The Effects of Child-Centered, Teacher-Directed, and Scaffolded Instruction on Low-Income, Latino Preschoolers' Task Performance, Motivation, and Private Speech

Angela Willson-Quayle

Advisor: Adam Winsler, PhD, Department of Psychology

Johnson Center, 237
April 02, 2001, 08:00 PM to 07:00 PM



Angela Willson-Quayle, M.A.

George Mason University, 2001

Dissertation Director: Dr. Adam Winsler

The present study set out to examine the effects of child-centered, teacher-directed, and scaffolded instruction on low-income, Latino preschoolers' task performance, private speech, and motivation (i.e., affect, on-task persistence, preference for challenge, independence, intrinsic motivation, and self-perceived competence). Using a pretest-posttest design, 61 preschoolers were randomly assigned to a child-centered, teacher-directed, or scaffolded teaching condition to work one-on-one with the experimenter on a Duplo Lego construction task. Levels of teacher control, structure, and responsiveness were differentially varied to reflect the basic elements of each instructional approach in a laboratory-type setting. Data were gathered by means of structured laboratory observation of children, parental interviews, and teacher report.

It was hypothesized that scaffolded children would display the highest task performance, the most optimal patterns of motivation, and the most advanced levels of private speech. Compared to child-centered children, it was further expected that the teacher-directed group would show higher task performance, less optimal patterns of motivation, and less advanced private speech.

The results only partially supported the hypotheses for the main outcomes. More specifically, almost all children showed posttest increases in task performance after the dyad session, although children in the child-centered group remained about the same. All children displayed less positive emotion after the dyad session due to increased boredom/frustration with the task. Scaffolded children showed significantly more persistence than child-centered participants, but not more than teacher-directed children. Interestingly, the pattern of posttest means revealed that while the persistence of the scaffolded and teacher-directed children increased, that of the child-centered group decreased. This is an important finding in that it suggests not only that some structure in teaching increases persistence but also that a lack of structure decreases children's efforts at completing a task. Children in the teacher-directed group produced less task-irrelevant private speech than children in the scaffolded group. Although overall results for task performance, affect, task-relevant, partially internalized, and total private speech were not statistically significant, the pattern of posttest means did fall in the hypothesized direction. In addition, no significant differences across group were detected for independence, challenge preference, intrinsic motivation, self-perceived competence, ability- and affectively-relevant self-statements, fragmented speech, and self-regulatory private speech. A major limitation of the study was the overall low frequency of private speech for children in this sample. By and large, the majority of children emitted low levels of private speech, but a sizeable minority produced high amounts. This resulted in an absence of normally distributed scores and overly large standard deviations. In terms of secondary results, it was found that children's English language proficiency was positively associated with performance on the Duplo Lego task whereas Spanish language proficiency was not. Children who preferred to use English (over Spanish) were more persistent and more intrinsically motivated. Children who preferred to speak English with the experimenter produced more positive self-evaluative statements than those who preferred Spanish.

It appears that no one teaching approach is clearly better than any other in terms of increasing low-income, Latino preschoolers' task performance, motivation, and private speech. Scaffolding appears to be best for increasing children's persistence on tasks and the pattern of posttest means for some non-significant findings suggests that some amount of structure (teacher-directed or scaffolding) is beneficial to such outcome variables as task performance and independence. Minimal structure, on the other hand, as evidenced by the child-centered approach did not appear to produce any gains. In addition, strong English language skills appear to be critical to these children's perceptions of self-competence in school-based activities and, as such, should be carefully considered in curriculum building and implementation.

In conclusion, it is evident that scaffolding as an instructional approach needs further exploration. Furthermore, it is clear that we should not assume that teacher-directed instruction has no merits nor that child-centered teaching (in its most liberal form) can do no wrong. Rather we should appreciate that teacher-directed and child-centered teaching are neither all bad nor all good. Finally, based on the present findings and prior research, it is evident that differing teaching approaches have differential effects on diverse groups of children. As a result, early childhood education should not be seen as an educational issue of one size fits all. Rather, the research depicts a more varied and complex picture, especially in the case of low-income, Latino preschoolers.