The Development of Knowledge Structures in Children

Deborah Roedder John, University of Wisconsin-Madison
ABSTRACT - Although schemas have been the subject of much interest in recent years, issues surrounding the development of these knowledge structures have received little attention. The purpose of this paper is to examine several conceptual frameworks which describe how schemas develop and provide a starting point for investigating schema development in empirical settings. These descriptions are then used as the basis for a preliminary study of schema development in children of different ages.
[ to cite ]:
Deborah Roedder John (1985) ,"The Development of Knowledge Structures in Children", in NA - Advances in Consumer Research Volume 12, eds. Elizabeth C. Hirschman and Moris B. Holbrook, Provo, UT : Association for Consumer Research, Pages: 329-333.

Advances in Consumer Research Volume 12, 1985      Pages 329-333

THE DEVELOPMENT OF KNOWLEDGE STRUCTURES IN CHILDREN

Deborah Roedder John, University of Wisconsin-Madison

ABSTRACT -

Although schemas have been the subject of much interest in recent years, issues surrounding the development of these knowledge structures have received little attention. The purpose of this paper is to examine several conceptual frameworks which describe how schemas develop and provide a starting point for investigating schema development in empirical settings. These descriptions are then used as the basis for a preliminary study of schema development in children of different ages.

INTRODUCTION

The importance of prior knowledge in learning and problem-solving has been a central concern of researchers for over fifty years. Bartlett (1932) was the first of many investigators to demonstrate the importance of knowledge structures (schemata) in guiding the comprehension and interpretation of new information. Recent studies have provided further support for the notion that individuals utilize schemata in a wide variety of processing situations for a variety of purposes. Individuals appear to have schemata that govern our expectations regarding the appearance of objects and places (scene schemata), the order in which certain sequences of events will happen (event schemata or scripts), and the structure of actions occurring in stories (story schemata). These schemata in turn provide a framework for perceiving (Minsky 1975), comprehending (Bransford and Johnson 1973), and recalling new information (Bower, Black, and Turner 1979; Brewer and Treyens 1981; Goodman 1980; Mandler and Johnson 1977).

Little attention, however, has been directed toward understanding the development of schematic knowledge structures. Because schemas play such an important role in guiding many aspects of information processing, the question of how schemas develop and how they change over tine would appear to be of key importance. Various researchers have suggested that new schemata are developed through experience and observation (Nelson and Gruendel 1981) or through transformations and modifications of existing schema (Rumelhart and Ortony 1977). As they develop, investigators believe that schemata become more complex and abstract in content (Schank and Abelson 1976; Taylor and Winkler 1980). Unfortunately, little empirical evidence has emerged to examine many of these contentions.

This paper addresses these questions by introducing and contrasting several views of schema development. These views provide the conceptual basis for further research by specifying changes in the content and utilization Of schemas as they develop. This paper also presents some preliminary evidence for these developmental changes in a study with children of different ages. Children of different ages seemed to be a particularly appropriate sample for detecting developmental change due to wide variations in knowledge bases which occur in younger and older children. The implications of these findings for future research and suggestions for extending this line of inquiry are also discussed in a concluding section.

CONCEPTUALIZING SCHEMA DEVELOPMENT

Although schema development has not received much empirical attention, several researchers have developed hypotheses about the changes that occur in the course of schema development. Each of these frameworks characterizes schema development as progressing in a series of stages or levels with each stage qualitatively different from one another. Stages are Usually described as differing in terms of the abstractness and complexity of the knowledge structure. Each framework is described in more detail below.

Taylor and Winkler

Taylor and Winkler (1980) describe schema development as occurring in four distinct stages: rudimentary (or episodic), stereotypic (or novice), relative expert, and automatic phase. At the earliest stage of development, knowledge may consist of a specific example which an individual has experienced or observed. In the next stage, the stereotypic or novice phase, individuals move from knowledge of specific examples to knowledge of stereotypic attributes which define a situation, person, or object. These attributes are likely to be based upon information which an individual encounters in initial or frequent contacts with the person, object, or situation. For example, firemen may be described in terms of stereotypic attributes such as fighting fires and saving cats in trees. During this phase, individuals are likely to overgeneralize on the basis of limited information and are not likely to be sensitive to information which i 6 discrepant with their stereotypes. In the third phase, relative expertise, individuals become more sensitive to discrepant information and may use it to refine and qualify existing knowledge structures. As a result, schemas become more complex and may contain conditional statements which qualify simple generalizations ("if a plant's leaves begin to drop off, it may be overwatering . . . unless it has been very hot, in which case the plant could be scorched ). In the final stage, the automatic phase, the content of the schema remains unchanged but the access to and use of the schema becomes automatic or habitual.

Abelson

Abelson (1976) describes the development of event schemata or scripts as progressing through several stages. As individuals acquire new experiences, their representations of the sequences of events common to these experiences include more abstract statements and more complex series of events. Script development is viewed as progressing through three levels: episodic, categorical, and hypothetical. The episodic level is characterized by an emphasis on single incidents with little generalization or abstraction from incident to incident. Scripts at this level are often represented by details of a single experience. At the categorical level, scripts begin to include general statements which demonstrate abstraction from specific details of any one incident. However, specific examples are often mentioned along with simple generalizations at this stage. As scripts progress to the hypothetical level, they are characterized by even more abstraction and more complexity. Scripts at this level are usually void of any specific details and contain only abstractions and general statements. m is level is also characterized by more complex sequences of actions organized hierarchically and contained in conditional clauses (if . . . , then . . .).

Additional clarification regarding the qualitative differences between the episodic, categorical, and hypothetical levels has been recently provided by Martin and her colleagues (Martin, Harrod, and Siehl 1980). To facilitate the coding of protocols into one of these three levels, these investigators have proposed several additional characteristics which distinguish different levels of script development: number of words indicating frequency, type of nouns, verb tense, and types of logical connectors. Words indicating frequency, such as "sometimes" and "usually," are characteristic of only the categorical and hypothetical level. Because episodic scripts typically describe a single incident, there is little need for frequency indicators to summarize multiple incidents. The types of nouns used to describe event sequences also varies depending upon the level of script development. Mention of nouns at a concrete level ("me" or "Sharon, my friend") are characteristic of the episodic stage, whereas abstract nouns ("salesperson" or "individual") are more frequently used in categorical and hypothetical scripts. When describing events, verb tense also tends to be different according to developmental stage. Events in episodic scripts are described in the past tense ("I looked for the salesman") whereas events in categorical scripts are described in the "eternal" present tense ( A shopper looks around before selecting an item"). The hypothetical level is characterized by the use of eternal" present tense or future tense. Finally, different types of logical connectors between events tend to be found in different stages of development. The episodic level is characterized by the use of simple connectors such as "and" and "while"; the categorical stage is characterized by connectors such as "or" or" for example"; ant, the hypothetical level is characterized by connectors such as "if," "then," and "unless."

Gruendel

Gruendel (1980) distinguishes between -two different levels of script development: simple sequences of acts or conditional sequences. At the lowest level of development, scripts are characterized by individual events listed in a tempo al order. In contrast, more sophisticated scripts are characterized by conditional sequences. a ese sequences are defined as series of events embedded within other events and series of events connected with conditional "if-then" and "when" clauses.

Comparison of Approaches

These conceptualizations of schema development share several features in common. First, all three describe schema development as progressing in a series of stages or levels which are qualitatively different from one another. Second, there is considerable overlap between these approaches in the dimensions they use to distinguish different stages of development. The degree of abstraction is a key element in the developmental sequence described by Abelson and Taylor and Winkler. Both approaches describe the lowest level of development as one in which single experiences or instances form the basis for the knowledge structure. Generalizations based upon multiple experiences or examples are characteristic of more highly-developed scripts. All three approaches view the highest level of development as including complex sequences of events often connected with conditional clauses. Sequences of events embedded within one another are described as occurring in Abelson's hypothetical stage and Gruendel's conditional sequence stage. Sequences of events connected by qualifiers and conditional phrases ("if-then ) are described as characterizing Abelson's hypothetical stage, Taylor and winkler's expert phase, and Gruendel's conditional stage.

Although the similarities are evident, these approaches also vary in several respects. First, each approach specifies a different number of stages in characterizing the developmental sequence. Taylor and Winkler describe four stages, Abelson specifies three stages, and Gruendel considers two stages. A more important distinction between approaches is the focus of developmental differences among stages. Both Abelson's and Gruendel's stage descriptions are based upon qualitative differences in the content and organization of schemas. Taylor and Winkler's stage descriptions are based upon differences in the content of schemas and differences in the way in which these schemas are utilized in everyday situations. In Taylor and Winkler's characterization, the difference between the relative expert and expert stage is largely attributed to the naive dependence on schematic knowledge structures exhibited by relative experts. The difference between the expert and automatic stage is based upon the degree of conscious effort which must be exerted to access and use schematic knowledge structures. Considering these differences, Abelson's and Gruendel' 8 approach seems better suited for studying structural changes in development whereas Taylor and Winkler's approach seems better suited for studying process differences in schema development.

PRELIMINARY EVIDENCE

Preliminary evidence regarding schema development was collected in 8 study of script acquisition in children. Scripts for a familiar purchase situation, grocery shopping, were elicited from children in three age groups (ages 4-5, 6-7, and 9-11). The age range from four to eleven was selected to be broad enough to detect changes in script development. Fourteen children in each age group were asked to describe "what happens when you go grocery shopping" and were instructed to remember the events in the order in which they generally happen. As children provided their descriptions, interviewers provided necessary prompts to encourage children to remember as much as possible about grocery shopping events. This procedure was adopted based upon its successful use in previous studies of knowledge development in children (Nelson and Gruendel 1979a, 1979b). Children's responses were recorded on cape and transcribed after the completion of the experiment. Based upon Abelson's framework, children's protocols were classified into one of three levels of development: episodic, categorical, or hypothetical. Examples of children's scripts classified by category are provided in Table 1.

Children's scripts were expected to vary across age groups by virtue of the fact that older children should have more experience and information regarding grocery shopping. If experience and observation are the basis for script development as researchers believe, older children with more experience should exhibit more well-developed scripts than younger children. Episodic scripts should be more frequent in younger children whereas categorical or hypothetical scripts should be observed in older children.

TABLE 1

EXAMPLES OF GROCERY SHOPPING SCRIPTS

TABLE 2

PERCENTAGES OF EPISODIC, CATEGORICAL, AND HYPOTHETICAL SCRIPTS

The percentages of children in each age group classified as having episodic, categorical, or hypothetical scripts are presented in Table 2. An examination of this table reveals that most of the scripts elicited from children were categorized as either episodic or categorical. Only one script approached the hypothetical phase in terms of abstraction, lack of specific details and examples, and use of qualifiers for general statements. E us, the major differences in this case appear to be concentrated at the episodic and categorical stages of development.

Differences between age groups were examined by comparing the proportion of children in each group classified as episodic, the proportion of children in each age group classified as categorical, and the proportion of children in each age group classified as hypothetical. As expected, episodic scripts were most common among the youngest children in the sample. More children in the youngest age group exhibited episodic scripts than children in the middle age group (Z-3.42 p<.01) or the oldest age group (Z-4.53, p<.01). [The r values reported here have been corrected for a rise in the error rate due to multiple comparisons.] Children in the middle age group also tended to exhibit episodic scripts at a slightly higher rate than children in the oldest age group (Z-1.75, p-.08). As predicted, the youngest subjects produced fewer categorical scripts than children in the middle age group (Z-3.42, p<.01) and children in the oldest age group (Z-4.16, p<.01). Children in the middle age group and oldest age group did not differ with respect to the proportion of categorical scripts produced (Z-1.08, p-.26). And, finally, children in all three age groups did not differ from one another in the production of hypothetical scripts (Z<1 for all comparisons).

These results support the existence of different stages of script development and lend further credibility to the contention that scripts developed the increasing levels of experience. As expected, older children with more experience exhibited more sophisticated scripts than younger children with little or no experience. Episodic scripts were characteristic of the youngest children whereas categorical scripts were more characteristic of older children.

One result which was unexpected was the absence of hypothetical scripts in this sample of children. This finding might be due to the age range included in this sample or by the nature of the subject area (grocery shopping) used in this study. It may be that hypothetical scripts develop at a much slower rate and are exhibited by children who are older than those included in this sample. It is also possible that the particular situation studied here may not be complex enough to produce hypothetical scripts. me sequences of events involved in grocery shopping may not be complex enough to produce the type of embedded sequences of events and conditional clauses which are typical of hypothetical scripts.

SUMMARY AND CONCLUSIONS

Despite the recent interest in schemas as memory structures which influence and guide many aspects of information processing, issues surrounding the development of these knowledge structures have received little attention. Several conceptual frameworks were reviewed here as a basis for investigating the changes in content and usage that accompany schema development. Findings supportive of these changes were also reported as a preliminary step toward understanding the progression of schema development.

These findings, and the conceptual frameworks upon which they were based, need to be extended in several ways. In terms of conceptualization, the characterizations of different stages of development provided by various frameworks need to be expanded and clarified. Most of the stage descriptions are too general in nature to be used as a basis for measuring schema development in empirical settings. More detail is needed to identify stages of development in a more precise manner. Extensions of general stage descriptions, such as the one developed by Martin and her colleagues (1980), would offer additional guidance in studying different developmental phases.

The empirical findings reported here also need to be extended in at least three ways. First, schema development should be studied with a variety of subject populations and topics to provide a fuller view of developmental differences. Additional data would serve to modify and expand our ideas about the changes which occur in schemas as they develop. A second area in need of further research concerns the differences in schema utilization as a component of the develop mental process. The evidence presented in this paper is limited to differences in the content of schemas at varying levels of development. As Taylor and Winkler's (1980) framework suggests, it seems plausible that differences in utilization as well as content should be involved in the course of development. And, finally, the findings reported here need to be extended to look at the dynamic aspects of schema development. The approach used here examined existing differences in schema development among groups which were thought to vary in terms of their level of experience and information about a particular topic. In order to examine schema development in a more dynamic manner, future efforts should actively vary the amount of experience or information individuals possess to insure better control and provide a better opportunity to observe developmental changes.

REFERENCES

Abelson, R. P. (1976), "Script Processing in Attitude Formation and Decision Making, in J. S. Carroll and J W. Payne (Eds.), Cognition and Social Behaviors Hillsdale. NJ: Lawrence Erlbaum Associates.

Bartlett, F. C. (1932), Remembering: A Study in Experimental and Social Psychology, Cambridge, England: Cambridge University Press.

Bower, G. H., Black, J. B., and Turner, T. J. (1979), "Scripts in Memory for Text," Cognitive Psychology, 11, 177-220.

Bransford, J. D. and Johnson, M. (1973), "Considerations of Some Problems of Comprehension, in W. G. Chase (Ed.), Visual Information Processing, New York: Academic Press.

Brewer, W. F. and Treyens, J. C. (1981), "Role of Schemata in Memory for Places, Cognitive Psychology, 13, 207-230.

Goodman, G. S. (1980), "Picture Memory: Row the Action Schema Affects Retention, Cognitive Psychology, 12, 473-495.

Gruendel, J. (1980), Scripts and Stories: A Study of a Children's Event Narratives, Doctoral Dissertation. Yale University.

Mandler, J. M. and Johnson, N. S. (1977), "Remembrance of Things Parsed: Story Structure and Recall," Cognitive Psychology, 9, 111-151.

Martin, J., Harrod, W. and Siehl, C. (1980), "The Development of Knowledge Structures," Research Paper No. 557, Graduate School of Business, Stanford University.

Minsky, M. (1975). "A Framework for Representing Knowledge ," in P. H. Winston (Ed.), The Psychology of Computer Vision. New York: McGraw-Hill.

Nelson, K. and Gruendel, J. M. (1979a), From Personal Episode to Social Script: Two Dimensions in the Development of Event Knowledge, Paper presented at the biennial meetings of the Society for Research in Child Development, San Francisco.

Nelson, K. and Gruendel, J. M. (1979b), At Morning It's Lunchtime: A Scriptal View of Children's Dialogues, Discourse Processes, 2, 73-94.

Nelson, K. and Gruendel, J. M. (1981), "Generalized event representations: Basic Building Blocks of Cognitive Development, in A. Brown and M. Lamb (Eds.), Advances in Developmental Psychology, Hillsdale, NJ: Lawrence Erlbaum Associates.

Rumelhart, D. E. and Ortony, A. (1977), "The Representation of Knowledge in Memory," in R. C. Anderson, R. J. Spiro, and W. E. Montague (Eds.), Schooling and the Acquisition of Knowledge, Hillsdale, NJ: Lawrence Erlbaum Associates.

Schank, R. C. and Abelson, R. P. (1977), Scripts, Plans, Goals, and Understanding, Hillsdale, NJ: Lawrence Erlbaum Associates.

Taylor, S. E. and Winkler, J. D. (1980), "Development of Schemas, Paper presented at the American Psychological Association annual meeting, Montreal, Canada.

----------------------------------------