4. About having fun together

5. Sometimes the result of perseverance or desire for friendship

6. About sharing

7. Not always developed sequentially from solitary to parallel to cooperative

8. Influenced by context:

a. Absence or presence of regular friends

b. Weather, e.g. inside/outside

c. Adult presence

d. Children's choices of activities: outside/inside, book corner, drinks trolley, slide, sandpit, monkey bars, swings, clay, mat time, music

9. Vehicles for language development and intercultural contacts

10. May work towards excluding or including others.

Individual reflections on data
The initial findings were extended as teachers and researchers reflected further on the data, as seen in their individual writing that follows. As the data was extremely rich the research team decided to meet again in early 2004, to continue the analysis. At this meeting, key themes were identified as being of interest to individual members of the team. The teacher-researchers then met in smaller groupings, and finally reflective pairs.

The first action research cycle culminated in individual writing by each member of the research team, showing the wide range of perspectives that were taken on the data about children's friendship interactions. The three short articles that follow illustrate the teacher-researchers' individual reflections on the data. Heather begins by considering if the timing of routines during the afternoon session had any influence on children's friendships. Pam follows with some thoughts on children's friendship non-verbal overtures and the influence of gender and age on these interactions. Finally Liz reflects on the frequency of friendships among mixed age groups of children.

Articles
The Time of Day and the Types of Interactions Children Have - By Heather Simmons
When I examined the Friendship Maps for each child I particularly looked to find:

• patterns of different interactions happening at different times during the session
• any differences in interactions after the group time (mat time, which happens at about one and a quarter hours into the two and a half hour session)
• when sustained play with others was happening during the session
• whether there were any links between the activities the children were involved in and the interactions the children were experiencing
• whether the Samoan children were mixing with the other children throughout the day, or whether they started the afternoon with the other Samoan children and branched out as the session continued.

These patterns seemed to have importance in the search for ways to enhance children's friendships and thereby include them further in the community of learners. My reflections especially relate to the settling-in period, catering for quiet children, the unified arrival of the Samoan children from Upu Amata, the presence or absence of special friends, the timing of sustained play, and the effects of group time on friendship interactions.

Settling-in period
During the initial settling-in period at the beginning of the session the children appeared to spend time making connections with a number of others. A lot of interactions during the settling-in period were of a very fleeting nature. Many examples in the data illustrate the different strategies the children used to get noticed by others during this time, such as:

• Introducing themselves to others by name, e.g. "Rupe has just joined the group, and Tama tells her that his name is Tama."
• Doing silly things to make the other children at the activity laugh, e.g. "Salesi goes in to the adults' toilet and comes back out. All the children laugh, so he does it again, going in and out of the children's toilet."
• Speaking to all the other children involved at the activity, e.g. "Cameron arrives at the sandpit pulling a trolley. Callum begins to talk to Cameron about the trolley, then turns to Litia who is beside him and talks to her as well before he goes back to digging in the sand."
• Choosing activities where other children are already involved, e.g. "Pania approaches the screen printing and watches Malia. Pania puts herself very close to Malia, she leans in so that she is part of the action."

Several children appeared to be less confident about making approaches to other children at the beginning of the session. Many examples occurred of these children watching others in a variety of activities during this part of the session. For instance, "At 1pm Leighton was at the water wall watching what the other children are doing. Children move in and out of this activity. Leighton watches quietly and makes no verbal contact." It is interesting that children like Leighton were more confident about becoming involved with others later in the session. For example, "At 2.09pm Leighton is at the slide. Hine joins him and smiles. They both begin to laugh and Leighton follows Hine back up the slide."

The Samoan children coming together from Upu Amata
As the group of Samoan children arrived together directly from the Upu Amata pre-school most of their interactions in the early part of the session were with each other. Some of these children appeared to seek out the activities where the other children are during the settling-in period. For instance, "At 1.20pm Semo is washing the trolley wheels. He is working alone but moves up to Silei, Ioane and Salesi (all children from Upu Amata pre-school) in the water trough. He looks and smiles at them as he moves around."

As the session progressed, the Samoan children seemed to have more sustained interactions either with their Samoan friends, or with other children. Regardless of how involved this group of children were in sustained play they still maintained contact with other children/adults in the larger group by calling out to them. For instance, when Adam and Semo were using the trolley they were calling to each other by name as they went.

Special friends
It appears that some of the children would start the day initiating interactions with many different children. If a child found a special friend for the day, he/she would then stay involved with that one friend for the rest of the session. The friends would then move together from one activity to the next, or one friend would follow the other to a new activity. An incidence of this occurred when Nyra joined Hine in the barrel swing at 1.40pm. At 2.00pm they got off the swing, got a drink together, and walked inside.

Sometimes a pair of friends would subsequently join a group of children playing together, or other children would join in with the pair's existing play. This only appeared to happen later in the session after the settling-in period. For example, "Luka and Semo return to their play with the trolley straight after mat time. Eru approaches and the boys let him climb in before they are off. Soon Sina comes up alongside and starts to help push. Luka moves over to make room. They talk while they push."

Sustained play at one activity
Some children did not meet up with a special friend even after making initial contacts. These children then appeared to become involved in one activity, and stayed at this one activity regardless of who else was present. These children interacted with other children or adults present at the activity, but remained working by themselves or alongside others. Malie's observational record provides an effective example of this:

• 1.45pm - Malie goes to the sewing table and works with the teacher
• 1.50pm - Malie is at the sewing table with Hira
• 2.05pm - Malie is working at the sewing table with two teachers
• 2.38pm - Malie is at the sewing table with Nyra
• 2.39pm - Malie is working alongside Adam at sewing table
• 2.39pm - Malie is at the sewing table with Nyra.

The effects of group time on interactions
Children who had already established a friendship grouping, playing with one special friend or within a small group, would choose to sit with these peers during mat time and continued interacting together during the group time. For example, Salesi sits very close to Taai on the mat. He later turns to face Taai while he sings Grace.

Children who were already making many interactions would continue this pattern during mat time. For instance, Mareta is quiet for the first ten minutes of mat time, then, when it is time for fruit, she laughs with Callum, Semo and Luka as they make funny faces with the orange skins across their teeth. In contrast, if children had been involved in sustained play with others prior to group time they appeared to rush back to this play afterwards. For example, "Hira, Adam and Tane have been playing in the sandpit, making a supermarket with a teacher. They go to the mat together holding hands. They talk together about the supermarket, and after mat time they walk back to the sand pit together." Such patterns indicate that mat time is merely an interlude in children's sustained play patterns. After mat time children often continued the play begun before mat time, and with the same group. In contrast, children who had not planned what to do next would start at another activity at the conclusion of group time and make new interactions with the children there.

Heather's concluding reflections on learning and teaching
It appears that all the children go through a time at the beginning of the session when they make a lot of interactions with others. Some of these interactions can be fleeting. Some of the children continue this pattern for the whole session. Others appear to become more involved with others during the session and have more sustained interactions with either one friend or within a group.

The Samoan children, who all arrive together at the beginning of the session, do appear to spend this settling-in period within the group. It seems that they seek out each other, but as the session moves on they may become more involved in sustained play with other Samoan children or children from other ethnic backgrounds. Later in the session, it appears that, for some of the children, the activity appears to be as important, if not more important, than the other children involved.

From looking at the Friendship Maps, the timing of the group time appears to make little difference in the type of interactions the children make with each other before or after being part of the larger group. However, it must be relevant that mat time happens each day after the children have worked through their settling-in period.

Investigating Friendship Overtures - By Pam McGarva
As a new researcher faced with so much data and so much information I initially found it was difficult to know where to start and what to include, but I began by looking at the role of verbal and non-verbal friendship interactions. This led me to consider what might influence friendship overtures in younger and older age groups, and whether gender differences might have an influence on friendships.

The role of non-verbal strategies in friendship interactions
I divided friendship interactions into verbal and non-verbal. I found that approximately 60% of the interactions between the children were non-verbal so I made a list of all the non-verbal strategies used by the children (17 in total):

• playing alongside and looking at the other child
• following another child
• watching and smiling
• sitting closely
• gaining eye contact and smiling
• making a child laugh, clowning
• taking another child's hand
• making space to let someone sit down
• showing something without saying anything
• imitating/copying actions made by another child
• pointing to something that belongs to another child
• watching, listening and nodding
• pushing another child on a swing without being asked
• rubbing another child's head in a friendly gesture
• stretching a hand out to another child
• raising eyebrows in greeting
• tickling another child to make them laugh.

It appears that these non-verbal strategies are important in friendship overtures. This supports the work of Corsaro (1985) who identifies that one of the most popular ways that children enter a group is through non-verbal strategies.

Friendship overtures in older and younger children
I also recorded which child had made each friendship overture and noted that child's age. There appeared to be a difference between the sixteen younger children (2.11 years to 3.5 years) and the ten older children (3.6 years to 4.10 years). In total, the children in the younger group made 97 overtures of friendship (an average of six each), while the children in the older group made 71 overtures of friendship (an average of 7.1 each). The older children therefore seem to have made slightly more friendship overtures than the younger children.

Next, I looked at how many overtures were received by each child. I found that the younger children had 58 friendship overtures made to them (or an average of 3.6 each). In contrast, the older children had fewer overtures made to them, just 22 (or an average of 2.2 each). In addition, there seemed to be evidence that a small group of children made a lot of the overtures, while others made fewer. When looking more closely at which children made fewer overtures, it appears that it was those who are generally considered to be less socially able.

In contrast, it was significant that although Salesi made many friendship overtures and his overall success rate was only about 25%, he didn't give up. Corsaro (1985) noted that, although the initial access attempts were frequently resisted, if a child persisted with a sequence of attempts, group entry was more likely (over 75% successful).

Gender differences in friendship overtures
I next decided to see if there were any gender differences in the children's friendship interactions. It was interesting to note that the twelve boys seemed to make and receive significantly more overtures of friendship. The boys totalled 89 friendship overtures (an average of 7.4 each), while the fourteen girls made/received 82 overtures (an average of 5.8 each). It is possible that girls had more sustained friendship interactions, therefore they did not need to interact with so many other children in order to find a friend.

Pam's concluding reflections about teaching and learning
Almost all children were involved in friendship overtures as either initiators or receivers. In contrast to the younger children (who tended to make many more fleeting friendship interactions over an afternoon), the smaller number of overtures made by the older children was somewhat unexpected. However, looking at the number of overtures made by the different age groups, as opposed to interactions, it may be that older children have more sustained friendship interactions and therefore do not need to make as many overtures. Likewise, the smaller number of overtures made by girls as opposed to boys may indicate more enduring friendships and stronger social competence in this group.

Some children lacked strategies for initiating friendships so they may need to be exposed to more modelled friendship overtures, and encouraged to build confidence in this area. Other children persisted in making friendship overtures, despite often being rebuffed. These children appear to accept that overtures often do not end in friendships. It was interesting to note the level of resilience demonstrated by children in this regard; encouraging persistence in making friendship overtures may therefore be important.

Friendships among Mixed Age Groups - By Liz Schimanski
I analysed the twenty-six Friendship Maps that were included in the data, looking at the age of the child, who they interacted with, the ages of the children where friendships were indicated, and the length of the children's friendship interactions.

While Pam identified differences between two groups: older and younger children, I noted variations in the friendship interactions of three different age groups. The largest group of children were the three-to-four-year-olds. There were fifteen children in this group. In contrast, only six children were older, aged between four and five-years-old, and just five children were aged between two and three-years-old. I examined the impact of different age groups on the children's ability to sustain friendships, the formation of vertical groupings, the use of non-verbal strategies in friendships, as well as the role of persistence in friendships.

Sustained Friendships
A total of 225 significant friendship interactions were identified in the data (indicated by starred boxes, see Appendix One). After analysing these interactions, it appears that a number of children had three or more significant friendship interactions with the same child and their play together lasted for several minutes as opposed to several seconds. For example:

• Nyra (3.3 years) had six friendship interactions with Hine (3.2 years)
• Luka (4.5 years) had twelve friendship interactions with Semo (4.3 years)
• Salesi (3.2 years) had three friendship interactions with Adam (3.4 years)

This indicates that some children are developing and maintaining friendships with others of a similar age. However, although friendships did occur across all age groups, the older age group seemed to have the majority of friendship interactions, regardless of whether they or someone younger initiated the friendship. As this group was not the largest, their greater number of friendship interactions perhaps relates to their higher level of social development. This supports Pam's findings.

Vertical grouping
It was interesting to note that twenty of the significant friendship interactions actually involved children within the two-to-three-year-old group. The younger children tended to gravitate towards and create friendships with the older children rather than become involved in friendships with their same age peers. The younger children also tended to have fewer interactions with those that are of a similar age. For example, Callum (2.11 years-old) had four friendships with Kahu and Mareta, who were both 3.3 years-old. From Callum's total of eleven interactions, only three were with children of a similar age. However, this could be influenced by the lower numbers of children in the two-to-three-year-old group.

It appears that aspects of Samoan culture may play a part in the way the older children, especially those closer to school age, feel a sense of responsibility towards the younger group. I found one example where two Samoan girls, who were almost five, took on the role as the main caregiver to a younger child, not yet three and who was just settling into kindergarten. Abby (2.11 years), was 'looked after' by Silei (4.1 years) and Isa (4.4 years). This led to other Samoan children taking on the 'caregiver' role with this child if one of the key caregivers was absent. For instance, Semo was seen looking after Abby when Silei was not at kindergarten. Nonetheless, this pattern has not meant that all other friendships have ceased to develop for the Samoan children while they looked after younger children; it merely means that the younger child becomes involved in their play; so Abby was taken along with Silei and Isa as they played with Salesi and Luka and all were involved in play throughout the afternoon.

Using non-verbal strategies to initiate friendships
Younger children might not say to each other: 'I am your friend, come and play,' but it does appear that non-verbal interactions such as smiling, laughing, having fun together are ways that children convey this same meaning. For example:

• Leighton (3.2 years) watched a lot during the first part of the session. When he was at the pole house, he smiled at Adam and they began to go down the slide together.
• Callum (2.1 years) at the play dough table, was able to initiate a friendship with Kahu (3.3 years) by using comedy.

Persistence
Rata attempted numerous times to initiate a friendship with Tama, but she was continually rebuffed. However, this did not immediately deter her. She persisted for some time although, unfortunately, a friendship did not ultimately occur. There could be many reasons why this continued interaction failed to develop into a friendship; for example, factors such as gender, proximity or physical attractiveness may be responsible. It is also possible that more salient factors, such as friendships already developed with other children, may be responsible for a child to be rejected (Rizzo, 1989).

Liz's concluding reflections on learning and teaching
The usual kindergarten practice is to have the younger age group (three-to-four-year-olds) attend the afternoon sessions, while the older ones (four-to-five-year-olds) attend the morning session, thereby separating the children based on age. The information we have collected suggests that, for enhanced social and emotional development, sessions including mixed age groups could have positive benefits. By having vertical grouping in our afternoon session we see children of all ages mixing and making friends with each other. Furthermore, it was interesting to find that older Samoan children seem to have taken the role of nurturing and settling in the younger, more unsettled children while still creating friendships with others throughout the afternoon session. This offers both a sense of security for the younger child as well as providing a model for dramatic/imaginative play.

It is evident that children use many different strategies to develop friendships. The friendships can be just for one afternoon or they can be more sustained and carry over to many sessions. However, each age group uses the same techniques when making friends, whether that is through non-verbal interactions or through means that are more direct. There may be ways that these techniques can be socially encouraged within the community of learners.

Reflections on intercultural and interlingual aspects of the study
A number of issues related to the intercultural and interlingual nature of the afternoon sessions were raised by this study. These matters were reflected on in the individual writing by members of the wider research team. In the section that follows, Eileen discusses teachers' findings on the practice of Samoan children calling out others' names, the children's code-switching practices, and their engagement in co-construction with adults. Next, Joy discusses the cultural imprints in children's friendships. Penny then reflects on some implications for the bilingual development of these children.

Samoan Children's Use of Names, Code-switching & Co-construction with Adults - By Eileen Woodhead
A number of interesting educational issues arose in the course of this study. For me the ones that stood out were Samoan children's use of names, the phenomenon of code-switching and their engagement in co-construction with adults in the kindergarten. These issues are commented on below.

Samoan children's use of names
One significant finding from the research was the discovery that Samoan children and adults' use of names differed from those with which we are familiar with in other New Zealand cultures. In many cultures, names appear to be used as a tool to enter or sustain friendships. However, during the study Samoan family members were observed to call out a valued friend's name or a family member's name for no practical reason, except apparently to signal I see you; I am still connected to you. It was also noticed that a Samoan adult or child would repeat a new name often, as if to get it clearly in their mind, and there was apparently no concern about asking several times for the other person to repeat their name. It is likely that in many English-speaking cultures this practice would be avoided as it is considered impolite not to quickly register the new name.

Morton (1996) describes similar behaviours associated with name usage in Tongan society, and has put forward some sociocultural reasons for this phenomenon. He explains that in Tongan families the practice of calling the names of absent people is used to distract and comfort very young children. He also notes that the use of parents and household members' names acts as a constant reminder to the child of their connections with others, even if the other is absent. The practice of calling out the names of others may therefore play an important part in children's development.

In this study, children from the Samoan culture seem to be using similar name-calling practices to those described in Morton. In addition, it has been noted that children from other cultures appear to have begun to imitate these name-calling practices as a method of entering and maintaining friendships with the Samoan children. The teachers in this study have reflected on what these findings have meant for their teaching. Some of their changes to practice include:

• being very conscious of pronouncing and spelling the Samoan child's name correctly; and, indeed, taking the time to do this
• using the child's name often to connect with them and encouraging other children to use names
• using affectionate terms such as sweetie, buddy, mate, honey at the expense of regularly using the child's name may undermine the identity of the child
• understanding that a child's name may be connecting and empowering
• having awareness that the name calling out is a valued practice in friendship interactions and providing a supportive emotional climate for this to happen freely.

Consideration could also be given to the value that European society places on the written name whereas it appears that Pacific Island cultures may value the spoken name more highly. Getting to know children's names may seem like one of the most basic tenets of teaching and therefore hardly the subject of in-depth research and reflection. However, in a climate of unpacking or deconstructing our beliefs about teaching and learning in early childhood, we are challenged to look at seemingly everyday tasks in a fresh light and become more aware of western cultural bias (Canella, 1997).

Possible reasons for code-switching in this setting
As noted earlier, one key area of interest in this study centred on code-switching. Baker (1996) indicates that this phenomenon occurs when a speaker changes from one language to another. Code-switching may occur for just a single word in mid-sentence, or for an extended part of a conversation. Baker suggests there are many reasons why individuals might choose to code-switch, including the need to express a concept that does not exist in the user's language of conversation or to include or exclude others from the conversation. Both Baker (1996) and Foster-Cohen (2003) however, mainly present code-switching in a positive light, suggesting that the bilingual speaker makes an active and positive choice to do this as a bilingual way of being.

In examining language samples that show code-switching, the monolingual listener/reader is often left guessing (albeit an informed guess) as to what the motivating factor was for any particular instance of code-switching. However, several incidents emerged during the study that may shed light on this phenomenon. For instance, a Samoan child would often speak fluently in Samoan and then unexpectedly interject an English phrase or expression. This sometimes appeared to be done when a child did not know the Samoan equivalent for a word or concept, or when a phrase was seen as embedded in the English speaking culture, such as 'OK Mate' (Silei to Ioane), and 'Stop it, I don't like it' although the latter is promoted within the kindergarten in both English and Samoan. It appears that when a concept has been learned in English then English becomes the preferred language to express the concept in the midst of Samoan discourse. Some examples that occurred are:

• Semo telling Peteru 'It's my turn,' a turn-taking phrase that is actively taught in the kindergarten.
• The word 'painting' occurring in the middle of a stream of Samoan language, gives rise to questions about whether the child has not yet learned this word in Samoan and whether this experience is unique to kindergarten for that child.
• In the midst of conversation in Samoan, Ioane sings Choo Choo the fast train is coming down the track - a song that is probably only sung at kindergarten in English.

It is reasonable to assume that a Samoan child may switch to English to advance a friendship (e.g. Luka, a Samoan child, uses English when interacting with Tama, an English-speaking Māori child). A similar reason for code-switching behaviour is also noted by Baker, and the data from this study indicates that the promotion of friendships may be a common motivation for code-switching by the Samoan children.

Foster-Cohen says: "Bilinguals code-switch for a number of reasons, but always because they feel comfortable doing so, and partly because it defines who they are" (p. 55). For this reason, the teacher-researchers affirm that it is not just acceptable but in fact desirable for the adults working with bilingual children to also code switch when they are motivated to do so, just letting it unfold naturally.

Co-construction and the role of adults in friendships
Adult scaffolding of friendships almost certainly happens over time and in subtle ways, but it was difficult to find many clear examples in the data of where an adult had obviously led the child into a friendship interaction with another child. However, children often called adults' names, especially the name of a Samoan teacher aide employed at the kindergarten. Another important finding is that each child received a meaningful teacher/child interaction at each session. This is likely to be a positive outcome of the favourable group size, low ratios and teacher awareness.

What did come through strongly from the data was the number of examples where an adult and child had enjoyed an interaction that could be described as co-construction. Dockett and Fleer (1999) describe this approach as one in which both two individuals collaboratively take an active part in the construction of knowledge and understanding, seeking to interpret and make sense of experiences, rather than following a model in which knowledge is simply transferred from adults to children. In the data on friendship interactions children appear to be having many co-constructive friendship interactions with adults in their learning community. These are marked by reciprocity, power sharing, turn taking, sharing an interest, being close, fun, laughter, touching, smiles.

Cultural Imprints on Friendship - By Joy Mepham
Informal observation shows that the children in Wycliffe Nga Tamariki Kindergarten are relatively unaware of ethnic differences and appear to form friendships with others that they share activities and interests with rather than those from the same cultural group. In an endeavour to authenticate these impressions, all interactions between children were coded for ethnicity and patterns of interaction between different ethnic groups were examined.

It appears that there were more interactions between children within the same cultural group, particularly for the Samoan children. Litia and Malia are the only exceptions, both interacting more with Māori children. It is interesting to note that these two girls also have the highest number of interactions with teachers. However it is important to remember that each child was observed for one session only, and the contextual information on Malia's Friendship Map explains that it was a smaller group than usual on the day she was observed and that her normal friends were not present that day.

Limited language understanding between English speaking and Samoan speaking children may contribute to fewer interactions between different cultural groups, as suggested in a conversation that took place between Adam and Malia:

Malia: "What's this, what's this?" [pointing to tee-shirt]

Adam: "A fish, dah a fish"

Malia: "Olava olava"

Adam: "A fish"

Malia: "What's this?"

Adam: "Sa dolphin"

Malia: "Agh"

Adam: "Heh, it's not funny"

Malia: "Have you seen my trolley, my trolley, it's not funny, not funny"

Adam: "What's that? .. Wow!"

Malia: "What's your name?"

Adam: "Adam, what's your name?"

Malia: "Malia"

Adam: "Name Malia, her name is Malia, her name is Malia"

After three more interchanges that were not heard clearly enough to transcribe Malia asks Adam something in Samoan to which he replies: "Na I'm garble, garble, silly billy, I'm off to work. Ow... I've got to get my one, see there's my one, thank you Malia - ow - get me that trolley," after which their interaction ends.

There is some evidence to suggest that, from about the age of three, children become aware of their own ethnicity (Haslett & Samter, 1997) and are able to identify children from other ethnicities. They also suggest that young children become friends with those who are the same age, race and gender whereas older children form friendships with others who share similar interests and attitudes (p. 195). In an early childhood environment such as Wycliffe Nga Tamariki, the sharing of similar interests and activities appears to also indicate the development of friendship in this group of young children. For instance, Tama is working in the sandpit with a wheelbarrow. Adam approaches him and offers to help. This play continues for 10 minutes and contains many indicators of friendship.

In another example, Callum and Kahu are sitting together and begin to bump their heads together. They laugh and continue to do it. They then get a poi each and begin to twirl them around. The poi gets tangled and the teacher helps to untangle them. There is lots of laughter. The activity with the poi widens to include others. These are both examples of intercultural friendships and the Friendship Maps also contain examples of boys and girls engaged in activities indicating friendship. Nevertheless the absence of data containing both intercultural and cross-gender friendships suggests that these may not be happening. More teacher involvement in children's friendships - as a friend, may stimulate such interactions.

An example of an intercultural friendship interaction between Abbey and Silei appears to indicate that cultural influences contribute to the friendship. The teacher-researchers wrote:

"When Abbey first started kindergarten she was very unsure about staying by herself and being independent. Some older (Samoan) girls realised Abbey's unease and started to help her. For example, Silei will hold Abbey's hand when she first arrives and then help her to settle into an activity. She will play near Abbey and help her to become part of the group. Silei also takes the role of 'mother', helping Abbey with such tasks as putting on her shoes. This has been a nurturing relationship that has developed. Abbey has found a true friend who has helped her to grow and mature. Silei has been happy to take on the role of helping another."

Such examples suggest that there are cultural imprints on children's friendships. However the current research suggests that there are very few important differences between cultures in children's basic understandings of friendship. Indicators of the 'friendship dance' are similar across most cultures, but it is possible that differences between cultures may influence the kind of behaviours expected of friends.

The Implications of Friendship for Children's Bilingual Development - By Penny Haworth
It seems likely that the 'dance of friendship' provides a way for children to hone their social, communicative and cognitive skills. This may be particularly important in the multicultural setting of Wycliffe Nga Tamariki Kindergarten where intercultural and interlingual competence develops alongside the development of cognitive concepts and social skills.

While it appears that both non-verbal and receptive communicative strategies were often used between children from different language backgrounds, more cognitively and linguistically sophisticated interactions tended to take place between children who shared the same language. In contrast, friendship interactions between English and Samoan speakers often appeared to be more fleeting in nature, have less communicative meaning, and focused on less complex topics. For bilingualism to be promoted within this setting, the involvement of adults, and the encouragement of proactive collaboration may therefore be critical. Not only may adults enhance children's sense of familial belonging, but the degree to which proactive collaboration between children is supported by adults, in assisting with both bilingual and cognitive development, may also be relevant to the children's developing competence as bilingual learners in this setting.

Cummins (1984, 1986) and Baker (1996) both affirm the need to provide opportunities for interaction on more complex topics. To do so within a bilingual context is likely to enhance the parallel development of bilingualism alongside the development of complex cognitive skills and concepts. Bilingualism that promotes more complex thinking and language is most likely to occur in scaffolded interaction around complex learning themes that arise in the kindergarten. Many teachers in such settings do not have sufficient bilingual proficiency to be able to capitalise on more complex learning events through the medium of the Samoan language. Therefore, increasing the opportunities for the children to engage in scaffolded discussion with Samoan adults within an extended community of learners will undoubtedly be useful. The current inclusion of Samoan early childhood student teachers within the kindergarten will probably have positive spin-offs in this direction. However, in the long term, Samoan parents may have a similar effect, particularly if teachers continue to initiate effective ways to inform them about topics of interest within the setting, share their praxis, and guide them in how to support their children's learning in these areas.

Changes to teaching and learning resulting from the study
Teachers' reflections on this first phase of the action research served as an impetus for some changes to be made in current teaching and learning practices, as well as confirming other present practices at WNT Kindergarten. These key outcomes are explained below, in association with each of the four key research questions.

Question 1: How are the children moving in and out of groups?
Frequently using others' names appears to be very important to children in this setting. The calling out of names, to both adults and children, appears to be a strategy to affirm friendship rather than to achieve a purpose (i.e. gain someone's attention). It is evident that some non-Samoan children appear to be also adopting this practice as a means to affirm friendship.

As a result of this finding, teachers are now placing further emphasis on knowing children's names, especially the Samoan children's names, which can at times present pronunciation challenges for non-Samoan people. Further related teaching activities have also been added to the kindergarten programme, such as:

• teachers intentionally using names in their incidental conversations
• singing that includes naming
• using photos and power point stories to support children's ability to name others.

Furthermore, the early data indicated that children's friendship interactions were more prevalent in the outdoor environment, with trolleys and the water trough being the most common physical sites. These findings have given teachers more confidence in the view that unstructured access to play experiences is highly desirable. When using large equipment, such as trolleys, it is essential for children to communicate, negotiate and take turns in order to participate successfully. The accessibility of these activities and resources may facilitate these interactions. In addition, teachers feel it is important to support intercultural friendships and to encourage more complex language interactions between peers during such activities.

Question 2: Do non-verbal communications initiate and sustain friendships and if so how?
As well as verbal communications, holding hands and touching were commonly identified as forms of connection during friendship interactions. Laughing - in the context of having fun together - also seems to emerge as common and important in terms of friendship interactions. In addition, smiles are used frequently for various purposes such as greeting, affirming and sharing pleasure. Teachers may now more consciously provide models of positive interactions for the children and thus help to maintain a happy working atmosphere in the kindergarten.

Question 3: Does code-switching occur in the forming of friendships?
There is also evidence in the early data to indicate that code-switching is prevalent during friendship interactions involving the Samoan children. A particular code is chosen according to the ethnicity of the friendship dyad of the moment. Code-switching appears to assist effective expression for Samoan children with limited spoken English. Furthermore, the initial data commonly shows two word English utterances occurring amidst fluent streams of Samoan language. These findings support teachers' belief that code-switching is both desirable and important for children's communication. This is in keeping with views expressed in current literature (Baker 1996; Foster-Cohen, Meade and Puhipuhi, 2002). Teachers will continue to respect and support this practice through frequent use of Samoan as well as English language in the kindergarten.

Question 4: What other factors contribute to making friends?
It appears that children have an ability to persevere in the face of many rejections. Vertical grouping also seems to influence the nature of friendship interactions. In addition, the sheer volume of interactions initiated by the very young children brings to question the early childhood paradigm of solitary, parallel and associative play as a predictable linear progression in social development. This theory, if deeply embedded, may influence teachers' observations of children. However, Canella (1997), in 'Deconstructing Early Childhood Education,' challenges teachers to identify firmly held beliefs and re-evaluate them with an open mind.

The volume and diversity of the very young children's interactions may suggest that successful social development is non-linear and variable according to factors such as the sort of weather on the day, the setting and the dispositions of the child. Children seem to utilise a `trial and error' strategy in this process, but dispositions for friendship, such as willingness to take an interest or persevere in the face of rejection, may also play an important part in children's ultimate success in establishing friendships with their peers.

Where to from here?
Through their initial engagement in this process of participatory action research, teachers have developed an increased awareness of, and greater empathy with, all of the children in the kindergarten. The in-depth observations and rich data have deepened each teacher-researcher's knowledge of individual children, resulting in greater advocacy for each child based on fuller knowledge and understanding. This has clearly contributed to more effective assessment of and planning for children.

The research team is still honing their skills in the collaborative, participatory action research process. However, their partnership has been strengthened by the shared excitement and enthusiasm for the rich data collected in this phase of the study, and reflections on what this means for children's learning and their teaching. The university researchers' vision has also been enhanced by the rich contextualisation of the data, to a degree not possible in more traditional researcher-driven investigations which are of necessity limited by time and resources. It is therefore good to know that the journey does not stop at this point. As Wadsworth (1998) points out: "Participatory action research is not just research which we hope will be followed by action! It is action which is researched, changed and re-researched, within the research process by participants" (p. 14).

After spending long hours talking, thinking and writing about the data from the first phase of the study the research team were especially intrigued with the discovery that those children who were involved in more sustained friendships also seemed to be the ones who were engaged in more complex learning. Furthermore, the role of adults emerges as a significant factor in facilitating children's friendship interactions. Interest has therefore grown around the question of how adults facilitate children to engage in more cognitively complex interaction and how they enhance children's language and thinking through the co-construction of ideas. The data so far has certainly confirmed Massey's (2004) notion that "teachers tend to act as magnets for pre-school children" and that the presence of adults appears to help to sustain interactions over a longer period of time. Enhancing children's working theories of their world in this setting, with its rich diversity of culture and languages, has therefore become the research team's new challenge; but that story will have to be told in a later chapter of this ongoing voyage of discovery!

References
Alton-Lee, A., & Nuthall, G. (1992). A generative methodology for classroom research. Educational Philosophy and Theory, 24(2), 29-55.

Baker, C. (1995). A parents' and teachers' guide to bilingualism. Clevedon, Avon: Multilingual Matters Ltd.

Baker, C. (1996). Foundations of bilingual education and bilingualism. Clevedon, Avon: Multilingual Matters Ltd.

Bell, J. (1997). Doing your research project. Buckingham: Open University Press.

Bodrova, E., & Leong, D. J. (2003). Chopsticks and counting sticks: Do play and foundational skills need to compete for the teachers' attention in an early childhood classroom? Young Children, 59(3), 10-17.

Burgess, F., & Mara, D. (2000). The development of Pacific Islands Early Childhood Centres: A suitable case for networking and mutual support across the sector. Early Education, 23, 19-23.

Burns, A. (1999). Collaborative action research for English language teachers. Cambridge: Cambridge University Press.

Canella, G. S. (1997). Deconstructing early childhood education. Social justice and revolution. New York: Peter Lang.

Cardno, C. (2003). Action research a developmental approach. Wellington: New Zealand Council for Educational Research.

Carr, M. (2001). Assessment in Early Childhood Settings: Learning stories. London: Paul Chapman Publishers.

Corsaro, W. A. (1985). Friendship and peer culture in the early years. Norwood, NJ: Ablex.

Cullen, J., & Allsop, G. (1999). Enriching the knowledge base of children's play. A paper presented at the Seventh Early Childhood Convention, Nelson, New Zealand.

Dockett, S., & Fleer, M. (1999). Play and pedagogy in early childhood: Bending the rules. Melbourne: Thomson.

Fleer, M. (2002). Socio-cultural assessment in early years' education: myth or reality? International Journal of Early Years Education, 10(2), 105-120.

Foster-Cohen, S. H. (2003). A review of bilingualism and second language acquisition in early childhood. Part C, Pasifika ECE Final Report. Wellington: Anne Meade Associates (Meade, PuhiPuhi, Foster-Cohen).

Hargreaves, A. (1994). Changing teachers, changing times. Teachers' work and culture in a postmodern age. London: Cassell.

Harley, E. (1999). Therapeutic model of play. A paper presented at the Seventh Early Childhood Convention, Nelson, New Zealand.

Haslett, B. B., & Samter, W. (1997). Conceptions of friendship: A developmental perspective. Children communicating: The first 5 years, pp.192-235. London: Lawrence Erlbaum Associates.

Kemmis, S., & McTaggart, R. (Eds.) (1988). The action research planner (3rd edition). Geelong, Victoria: Deakin University Press.

Krashen, S. (1982). Principles and processes of second language acquisition. Oxford: Pergamon Press.

Lindquist, G. (2003). The dramatic and narrative patterns of play. European Early Childhood Education Research Journal, 11(1), 69-78.

Massey, S. L. (2004). Teacher-child conversation in the preschool classroom. Early Childhood Education Journal, 31(4), 227-231.

Meade, A., Puhipuhi, H., & Foster-Cohen, S. (2002). Pasifika Early Childhood Education, Final Report, Anne Meade Associates (unpublished).

Ministry of Education (1996). Te Whāriki: Early childhood curriculum. Wellington: Learning Media.

Ministry of Education (2002). Pathways to the future: Ngā Huarahi Arataki. A ten year strategic plan for early childhood education. Wellington: Learning Media.

Morton, H. (1996). Becoming Tongan: An ethnography of childhood. Honolulu: University of Hawaii Press.

Richards, J. C., & Lockhart, C. (1994). Reflective teaching in second language classrooms. Cambridge: Cambridge University Press.

Rizzo, T. A. (1989). Friendship development among children in school. Norwood, New Jersey: Ablex Publishing Corporation.

Rogoff, B. (2003). The cultural nature of human development. New York: Oxford University Press.

Saravan, V. (2004). Bilingual children's language proficiency and language choice patterns. Journal of Australian Research in Early Childhood Education, 11(1), 13-25.

Sharman, C., Cross, W., & Vennis, D. (2000) Observing children: A practical guide (2nd edition). London: Cassell.

Vygotsky, L. (1978). Mind in society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.

Wadsworth, Y. (1998). What is participatory action research? Action Research International. Paper 2, retrieved from http://www.scu.edu.au/schools/gcm/ar/ari/p-ywadsworth98.html on 31 October, 2006.

Wragg, E. C. (1994). An introduction to classroom observation. London: Routledge.

Yong, H., & McMullen, M. B. (2003). Understanding the relationship among American primary-grade teachers and Korean mothers: The role of communication and cultural sensitivity in the linguistically diverse classroom. Early Childhood Research and Practice, 5(1), 1-17.

Footnotes
1 The procedures for this study were scrutinised and approved by the Massey University Human Ethics Committee (PN Protocol 03/84) on 11 August 2003.