Learning, Culture and Social Interaction

Cultural patterns of collaboration and communication while working together among U.S. Mexican heritage children

Maricela Correa-Chávez ⁎ California State University, Long Beach, United States

a r t i c l e i n f o a b s t r a c t

Article history: Received 13 November 2015 Received in revised form 9 August 2016 Accepted 15 August 2016 Available online 9 September 2016

To examine the cultural organization of collaboration, 50 U.S. Mexican-heritage sibling pairs (ages 6–11) were videotaped as they participated in a puzzle construction activity. Half were from families with more recent connection with rural practices, and limited schooling (“pueblo families”) and half from “high schooling families” (more connection with middle-class prac- tices, higher schooling). Children were given a previously constructed model, parts to construct another, and left alone. Every 10 s coders noted how the siblings coordinated either: jointly en- gaged, checking-in, solo, or off-task and if collaboration was organized either verbally, nonver- bally, or with multiple means. Children from “Pueblo” families engaged jointly and used nonverbal and multiple means of communication more than children from “high schooling” families who more often worked solo, were off-task, and used talk to communicate. Results are linked to practices in Indigenous American communities where children's Learning by Ob- serving and Pitching In is common.

© 2016 Elsevier Ltd. All rights reserved.

Keywords: Collaboration Nonverbal communication Learning by Observing and Pitching In Children from immigrant families Talk

1. Introduction

This study examined the cultural organization of collaboration among sibling pairs of Mexican heritage whose families had varying experience with Indigenous and Western cultural practices. Of particular interest was whether children from families that presumably have more experience with Indigenous forms of organizing teaching and learning would coordinate more collab- oratively compared to children whose families have more experience with middle class European American cultural practices (as indexed by extensive maternal schooling). A secondary question examined cultural patterns in how children communicated while they were collaborating. The need to understand variability in forms of collaboration and communication is especially important in light of the fact that among Mexican immigrants to the United States there is large variability with school experience (National Task Force on Early Childhood Education for Hispanics, 2007). An emerging body of literature suggests Mexican mothers with fewer grades of school may have more experience with a form of organizing learning common in Indigenous communities of Me- soamerica called Learning by Observing and Pitching In (LOPI) (Correa-Chávez, Mejía-Arauz, & Rogoff, 2015; Rogoff, Paradise, Mejía- Arauz, Correa-Chávez, & Angelillo, 2003).

1.1. Learning by Observing and Pitching In

A long line of research has shown that in communities with Indigenous North and Central American histories children have wide access to family and community activities and are treated as “legitimate peripheral participants” (Lave & Wenger, 1991).

Learning, Culture and Social Interaction 11 (2016) 130–141

⁎ Corresponding author at: CSULB, Psychology Department, 1250 Bellflower Blvd., Long Beach, CA 90840-0901, United States. E-mail address: [email protected].

http://dx.doi.org/10.1016/j.lcsi.2016.08.001 2210-6561/© 2016 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Learning, Culture and Social Interaction

journal homepage: www.elsevier.com/locate/lcsi

LOPI details the cultural organization of this form of teaching and learning where children are included in family and community activities and frequently pitch in, or help out ongoing activity out of their own initiative (Alcalá, Mejía-Arauz, Rogoff, Coppens, & Dexter, 2014; Rogoff, 2014; López, Najafi, Rogoff, & Mejía-Arauz, 2012; Mejía-Arauz, Correa-Chávez, Keyser Ohrt, & Aceves-Azuara, 2015).

This collaboration and initiative in activity is often evident in the ethnographic reports of children working with adults in com- plex tasks from helping to run small business, to running errands, to translating, and caring for sibling (Lancy, 2008; Ochs & Izquierdo, 2009; Mejía-Arauz et al., 2015). Additionally, the helpfulness reported is not only limited to the home environment, Mexican and Central American descent students in one Los Angeles area school volunteered to help out in school even when they were on vacation in addition to helping out at home (Orellana, Dorner, & Pulido, 2003). Interviews with mothers provide more support for these observations and also indicate that often children participate in these activities from their own initiative, without being forced or cajoled into helping (Alcalá et al., 2014; Coppens, Alcalá, Mejía-Arauz, & Rogoff, 2014; Mejía-Arauz et al., 2015). This study builds on the ethnographic work showing cultural differences in children's helpfulness and on the work of Rogoff and her colleagues who have suggested that a key part of childhood in communities with Indigenous histories involves collaboration and initiative in participating in family and community activities (Correa-Chávez et al., 2015; Rogoff et al., 2003). These cultural patterns of encouraging helpful collaboration at home and in the community are often different than the patterns of work encouraged by school.

1.2. Cultural patterns of collaboration

Traditionally schools have been places where individual work is prioritized and collaboration is either actively discouraged (called cheating) or simply not encouraged in day to day interactions where the teacher/whole class or teacher/individual student format predominates (Mehan, 1979). When children from a traditionally organized classroom were asked to collaborate, they often used test and quiz formats rather than interactions where they built off other one another's efforts (Matusov, Bell, & Rogoff, 2002). Of particular importance to this study is the fact that children must learn the cognitive skills of effectively collab- orating with one other people, and just as modes of interaction vary across cultural communities, forms of organizing collabora- tion should also vary across cultural communities (Correa-Chávez & Rogoff, 2005).

Research suggest that working together in a way that skillfully blends with ongoing interaction appears to be a skill learned early on in interaction with others in communities with Indigenous history. Mayan toddlers were more likely to request help from their already occupied mothers using gaze, touch, and body posture compared to European American toddlers who rarely used those methods in asking for help (Rogoff, Mistry, Göncü, & Mosier, 1993). Among school aged children, U.S. Mexican heritage children were ten times more likely to wait patiently and check that an adult was not busy before asking her for help compared to European American children who loudly and frequently interrupted ongoing activity (Ruvalcaba, López, Rogoff, Correa-Chávez, & Gutierrez, 2015).

Skills at observation and integration in ongoing activity may be related to the different forms of collaboration observed in com- munities that have Indigenous history. When working on a collaborative project with multiple participants, Mayan mothers with 2 or fewer years of schooling were more likely to build off the work of participants. Mayan mothers with 12 or more years of schooling were more likely to subdivide the project into smaller tasks (Chavajay & Rogoff, 2002). Similarly when 3 siblings were engaged with a novel science exhibit U.S. Mexican heritage siblings coordinated in a way that rarely interrupted ongoing activity, but rather blended agendas between participants. European American siblings were more likely to interrupt ongoing ac- tivity and organize the activity through turn taking rather than working together (Angelillo & Rogoff, 2005). Even when working on individual projects, U.S. Mexican heritage children whose families were more familiar with the ways of rural Mexico and had fewer years of schooling were more likely to engage as a group compared to U.S. European heritage children and Mexican heri- tage children whose mothers had 12 or more years of schooling (Mejía-Arauz, Rogoff, Dexter, & Najafi, 2007).

The cultural patterns in collaboration found in the research may be learned in interaction as children work alongside adults in meaningful family and community activity. However this pattern of involvement tends to change as schooling becomes more prominent in family life. Studies examining the helping behaviors of Mexican children at home have consistently found that as parents have more school experience children contribute less often and less meaningfully to the family, and childhood is reframed as a time where children are supposed to be dedicated primarily (if not exclusively) to school (Alcalá et al., 2014; Coppens et al., 2014; Mejía-Arauz, Keyser-Ohrt, & Correa-Chávez, 2013). However as previously stated, traditional schools do not tend to pro- mote collaboration (Matusov et al., 2002), focusing instead on individual accomplishments and verbal competency (Hart & Risley, 1995).

1.3. Talk and joint activity in collaboration

Although some scholars have focused on the role of talk in collaboration, if children are accustomed to integrating themselves into community and family work without interrupting others, collaborative interactions may rely heavily on the act of working together (joint activity) as an organizer in addition to talk. In some circumstances engaging in large amounts of talk may not nec- essarily lead to the most fruitful collaboration. For example nonverbal interaction and modeling were common and effective teaching strategies for young siblings working together (Azmitia & Hesser, 1993) as well as for college students working together (Azmitia & Crowley, 2001). In some cases, such as in designing structures “rapid prototyping,” (building and testing ideas with

131M. Correa-Chávez / Learning, Culture and Social Interaction 11 (2016) 130–141

minimal discussion) led to designing better structures than collaborations involving intensive discussion (Apedoe, Mattis, Rowden-Quince & Schunn, 2010).

However in communities where people have had many years of schooling, as well as in schools themselves, talk is often pri- oritized as a marker of engagement and learning (Dixon, Levine, Richman, & Brazelton, 1984; Kim, 2002). In teaching interactions between Zinacantec Mayan siblings, children with only a few years of school used more talk when teaching a younger sibling compared to children who had not been to school. The children who had not been to school tended to use more bodily closeness and bodily guidance as teaching tools (Maynard, 2004). When children were placed in groups of three with an adult who was teaching them a new task, European American children were more likely to communicate with one another using talk, whereas U.S. Mexican heritage children whose families had few years of schooling (and whose families had immigrated from rural Mexico) were more likely to engage in multiple nonverbal turns at communication called nonverbal conversation (Mejía-Arauz et al., 2007). The results from these studies as well as previous cultural research indicate that the patterns of verbal communication many believe are necessary for collaboration are cultural in nature and may be related to familial participation in the cultural in- stitution of school.

1.4. Schooling as a cultural practice and familiarity with indigenous ways

The institution of school has organized child life and learning in many middle class communities for generations, although its role is often overlooked (Hernandez, 1997; Rogoff, Correa-Chávez, & Navichoc-Cotuc, 2005). Many practices common in schools such as engaging in child focused activities, and mini language lessons, are common in communities that have an extensive his- tory with school, but uncommon in other communities (Gaskins, 1999; Lancy, 2008; Morelli, Rogoff, & Angelillo, 2003; Scribner & Cole, 1973).

In Mexico, mass schooling was also important as a way of forging a national “Mexican Identity” out of the many Indigenous groups of the country following the 1910 revolution. Part of the goal of schooling was to “modernize” the countryside by replacing Indigenous languages and traditions. As a result of these efforts many rural communities of Mexico no longer consider themselves Indigenous even though many still engage in some traditional practices and ways of life (Flores, Urrieta, Chamoux, Lorente Fernandez, & López, 2015). Migration to the United States has been most common from these rural areas (Consejo Nacional de Población, 2001). Silva et al. (2010) and López et al. (2010) argue that in many immigrant Mexican communities in the United States with limited schooling people may be familiar with the cultural pattern of LOPI. Therefore in this study, children's patterns of behavior in interaction are not seen as arising from a deficit, or “lack” of knowledge of school ways, rather from engagement with another cultural form of supporting learning.

According to the U.S. National Task Force on Early Childhood Education for Hispanics (2007), among Mexican immigrant mothers of 8 year old children, 48.6% completed high school, and 51.4% had fewer than 11 years of school. Maternal participation in the cultural practice of school has far reaching implications for children of the next generation, as mothers may organize inter- actions with children in ways that reflect the organization of school (Richman, Miller, & LeVine, 1992; Rogoff et al., 1993).

The different forms of interaction are also seen over generations where children whose mothers have many years of schooling tend to exhibit forms of interaction and group organization common to school. Mexican heritage children in the U.S. whose mothers averaged 7 grades of school relied more on observation as a source of information compared to Mexican heritage chil- dren in the U.S. whose mothers had 12 or more years of schooling (Mejia-Arauz, Rogoff & Paradise, 2005). The U.S. Mexican her- itage children whose moms had fewer years of schooling also attended simultaneously to multiple ongoing events more often than U.S. Mexican heritage children shoes moms had 12 or more years (Correa-Chávez, Rogoff, & Mejía-Arauz, 2005). Mayan chil- dren whose mothers averaged 3 grades of schooling attended more to information directed to others compared to Mayan children whose mothers averaged 12 or more years (Correa-Chávez & Rogoff, 2009). Additionally U.S. Mexican heritage children whose mothers averaged 7.5 years of schooling attended more to other's activity compared to U.S. Mexican heritage children whose mothers had 12 or more years of schooling (Silva et al., 2010; López et al., 2010).

Although many patterns of teaching and learning seem to change with increased parental participation in school, it is impor- tant to point out that increased schooling in Indigenous communities is associated with many other demographic changes. Fam- ilies with more schooling tend to have fewer children in the family, more limited involvement in the extended family, more urban experience, different migration patterns, and occupations requiring credentials — all of which might influence child life (LeVine, LeVine, & Schnell, 2001; Richman et al., 1992). Therefore in this study, school experience is used a proxy for increased familiarity with European American middle class cultural patterns of child rearing and child learning and not seen as the sole or “active in- gredient” in changing community patterns.

1.5. Present study

This study examines how U.S. Mexican heritage siblings whose families have differing cultural backgrounds with regard to fa- miliarity with school ways of organizing learning, or familiarity with LOPI engage in joint activity when they are working together on a difficult task. Few studies have focused on the cultural aspects of nonverbal communication in collaboration among children. In previous work some studies have focused on cultural aspects of collaboration in the presence of adults (Chavajay & Rogoff, 2002), and others have focused on nonverbal conversation in situations where children were working individually (Mejia- Arauz et al., 2007). However in this study the sibling pairs were working independent of any adult supervision — free to organize their interaction anyway they saw fit while working on something together. Additionally as siblings, the children likely already

132 M. Correa-Chávez / Learning, Culture and Social Interaction 11 (2016) 130–141

had a repertoire of communicative tools at the ready thereby making it likely they would use nonverbal aspects of communication if they so desired.

Based on previous comparative and ethnographic data it was expected that there would be differences in the patterns of co- ordination, with the sibling pairs whose families have more recent immigration from rural areas of Mexico and less extensive ex- perience with school (pueblo group) engaging collaboratively more and the sibling pairs from families more familiar with middle class ways through extensive schooling and related cultural practices (high schooling group) engaging more in solo work. It was also expected that children from the pueblo families would organize their collaborative interactions relying more on nonverbal and multiple means of communication compared to children from the high schooling families.

2. Method

2.1. Participants and their communities

All of the children were in school in the United States and were recruited through afterschool clubs at three public elementary schools in the Los Angeles area. All three schools were in areas that have large Latino populations and where the schools serve almost exclusively Mexican heritage children. Almost all of the schools' students (93%) participated in a free or reduced-price lunch program. The majority of the children's families (46 of the 50 sibling pairs) had historical roots in Mexico, 2 of the sibling pairs' families were from Guatemala and 2 from El Salvador (which were included because of a similar history of Indigenous prac- tices and access to schooling). Parents provided family demographic information in responding to the permission slip sent home from school.

All 24 mothers in the “pueblo group” were born outside the United States: 21 in Mexico, 1 in El Salvador, and 2 in Guatemala. Among the fathers in this group 17 were born in Mexico, and 3 in Guatemala (4 declined to provide information). All of the mothers in this group completed their education outside of the United States, and typical occupations for the mothers included cook, garment worker, and housekeeper. Of the 16 fathers who provided schooling information, 14 had 9 or fewer years of schooling and 2 had more than 9 grades. All of the fathers that reported schooling information completed school in Mexico. Typ- ical occupations for the fathers included gardener, busboy, mechanic, and cook.

At the time of the study 8 families in the pueblo group had only the 2 children that participated in the study, 10 families had 3 children and 6 families had 4 children. Seventy percent of the children in the pueblo group reported speaking mostly Spanish at home and 46% reported having visited Mexico. Twelve of the sibling pairs were of the same gender (5 pairs of sisters, and 7 pairs of brothers). In the 12 mixed gender sibling pairs, 7 had a sister as the older sibling. Average ages of the siblings were 9 and 7 ½ years (see Table 1). All of the children were in the prescribed grade for their age (for example 7 year olds in second grade). While participating in the activity, 16 of the 24 pairs used mostly English during the puzzle activity, 7 pairs used mostly Spanish and one pair used both English and Spanish.

In the “high schooling” group, 16 of the 26 mothers were U.S. born, 9 were born in Mexico, and 1 in El Salvador. Eighteen of the mothers in this group completed their schooling in California and typical occupations included teacher, home maker, nurse, and office manager. Nine of the fathers were born in the U.S., 11 in Mexico, 1 in Guatemala, and 1 in El Salvador (4 declined to provide information). Of the 19 fathers who provided schooling information, 3 had 9 or fewer years of schooling, 7 completed high school, and 9 attended school beyond high school (14 the fathers completed their schooling in California). Typical occupa- tions for the fathers included customer service representative, graphic designer, salesman, and teacher.

Table 1 Maternal schooling paternal schooling, gender, and age of siblings.

Mexican-heritage Pueblo 24 sibling pairs (48 children)

Mexican-heritage High schooling 26 sibling pairs (52 children)

Average maternal schooling (Information available on all 50 mothers)

7.3 grades (Range: 4–11 grades) SD = 2.1

14.3 grades (Range: 12–16 grades) SD = 1.8

Average paternal schooling (Information available on 35 fathers)

8.3 grades (Range 6–12 grades) SD = 1.9

13.0 grades (Range 8–16 grades) SD = 2.6

Older sibling Gender 12 girls, 12 boys 15 girls, 11 boys Average age 9.3 years

(Range 8–10 years) 9.1 years (Range: 7–11 years)

Younger sibling Gender 10 girls, 14 boys 14 girls, 12 boys Average age of focal child 7.4 years

(Range: 6–10 years) 7.0 years (Range: 6–10 years)

133M. Correa-Chávez / Learning, Culture and Social Interaction 11 (2016) 130–141

At the time of the study 8 families in the high schooling group had only the 2 children that participated in the study, 11 fam- ilies had 3 children, 3 families had 4 children, and 2 families had 5 children. Forty percent of the children in the high schooling group reported speaking Spanish at home and 23% reported having visited Mexico. Seventeen of the siblings in this group were same gender (10 pairs of sisters, and 7 pairs of brothers). In the 9 mixed gender sibling pairs, 5 had a girl as the older sibling. Average ages of the siblings were 9 and 7 years, and all of the children were in the prescribed grade for their age. During the puz- zle activity, 25 of the 26 pairs whose mothers had high schooling used mostly English during the puzzle activity, and one pair used both English and Spanish.

Chi-square analyses showed that there were differences between the groups in how likely they were to speak Spanish at home (X2 (1) = 5.27, p b 0.05), how likely the children were to have gone to Mexico to visit (X2 (1) = 3.93, p b 0.05), and language used during the activity (X2 (2) = 8.91, p b 0.05). Among the 35 sibling pairs who provided both maternal and paternal schooling information, there was a 0.82 correlation between maternal and paternal schooling (p b 0.01) such that the more schooling the mother had, the more school the father tended to have.

2.2. Procedure

As part of a larger study the sibling pairs were invited to be videotaped as they worked together to construct a three- dimensional honey bee puzzle. The puzzle was chosen because it was very difficult to put together with many pieces similar looking pieces, therefore the children would benefit from working together as they constructed it. It was also a puzzle that had been discontinued by its manufacturer therefore it was unlikely the children had experience with that particular puzzle. This was confirmed when we asked the children if they had ever constructed a puzzle like it before. All of the children reported having had some experience with 3D puzzles either at school or at home (usually involving dinosaurs or buildings), but no child had ex- perience with the honey bee puzzle presented.

The children were seated together at a long rectangular table opposite the video camera. The female Research Assistant (RA) then brought an already completed honey bee model that had been glued together and placed it in front of the chil- dren on the table. “Now you are going to work together to make a puzzle that looks like this (hands them the model). You can look at this one and touch it and move it around as much as you want. Here are all of the pieces you need to make an- other one (lays out all of the parts on the table in front of the children). I have to go outside for a little while, but if a piece breaks you can go outside and let me know and I'll get you another piece. You can be as loud as you want while you are here.”

The RA left the room for 10 min to ensure the children would feel comfortable talking, gesturing, and being loud with one another as they worked on the puzzle. This was also done to emphasize to the children that they would be figuring out how to construct the puzzle on their own without the help of the RA. The instruction “you can be as loud as you want while you are here” was also meant to emphasize to the children that this space was different from school and they did not need to behave or use “inside voices” as they worked. After 10 min the RA walked back into the room and sat in a corner doing some “work” in her notebook until the children finished. If the siblings had not finished by 15 min, the RA checked on them asking them if ev- erything was OK and if they wanted to continue. If any pair wished to stop at this point they were allowed to do so, otherwise the RA went back to “work” and waited for the siblings to finish. Once they were done, they were taken back to their after school club by the RA.

2.3. Coding

The videotape record was divided into 10 s segments during which a bilingual coder unaware of the hypothesis of the study identified how the pair was organized as they completed the puzzle: either jointly engaged, checking in, solo, or off task. Ten sec- onds was chosen because this amount of time was long enough that the children could build off of one another's ideas and ac- tions, but not so long that other forms of interaction might be missed or lost. Only one form of interaction was coded per segment. In order to not miss any segments involving collaboration, any segment that contained joint engagement and any other form of social organization, for example “checking in,” was coded as joint engagement. Although this meant that collabora- tion could take place for 4 s only, and still be coded collaboratively, this approach was actually conservative in that it was possible that it slightly reduced the possibility of finding differences between the groups. If any segment contained checking in and either solo or off task, it was coded as checking in, and if any segment contained solo and off task it was coded as solo. Again, this may have led to fewer segments coded as off task.

Although each 10 second segment was coded independently, the information from neighboring segments was used to make sense of what was occurring in any one segment. So for example if both children were looking for a particular piece because they had both agreed to try to find it in the previous segment, and they both spent the 10 s looking for it they would have been coded as collaborative rather than solo since they were both working on something together that they had previously agreed upon. The four main categories: Jointly engaged, checking in, solo, and off task are described in more detail below. Additionally if a segment was coded as jointly engaged, the coder also identified if the children were coordinating with one another primarily through nonverbal joint activity, primarily through talk, or through multiple means of communication (these are also described more fully below at the end of this section).

134 M. Correa-Chávez / Learning, Culture and Social Interaction 11 (2016) 130–141