Maternal Strategy Use and Preschool Categorization Abilities.

Abstract. This study identified the strategies that mothers employed while they were engaged in an animal categorization activity with their preschool children, examined the verbal behavior of the preschool children, and determined the relationship between the maternal strategies and the preschoolers' verbalizations during the interaction. The results revealed that mothers' strategy use varied according to the type of information being taught. For example, when sorting the animals, mothers asked labeling and perceptual questions. In contrast, when clustering or grouping the animals, mothers asked questions about the similarities and differences among the grouped animals. These strategies were differentially linked to the preschoolers' verbalizations as well. That is, maternal use of perceptual questions was positively related to the preschoolers' verbal responses, but negatively related to their independent elaborations. On the other hand, during the clustering phase, maternal contrasting information question s were positively linked to the preschoolers' verbal responses and negatively linked to their independent elaborations. Education implications, as well as implications for the social facilitation of preschool categorization abilities, are discussed.

Within the past two decades, the role of adult influences on young children's cognitive and problem-solving competence has received considerable attention in the developmental and early childhood education literature (see Azmitia & Perlmutter, 1989, for review). One reason for this interest is the belief by some child development researchers that children learn the requisite skills for independent problem solving and academic performance in the context of social interaction (Pianta, 1997).

Theoretically, much of this work has been guided by the sociocultural theory of Vygotsky (1978). The central premise of the theory is that cognitive growth is a socially mediated process that occurs as a result of interaction on two levels: an interpsychological level, where the adult assumes the responsibility for structuring and guiding the interaction; and an intrapsychological level, where the child gradually begins to participate in the interaction, eventually internalizes the skills and strategies demonstrated by the adult, and performs the task with minimal adult assistance.

Empirically, investigations using the Vygotskian model have explored how mothers structure and guide the interaction while teaching their children problem-solving and memory tasks, and they also have examined the influence of these maternal behaviors on the children's independent test performance.

Saxe, Gearhart, and Guberman (1984), for example, examined the initial task structure (instructions) that mothers provided while engaged in a numbers reproduction activity with their preschool children. They discovered that children whose mothers gave detailed task instructions displayed a recall of the numerical information that was more accurate than that of children whose mothers had only provided brief instructions about the nature of the activity.

On the other hand, Rogoff, Ellis, and Gardner (1984) assessed the relationship between maternal use of such specific strategies as directives and imperatives, and the children's recall of items on an independent sorting task during a task of teaching children to group school or household items. The researchers failed to find a relationship between maternal behavior and the children's recall of the information. Using similar methodologies, however, Kontos (1983) and Freund (1990) found that mothers' emphasis on metacognitive information was strongly predictive of the preschoolers' independent task performance on both spatial and problem-solving tasks.

Although Rogoff et al. (1984), Kontos (1983), and Freund (1990) have provided rich information on how mothers guide the interaction by their use of strategies, their descriptions of the children's participation in the interaction, unfortunately, have been limited to assessing their independent test performance. Interaction researchers argue that children acquire cognitive abilities within the context of social interaction, because they are actively involved in transforming the information (Day, French, & Hall, 1985). These transformations should be evident in the content of their verbal behavior during the interaction.

A preliminary investigation examining children's verbalizations during the interaction was conducted by Diaz, Neal, and Amaya-Williams (1990). In this study, four categories of the preschoolers' verbal behavior were coded (e.g., questions, responses, repetitions of maternal behavior, and independent task verbalizations). The researchers were interested in determining which maternal strategies (e.g., commands, directives, questions) were predictive of the preschoolers' verbal behavior. They discovered that of the maternal strategies, commands, and questions were linked to the preschoolers' verbal responses to the mother during the interaction.

The goal of the present study was to investigate the structure and guidance that mothers provide when teaching their preschool children about categorical relationships, to determine how children participate verbally in a categorization activity and to assess the relationship between maternal strategy use and the preschoolers' verbal behavior.

Researchers believe that the acquisition of categorical knowledge, like the acquisition of problem solving, memory, and language abilities, is rooted in social interaction. Through direct teaching, they speculated that mothers assist their children in acquiring important knowledge about how categories are created, as well as about the distinguishing features between and among category members and groups. However, the research examining how mothers teach their children about categorical relationships is sparse.

Previous investigations have identified two types of verbal labels (i.e., ostension, and contrasting information statements) that mothers use to describe individual items from a category and to describe how items are combined to create subordinate and superordinate categories (Callanan, 1985; Gelman, Coley, Rosengren, Hartman, & Pappas, 1998; Shipley, Kuhn, & Madden, 1983). The mothers' use of these labels varies depending on whether they are teaching basic level information (e.g., names of individual items) or superordinate level information (e.g., names of grouped objects).

To further examine how mothers teach their children about categorical relationships and to explore how their teaching behaviors are related to the preschoolers' verbalizations during the interaction, the following issues were explored in the present study.

First, similar to the Saxe et al. (1984) study, the level of task instructions provided by mothers was examined. The specific question was to determine the influence of maternal task instructions on the preschoolers' verbal engagement (e.g., verbal responses, questions, and independent elaborations) in the interaction.

Second, in line with previous research measuring maternal strategy use (i.e., cues, questions, statements) the goal of this study was to investigate whether mothers would show systematic differences in thier use of strategies when teaching their children a categorization activity that involved sorting and clustering items.

Although the researchers' specific interest was similar to the categorization research, they also investigated whether mothers would show systematic differences in their use of strategies when teaching their children to sort or to cluster information.

Third, consistent with the Diaz et al. (1990) research, the relationship between maternal strategies and the preschoolers' verbal assessment in the interaction was assessed.

Method

Participants

The participants were 50 mothers and their preschool children (M = 54.8 age in months, SD = 5.51) recruited from child care centers throughout a tri-county area. There were 25 females and 25 males among the children in the study. To recruit the participants for the study, preschool teachers were asked to give permission letters to mothers of 4-year-old children. Those mothers returning a signed form indicating their willingness to participate in the study were contacted to schedule a time for the videotaping. The mothers and children were

primarily Caucasian (n=34); however, 16 of the dyads were African American. The mothers were, on average, 31 years of age (SD= 8.7), and married (n= 39); 35 of the mothers had completed, on average, 14.7 years of school, SD = 2.4. Each of the mothers was the preschooler's primary caretaker. The mothers were offered a $10 honorarium for their participation in the study.

Materials and Research Procedures

The videotaping of the interaction was conducted in the principal investigator's lab, in the psychology department, by a female research assistant. The mothers were given the following instructions:

"Today, I have an activity that I would like you to do with your child. This activity involves pairing and grouping these 13 pairs of animals (sheep, cows, dogs, monkeys, elephants, zebras, alligators, kangaroos, turtles, pigs, snakes, raccoons, foxes). First, I'd like you to teach your child the names of the animals and then I'd like you to teach your child how to group the animals (e.g., put those animals together that you think go together). Please take all the time that you need and feel free to terminate the session at anytime. There is no right or wrong way to do this activity."

The animals were selected from the Animals Families Memory Game developed by Milton Bradley. To make sure that the mothers understood the purpose of the activity, they were allowed to practice with the examiner prior to teaching their children.

Scoring System: Maternal Behavior. Maternal behavior was coded for task instructions and for strategy use. The task instructions category assessed the level of task instructions provided by the mothers about the goals of the activity and was coded as follows:

Low Explicit (1): the mother begins the game by verbally labeling the animals (e.g., "This is a parent alligator").

High Explicit (2): the mother provides an overview of the activity.

The following strategies were scored for the frequency of occurrence during the sorting (naming) and the clustering (grouping) phases of the interaction:

Labeling (L): the mother names each animal or animal group.

Perceptual Cues (PC): the mother makes references to the perceptual attributes of the animals.

Concept Instruction Statements (CIS): the mother provides a rationale for grouping the animals or sorting the animals (e.g., "This is a parent cow and this is a calf, so these belong together," "These are pigs and these are sheep, so they belong on a farm").

Labeling Questions (LQ): the mother asks the child to name an individual animal or a group of animals.

Perceptual Questions (PQ): the mother asks the child about the size, shape, color, or some physical dimension of the animals (e.g., "What color is this one?" "How many legs does this one have?").

Contrasting Information Questions (CIQ): the mother asks the child to make a comparison between animals (e.g., "How are these two animals alike?").

Praise (P): the mother praises the child for a job well done.

Metacognitive Information (M): the mother makes reference to the child's memory abilities (e.g., "You're a good rememberer"), to places (e.g., "You've been to the zoo before"), and to objects/animals (e.g., "You have a zebra printed on your sheets").

Scoring System: Preschool Task Engagement Behaviors. The following scoring system is a modification of the Diaz et al. (1990) scoring system. Each behavior was scored for the frequency of occurrence during the sorting and clustering phases of the interaction.

Verbal Responses (VR): the child responds to maternal task relevant questions and statements.

Clarifying Questions (Q): the child asks the mother questions about the task (e.g., "Does the animal go here?" "What's this animal, or group of animals, called?").

Independent Elaborations (IE): the child states plans about the task (e.g., "I'm gonna put the snakes together") or monitors his or her own behavior (e.g., "This doesn't belong here").

Reliability of Coding

To obtain reliability, 40% of the videotapes were coded by two undergraduate students. Cohen's Kappa was used to assess reliability, and discrepancies were resolved through discussion. The remaining tapes were divided and coded independently by the scorers. Table 1 presents the Kappa co-efficients for each of the variables.

Results

Prior to data analysis, the eight maternal teaching strategies and the three preschoolers' verbal behaviors were converted to proportions.

Maternal Task Instructions and Preschoolers' Verbal Behavior

To examine the influence of maternal level of task instructions on preschoolers' verbal engagement, a one-way Multivariate Analysis of Variance (MANOVA) was performed, with the maternal level of task instructions serving as the independent variable and the three preschoolers' verbal engagement behaviors functioning as the dependent variables. The results of the MANOVA indicated that maternal level of instructional support was not significantly associated with the preschoolers' overall verbal engagement in the interaction, F (10, 32)=.65, p[greater than].05.

Maternal Strategy Use During the Sorting and Clustering Phases

To investigate the question of whether mothers would show systematic differences in their use of strategies when teaching their children how to sort and cluster the animals, t-tests were applied to the data. The mean proportions for the maternal strategies are presented in Table 2. Mothers asked significantly more Labeling Questions, t (49)=2.10, and Perceptual Questions, t (49)=7.55, during the sorting phase of the interaction than during the clustering phase of the interaction. On the other hand, they asked significantly more Contrasting Information Questions during the clustering phase of the interaction than during the sorting phase, t(49)=11.45. There were no significant differences in maternal use of Labels, Concept Instruction, Praise, Perceptual Cues, and Metacognitive Information during the sorting and clustering phases of the interaction.

Maternal Strategy Use and Preschoolers' Verbal Behavior

To examine the relationship between maternal strategy use and the preschoolers' verbal engagement, Pearson Product Moment Correlations were used. Maternal labeling and concept instruction statements were negatively related to preschoolers' verbal responses, whereas maternal use of perceptual questions was positively linked to children's verbal responses. Maternal use of perceptual questions was negatively associated with the preschoolers' independent elaborations. The means and standard deviations for the preschoolers 'verbal behavior are reported in Table 3. Table 4 reports the results of the correlations for the sorting phase of the interaction.

The correlations for the clustering phase of the interaction are reported in Table 5. Maternal labeling and concept instruction statements were negatively associated with the preschoolers' verbal responses, whereas contrasting information questions were positively linked to the preschoolers' verbal responses. Maternal labels were positively associated with the preschoolers' independent elaborations, but contrasting information questions were negatively related to their independent elaborations.

The simple correlations demonstrated that several of the maternal strategies for both the sorting and clustering phases of the interaction were related to the preschoolers' verbalizations. To determine which of the maternal strategies emerged as the strongest predictor of the preschoolers' verbalizations, multiple regression analyses were performed for the sorting and clustering phases of the interaction.

Sorting Analysis

In the first regression analysis, the predictor variables in the model were the maternal teaching strategies (Labeling, Perceptual Cues, Concept Instructions Statements, Labeling Questions, Perceptual Questions, Contrasting Information Questions, Praise, Metacognitive Information), and the criterion variable was the preschoolers' verbal responses. The regression analysis approached significance, F (8, 49)=24.30, p [less than].06, with maternal perceptual questions emerging as the strongest predictor variable in the model.

The second regression model included the maternal teaching variables as the predictor variables and the preschoolers' questions as the criterion variable. This model failed to reach significance, F (8, 49)=.42, p [greater than].05.

The third regression model, which included the maternal teaching behaviors as the predictor variable and the preschoolers' independent elaborations as the criterion variable, failed to reach statistical significance, F (8,49)=l.64,p [greater than].05.

Clustering Analysis

The first model--with the maternal teaching strategies as the predictor variable and the preschoolers' verbal responses as the criterion variable--was significant, F (8, 49)=2.38,p[less than].05, with maternal contrasting information questions emerging as the significant predictor variable in the model.

The second regression model, which included the maternal teaching strategies as the predictor variable and the preschoolers' independent elaborations as the criterion variable, was significant, F (8, 49)=2.28, p [less than].05, as well. The significant predictor variable was maternal use of contrasting information questions.

The third model--with the maternal teaching strategies as the predictor variable and the preschoolers' questions as the criterion variable--did not emerge as being statistically significant, F (8, 49)=.46p [greater than].05.

Discussion

The major goal of this study was to investigate the methods mothers used to teach their 4-year-old children about categorical relationships, and to examine how the preschool children participated in the interaction, as expressed through their verbalizations.

How do mothers initially structure an interaction about categorical relationships?

Although the majority of the mothers provided brief instructions about the nature of the activity, there were variations in their brief preparatory statements about the task. For example, some mothers began the task by labeling the animals (e.g., "This is a raccoon"), while other mothers began the task by asking the child to label the animal (e.g., "What's this animal called?"). Mothers also provided brief instructions about clustering the animals (e.g., "Now we're gonna put all of the animals together that belong together").

This method of segmenting the task instructions was found in the Saxe et al. (1984) study. Those researchers discovered that mothers gave additional detail about the task as the array increased in size. This pattern of teaching suggests that first, mothers are sensitive to their children's limited information processing capacities, and that second, they are aware of their children's growing knowledge base about categories.

What specific strategies do mothers employ when teaching their preschoolers about categorical relationships?

The mothers used different strategies for describing the individual animals than the ones they used for describing groups of animals. For example, when sorting the individual animals, mothers engaged in name recognition by asking their children to name the animals, and they engaged in elaborative questioning by asking their children about the perceptual qualities of the animals. In contrast, when mothers were teaching their children how to combine the pairs into larger groups, they asked questions about the similarities and differences between the grouped animals (e.g., "How does the egg group differ from the farm group?").

The link between the maternal strategies and the preschoolers' verbalizations was quite different for the two phases of the interaction. The variation found in two of the children's verbalizations (i.e., verbal responses and independent elaborations) was in part due to the types of cues and questions posed by the mothers. For instance, during the sorting phase, the children were more likely to respond when mothers asked questions that focused on the perceptual attributes of the pairs. Interestingly, those same questions suppressed their independent elaborations. On the other hand, when the information processing demands of the task increased and the mothers were teaching their children how to group the animals, the more that they asked questions about the similarities and differences between the animals, the more the children engaged in verbal responding to those questions. However, the children generated fewer independent elaborations when mothers asked similar types of questions about the grouped animals.

Why might two distinct patterns of teaching and relationships exist when discussing categories?

First, mothers may have anticipated the difficulty that children would have with understanding how individual animals are combined to create grouped animals, and so the mothers may have adjusted their teaching to provide explicit information that would assist the preschoolers in distinguishing the animals at the two levels. Callanan (1985) and Shipley et al. (1983) came to similar conclusions, based on their study of mothers' labeling basic level and superordinate level information. Hupp, Conroy, and Able (1984) also found strategies that differ for describing individual objects and grouped objects were effective instructional methods to use when interacting with children in a categorical activity.

Second, in terms of the relationship between maternal teaching and the preschoolers' verbal engagement in the activity, the data indicate that certain maternal strategies were strong predictors of the preschoolers verbalizations, while others were not. One implication that may be tentatively drawn from this result is that both mothers and children perceived that basic level and superordinate level information must be treated differently, as evidenced by the pattern of their verbalizations. A second implication pertains more specifically to the role of children's verbalizations during an interaction. According to Diaz et al. (1990), children's verbal responses and independent elaborations during an interaction serve two purposes. One, when children verbally respond, they provide the mothers with information about their attention and motivation levels. When they generate independent elaborations, they demonstrate to the mothers that they are transforming the information to match their knowledge base, monitoring their progress, and entering the initial stages of assuming "cognitive" control over the interaction. Diaz et al. (1990) speculate that because independent elaborations represent a more advanced method of approaching the material than do verbal responses or questions, future studies should identify those maternal strategies that increase the likelihood of children generating independent elaborations.

In summary, the focus of this study was to provide preliminary information on how mothers structure and guide a teaching activity about categorical relationships. The results, at some level, support and extend the social interaction model of preschool cognitive performance in the following ways: In support of the model, the data suggest that mothers are flexible in their initial approach and subsequent use of teaching strategies while engaged in a categorization activity with their preschool children. Second, in the specific exploration of the nature of children's verbalizations during the interaction and the measurement of the association between maternal teaching strategies and the preschoolers' verbalizations, the model has been extended. Our findings indicated that several maternal strategies were related to preschoolers' verbalizations. Certainly, caution must be taken in interpreting the direction of effects. Children bring different attitudes and behaviors to a learning situation that clearly influenc e the types of strategies that mothers employ while teaching.

What is the implication of this research for early childhood education? One major implication is that preschool children, with adult guidance and structure, can begin to develop rudimentary classification skills far earlier than expected. Children in their home environment often are involved in discussing with their parents such categories of objects as shapes, colors, and animals. The critical ingredient in fostering their understanding of class inclusion relationships is the type of verbal strategies that parents employ. Based on the findings of this study, verbal strategies should not only include labeling similarities and differences between objects, but also involve questioning the preschoolers about those attributes.

References

Azmitia, M., & Perlmutter, M. (1989). Social influences on children's cognition: State of the art and future directions. Advances in Child Development and Behavior, 22, 89-144.

Callanan, M.A. (1985). How parents label objects for young children: The role of input in the acquisition of category hierarchies. Child Development, 56,508-523.

Day, J., French, L. A., & Hall, L. (1985). Social influences on cognitive development. In D. L. Forrest-Pressley, G. E. MacKinnon, & T. G. Waller (Eds.), Metacognition, cognition and human performance (pp. 33-53). Orlando, FL: Academic Press.

Diaz, R., Neal, C., & Amaya-Williams, M. (1990). The social origins of self-regulation. In L. Moll, (Ed.), Vygotsky and education: Instructional implications and applications of socio-historical psychology (pp. 127-154). New York: Cambridge University Press.

Freund, M. (1990). Maternal regulation of children's problem solving behavior and its impact on children's performance. Child Development, 61, 113-126.

Gelman, S., Coley, J., Rosengren, K, Hartman, E., & Pappas, A. (1998). Beyond labeling: The role of maternal input in the acquisition of richly structured categories. Monographs of the Society for Research in Child Development, 63 (1, Serial No. 253).

Hupp, S., Conroy M., & Able, H. (1984). Designing instructional programs to facilitate generalization of object categories by young handicapped children. Journal of the Division for Early Childhood, 10, 149-155.

Kontos, S. (1983). Adult-child interaction and the origin of metacognition. Journal of Educational Research, 77, 43-64.

Pianta, R. (1997). Adult-child relationship process and early schooling. Early Education and Development, 8, 11-26.

Rogoff, B., Ellis, S., & Gardner, W. (1984). Adjustment of adult-child instruction according to the child's age and task. Developmental Psychology, 20, 193-199.

Saxe, G., Gearhart, M., & Guberman, S. (1984). Social processes in early number development. Monographs of the Society for Research in Child Development, 52, (2, Serial No. 216).

Shipley, E. F., Kuhn, I. F., & Madden, E. C. (1983). Mothers' use of superordinate category terms. Journal of Child Language, 10, 571-588.

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

Table 1
Reliability Co-efficients for Maternal and Child Strategies
Perc. Cues                       .80       .85
Conc Instr.                      .85       .82
Labeling Q.                      .92       .85
Percept. Q.                      .90       .96
Constr. Q.                       .85       .95
Praise                           .90       .80
Meta                             .95       .80
Preschoolers' Verbal Behavior  Sorting  Clustering
Verbal Responses                 .90       .90
Questions                        .85       .80
Ind. Elaborations                .80       .85
Table 2
Means and Standard Deviations for Maternal Strategies
Phase of the Interaction  Sorting   Clustering
Strategies
Labeling                  .13(.10)   .08(.02)
Perc. Cues                .08(.06)   .08(.07)
Conc Instr.               .29(.12)   .26(.14)
Labeling Q.               .06(.06)   .04(.04)
Perc. Q.                  .22(.1O)   .11(.11)
Contr. Q.                 .02(.03)   .23(.12)
Praise                    .19(.08)   .17(.09)
Meta                      .01(.02)   .02(.02)
Table 3
Means and Standard Deviations for Preschoolers' Verbal Behavior
Phase of the Interaction  Sorting   Clustering
Strategies
Responses                 .44(.22)  .50(.17)
Questions                 .09(.09)  .10(.08)
Elaborations              .46(.20)  .39(.16)
Table 4
Correlations Between Maternal Strategy Use and Preschoolers'
Verbal Behavior During the Sorting Phase
Preschool Behavior     Responses  Questions  Elaborations
Maternal Behavior
Labeling                 -.27 *      .21         .16
Perceptual Cues          -.24        .07         .20
Concept Instruction      -.27 *      .16         .18
Labeling Questions        .13       -.20         .02
Perceptual Questions       55 **     .15        -.47 **
Contrasting Inf. Ques     .11       -.05        -.08
Praise                    .05       -.16         .02
Metacognitive             .03       -.01        -.02
(*)p[less than].05,
(**)p[less than].001
Table 5
Correlations Between Maternal Strategy Use and Preschooler's
Verbal Behavior During the Clustering Phase
Preschool Behavior           Responses  Questions  Elaborations
Maternal Behavior
Labeling                       -.44 *      .12        .43 **
Perceptual Cues                -.11        .11        .07
Concept Instruction            -.31 *      .24        .21
Labeling Questions              .14       -.28 *     -.01
Perceptual Questions            .19       -.10       -.15
Contrasting Info. Questions     .38 **    -.10       -.37 **
Praise                          .08       -.13        .06
Metacognitive Information       .07        .07        .06
(*)p, [less than].05,
(**)p, [less than].001