Reactions to Corsaro

Corsaro, W. (1985). Friendship and Peer Culture in the Early Years.

Norwood, NJ: Ablex.

Lorraine Sherry

June 23, 1997



Chapter 1.
Entering the Child's World: Research Strategies for Studying Peer Culture

Corsaro's overall structure is surprisingly coherent, despite the narrative form of his report. In the first chapter he explains the purpose and context of his research study, ties it to relevant, related literature, describes the setting and population, defines the unit of analysis, describes the data collection process, and explains how he analyzed the data. It is very different from Heath's ethnography in which her methodology is quite hidden. Corsaro starts out by alluding to an initial statement of the purpose of his research and then clarifies it in the cover letter to parents:

Thus, from the very beginning, his stated purpose is very different from mine. Whereas he wants to explore how children begin to construct their concepts of roles, status, and their place in society, I am interested in solving practical problems such as overcoming barriers to the use of e-mail and Internet use by new users, developing the most appropriate staff development program for an educational organization, or helping educational institutions adopt, use, and integrate technology into their curriculum.

His rationale for doing an ethnographic study is that they are sparsely represented in the literature for theoretical reasons (a bias toward developmentally-centered research) and methodological reasons (researchers interpret the data from an adult perspective rather than interacting with children in a natural setting). The dozen or more studies that he cites do not capture background on the children's perceptions of their activities and social-ecological environments. Thus, he orients the reader and makes clear the relationship of the purpose of his exploratory study to previous research in the first few paragraphs.

Was the setting purposefully selected? This is not clear. If it is was, then he should have stated so and given a rationale for its selection; if not, then he should have stated that it was a convenience setting. All we know is that he says, "I was a stranger to this part of the country" who had only corresponded with the director by letter prior to the beginning of his study.

The school contained two representative samples (one boy, one girl, with a birthdate in each month of the year) of children in the 2.1-3.1 and the 3.9-4.10 year age bracket. It also had an excellent area for unobtrusive observations that had been used by previous researchers. This was his way of keeping initial observer interference to a minimum. His initial stay of one month prior to beginning his investigations is carefully documented, especially his "negotiations with the gatekeepers", in which he carefully avoided the mistakes of prior researchers at that school (and at other schools such as the Window Rock, Arizona fiasco!)

Entry into a school site is not an easy process. More than just phone calls to the principal, a cover letter, consent forms, and a letter of permission from a Human Subjects committee (documents I'd usually include in an appendix), it involves a very careful entry into another culture with norms of behavior which may be very foreign to the researcher. For example, when I had to conduct observations in an elementary school in Boulder, I was only able to do so in a school that had been previously observed by other researchers from RMC and McRel, and only in a selection of classrooms where teachers and students were used to being observed by outsiders. Clearly, this gave rise to a Hawthorne Effect, though it allowed me to gain entry and conduct observations during the time period that the contract allowed - the last two weeks of the term, which is probably as tense a time as the first two weeks of school.

Corsaro had carefully planned ahead and used unobtrusive observation during the initial tense period, which I was unable to do. Moreover, he stayed at the site for an extended period of time, helped children put their coats on, and participated in the usual chores of running a school as a volunteer staff member rather than strictly as an outside professional observer. This was his way of enculturating himself into the school's ecological environment. One reason why I'd like to carry out my dissertation at the UCD School of Education is that I have already gained entry there, set up communications networks with other students, learned alongside them and mentored them - I am part of that culture.

Corsaro's rationale for representativeness of data collection across all subjects (how often to observe, where, and when) is explained on pp. 32-33, in which he describes the inventory of the participants' activities, social-ecological areas, and schedule of the nursery school that he created and revised several times during his initial unobtrusive observations. He refers to four other authors who have insured sample representativeness in previous studies by collecting data across people, places, time, and activities. He does this to assure himeself that the sample is representative of how the whole set of interactive episodes is distributed. He also talks about theoretical sampling, which would enable him to look for anomalies and exceptions to the general pattern that would emerge from the representative sampling.

Corsaro carefully describes his process of data collection (where he hid the microphones, how he used hand signals to communicate with his assistant during videotaping sessions), the way in which he generated an unambiguous definition of a consistent and stable sampling unit (rather than a unit of analysis) - the interactive episode (p. 24), and the form he used for his field notes (p. 25), with people, places, time, activities, and four types of notes clearly explicated. That is how he bounded his observations: an interactive episode has a beginning, a body of information, and an end, clearly defined. This sampling unit, together with the sampling method, addresses the issue: "could another researcher replicate this?" The answer is "yes" - provided replication was done in the exact same setting.

This field note format and the sampling unit create a good alternative to the usual data-collection matrix that I have used. Generally, I create a grid to match specific data gathering methods with predefined research questions - an "epistemic form" that guides my data collection process. Corsaro had no such luxury, because he started out with only a social-constructivist theory of learning and a rather vague research question - he wasn't looking for any causal links to learning or performance outcomes. He used his theoretical notes to make later sampling decisions and generate working hypotheses as the study progressed. However, the use of a consistent field note format allowed him to gather quantitative data such as the number of children who asked specific questions about his identity and the number of times children noticed the microphone, thereby allowing him to track trends and see patterns among these and other activities. Participant observation activities were directly related to emerging patterns in his field notes (p. 34).

Corsaro is very explicit about his procedures and provides a clear accounting of the process by which he analyzed evidence over time to arrive at his set of patterns and assertions. This process of data collection, analysis, and subsequent data collection was constructed with an eye to increasing reliability and validity, though he does not use those terms per se. He does not, however, deal in any depth with the issue of generalizability, as Cherland did in her dissertation (she cites Glaser; I cite Yin; Corsaro cites neither). Do other children in different settings act the same way? Do they act differently? We cannot know. Perhaps Corsaro is trying to generalize to the ways in which children try to construct meaning out of their world and of people's roles in it, via peer-to-peer interactions, but he does not tell us that up front.

Corsaro identified consistent patterns in his field notes; developed working hypotheses and collected additional data to "flesh out" his hypotheses and predict future trends in activities; catalogued and transcribed his recordings into a database of sorts ("cataloged summaries") that linked the taped sequence of actions and verbal interchanges to the participants and the context in which it occurred; carried out a microlinguistic analysis of videotaped episodes (including both verbal and non-verbal communication) to refine his working hypotheses, and "indefinite triangulation" to estimate the validity of his interpretations. Besides having some of the coding done by an independent researcher and then comparing notes, Corsaro played back the tapes to teachers, parents, research assistants, and the children themselves, audiotaping them and later comparing them with his own interpretations. The examples he gives in the report are good, though I'd have liked to see more of them. I think it's important to include a lot of direct quotes in the final narrative.

Sharing taped episodes with participants is an excellent technique that helps uncover any misinterpretations of audio data that might have gone undiscovered. Though Corsaro does link verbal interchanges with videotaped interactions, and though he does strive for his own version of inter-rater reliability, this is not what I consider "triangulation" in the ordinary sense of the word - i.e., converging lines of inquiry using a variety of qualitative methods such as interviews, observations, and a study of artifacts.

Chapter 2.
Theoretical Issues in Research on Peer Culture

Rather than concentrating on a few important underlying themes or threads, as I have done in my synthesis papers, or giving a broad overview of the literature in general without concentrating too heavily on patterns, as David Lebow did in his dissertation, Corsaro picks a few exemplary researchers (e.g., Piaget, Vygotsky, Mead, Cicourel), delineates their theories, shows how others have built on them, and then relates their theory bases to aspects of children's peer interactions that he intends to address, pointing out areas of agreement and disagreement. For example, he notes the fact that children's self-reports in response to hypothetical questions or dilemmas cannot reflect features of peer culture that can be captured by direct observation of interactive events (p. 57). In contrast, he states that he agrees with Vygotsky's view that social development results from children's everyday practical activities in their life worlds (p. 61). He also states that though he agrees with Mead's concept of children developing a collective identity as "children" as distinct from adults, but he says that when children reach the game stage, their sense of self as cooperating individuals within a structured game with rules begins to differentiate itself from the "generalized other" of Mead. These are small points, but they show intense reflection on Corsaro's part. I would have liked to see Corsaro spend more time deliberating on Habermas' idea that identity is produced through socialization by appropriate symbolic generalities, and the role that shared representations share in this process.

Personally, I prefer the "themes and categories" approach used by Cherland. Though it takes a good lit review and some exploratory research to come up with a few emergent themes and categories, which can then be reworked bit by bit, they eventually lend themselves to a cleanly structured data collection matrix. Like Cherland, I see good qualitative research - particularly case studies - as generalizing to basic social processes that underlie the issues and problems of other settings (Cherland, 1990), or to theoretical propositions (Yin, 1994), and not to other populations or universes. The approaches used by Corsaro, Cherland, and myself may be rather different, but in the end, I think the purpose is the same. My approach is to get a good feel for the issues and problems that have been studied and cited by other researchers, do some "broad-base" survey research to capture emerging patterns and themes, and then to use qualitative methods to delve into participants' perceptions of the issue being studied, be it reflections on the utility of a training program, potential uses of a new innovation, or personal views of the teaching-learning process.

As researchers, we all stand on the shoulders of giants who have gone before us, and whose views we espouse, either in whole or in part. Corsaro is no exception. His critiques of the works of the four key researchers who have influenced his own thinking is brief but thorough, and he ties their theories together with his intended purpose in carrying out his own study. He gives a good account of Cicourel's work on adult-child interactions and ties it together with his own observation of children playing a "monster" game. The key to this interaction is that the monster (Martin) was able to apply his developing reasoning abilities to the problem of making sense out of his friends' activities of fleeing from him (the "monster"), and was able to come up with an imaginative activity that was totally appropriate to the situation. However, as Corsaro points out, it would have been almost impossible for Martin to explain his actions to an adult. This further solidifies his argument that his research is important because it fills in a gap in the literature, namely, observational study of peer-peer social interactions and activities among preschool children in normal everyday settings, as distinct from clinical observations, adult-child interactions, and explanations of actions by children to adults.

In the final section of Chapter 2, Corsaro introduces his "interpretive approach" to childhood socialization, namely that they actively construct knowledge through social interactions, and that they develop their language and discourse skills in this process. The confusions and ambiguities that they perceive in interactions with adults are played out in their peer-to-peer interactions; that is where they attempt to make sense out of them, in light of their own developing world-picture and sense of identity. Thus, Corsaro breaks with the prior researchers who theorize that child development takes place linearly, through stages in which basic skills are acquired through experience. In contrast, he stresses the use of language to generate social interaction; it is through social interactions that the child and his/her peers begin to construct a common core of social knowledge - shared experiences and understandings. He notes that the adult-child language interactions of infants have been studied in detail, but there is little research on "the interface of further language learning with social development experiences" (p. 75).

The purpose of his research is to "chart a beginning for needed research on peer cultures throughout childhood and adolescence, by focusing on friendship and peer culture in the early years" (p. 75). Specifically, Corsaro wants to "extend the constructivist conception of the nature of the developmental process" (p. 73) by "viewing development as a productive-reproductive complex in which an increasing density and reorganization of knowledge marks progression" (p. 74), with a major emphasis on the use of language to carry out progressive discourse and to "acquire control over their lives through the establishment of a collective identity" (p. 75).

Interestingly, this is a line of thinking that Charles Crook (1994) explores thirteen years after Corsaro's study was published. Crook deals primarily with mediated environments in which participants construct a "longitudinal continuity" of shared experiences and understandings through a knowledge-building discourse that takes place via computer-mediated communication, independent of time and place. He is interested in studying the cultivation of intersubjective attitudes by a community of learners - "in talk and action that serve to create that which becomes held in common - and known by the participants to be held in common" (Crook, 1994: 118). Thus, Crook becomes the link who ties my research interests (computer-supported collaborative learning or "CSCL") together with Corsaro's interest in children's "joint or communal attempt to acquire control over their lives through the establishment of a collective identity" (p. 75).

It would be wonderful if I were really able to pursue this view of myself as a researcher; however, I find that I am ahead of my time in that regard because of the slowness of the technology adoption process within the School of Education and RMC Research. On the other hand, by participating in electronic discussions with a global community of educational technologists, I have become firmly convinced that mediated construction of knowledge, and more than that - of shared culture - among communities of learners, will be the future form of instructional interactions in the millennium.

Chapter 3.
Children's Conceptions of Cultural Knowledge in Role Play

In Chapter 3, Corsaro addresses some of the issues that were unresolved in Chapter 1. He actually deals with inter-rater reliability, noting that when a subsample of the data were coded independently by himself and a co-worker, the proportion of agreement between coders was 0.904 (p. 81). He incorporates a good number of rich examples of children's conversations into his narrative. He also describes a comprehensive set of communicative functions or categories (I've heard these called "language functions" by people who are studying e-mail communications), and tabulates the patterns and frequency of their use.

This is really important, because it is through language that children learn and develop fundamental ways of looking at the world and share these understandings with others. Learning is an active process in which meaning is developed through action and experience; the child is actively building an internal representation of knowledge, a personal interpretation of experience. Conceptual growth comes from the accumulation of experience, from the sharing of multiple perspectives, and from the simultaneous, ongoing, changing of those internal representations in response to the perspectives that are communicated by others - here, within the peer group. Since concepts cannot be observed directly, Corsaro wants to explore this process of conceptual growth by observing children's language interactions with their peers.

He codes the verbal exchanges based on his knowledge of discourse analysis and drops them into what is hopfully a complete set of communicative functions to examine their relative frequency of use among superordinate-subordinate peers and peers of equal status. Is there a correlation between high status and the number of imperative commands given, between low status and the number of requests for permission issued? Are there differences in frequency of use of various language functions between peers of equal status versus between peers of unequal status?

He conducts a fairly quantitative analysis of the language functions that children use in their role play as a way of uncovering their emerging conceptions of status and role, not for the purpose of accepting or rejecting any hypotheses per se. For example, those with higher status use more imperative language functions in a role play than those with subordinate status. His interpretations of these quantitative results, e.g., "clearly, the children used language to exert authority depending on the social positions they occupied in role play" (p. 80), and "children's infrequent use of these communicative patterns in subordinate -> subordinate interaction in the presence of superordinates...suggests that children may be developing a rule which restricts conversation between subordinates in the presence of superordinates (p. 85), are sensible in the light of the purpose he has explicitly stated in the permission letter: "I focus on what young children `do' and `say' together, especially on how they use language to generate meaning in peer interaction". As Guba points out, the call for qualitative (or quantitative) methods is by itself not a call for a paradigm shift (Guba, 1992: 22).

He notes that studying frequency distributions alone is insufficient to capture how children articulate their conceptions to ongoing interactive scenes (p. 85). Thus, his next step is to use (and cite) examples of transcripts to uncover the process by which these status differences are played out. For example, in examples 3.1 and 3.2, he interprets the number of imperatives issued by "parents" as a way of maintaining control over low status interactants (be they children or "kitties"), and misbehavior from subordinates as a way of feeding into this exertion of authority. In example 3.3, the transcript provides "direct support for the contention that higher status means power and giving orders" (p. 95). Similarly, the role confusion in example 3.5 (same-status interaction) indicates that children's knowledge of role expectations may be less developed than their knowledge of status. Finally, in example 3.7 (the hunters), he discusses how role play is not only important for the children's development and use of social knowledge (roles and status), but also about how it becomes part of peer culture (using clever ploys to maintain a role play that goes against the school's shared beliefs and norms of behavior, such as the rule about not pretending to use guns because they hurt people).

On p. 100 he talks about the difference in length of episode between same-status and cross-status interactions, but is careful not to draw any major conclusions because of the number of confounding variables. His care in not jumping to conclusions is admirable! Throughout the discussions that accompany each of these transcripts, Corsaro tries to tie his interpretations to the results of other researchers, to point out differences and similarities between his findings and theirs, and to interpret his own findings in the light of other theoretical frameworks (e.g., Bateson's emphasis on the relationship between context and role, on p. 111).

One of the lessons that I take from this chapter with regard to my own work is that a mix of qualitative and quantitative methods can be used appropriately, not only in ethnography, but also in needs assessment, evaluation, and action research - provided I am very clear about the purpose for which I am using each of these methods. For example, when I tabulate frequencies and purposes of e-mail or Internet use, it is to explore emerging and changing patterns of use. Patterns, however, give no insight into the personal interpretations of experience of participants. This must be explored via interviews and focus groups, and a sufficient number of direct quotations must be cited and directly linked to any inferences I intend to make concerning participants' concerns, comfort level, perceived barriers and facilitators, attitudes, and other nonobservable constructs regarding the reasons for use or non-use of technology. Steve Alessi (personal communication) has mentioned that this direct linkage of data to inferences is a "must" in published work.

References

Back to Home Page

Lorraine Sherry
Uploaded August 8, 1997