On “Adults’ cognitive demands at home and at nursery school”…
Tizard, Hughes, Pinkerton, and Carmichael (1982) were interested in four-year-olds as they compared adults’ cognitive demands on girls at home and at nursery school. The two treatments for each girl were being at nursery school with a teacher(s) in the morning and at home with her mother in the afternoon. The independent variable was the number of cognitive demands, per each of four categories, made on the child by either the teacher or the mother. Intercoder reliability on different aspects ranged from 84% to 91%, and the content validity of the codes was explained by the authors. The static—group comparison design (Campbell & Stanley, 1963) was used.
Fifteen middle-class and 15 working-class girls were selected for the study, and all parents of the selected children agreed to participate in the study. It was reported that the working-class children were probably typical of the majority of those who attended half-day nursery school (while no such comment was made about the middle-class subjects). The statistical design was apparently a two-by-two, two—way ANOVA, comparing home to school and middle- to working-class. The authors did not explain whether they transformed the nominal data to an interval-type form. If they did not, ANOVA was inappropriate. If they did, each child was observed in 100 “turns of talk” so an effect size of .40 could have been assigned (E. W. Courtney, personal communication, July 16, 1986). With alpha at .05 and beta at .20, the minimum sample size should have been 26 subjects per cell (Cohen, 1969); the researchers had 15.
Tizard, Hughes, Pinkerton, and Carmichael (1982) found, at statistically significant levels, that:
1. Cognitive demands were used more frequently by mothers of both social class groups than by teachers.
2. More often cognitive demands were not answered by children at school than at home.
3. There were fewer questions asked by children at school.
4. There was no social class difference in the hourly rate, percent, or range of cognitive demands addressed to children at home or at school.
The weaknesses of the study include the internal and external threats to validity (selection, mortality, interaction of selection and maturation, etc., and interaction of selection and the treatment) for the static-group comparison design (Campbell & Stanley, 1963), the small sample size, and the ambiguity of how the statistical analysis was executed. The researchers reported that they were careful to get homogeneous groups of girls for their study. Keeping in mind the design weaknesses, the researchers’ findings are still fairly believable, because the data are so definite in the direction of the conclusions that they made.
Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi—experimental designs for research. Boston, MA:
Houghton Mifflin Co.
Cohen, J. (1969). Statistical power analysis for the behavioral sciences. New York, NY: Academic Press.
Tizard, B., Hughes, M., Pinkerton, G., & Carmichael, H. (1982). Adults’ cognitive demands at home and at nursery school. Journal of Child Psychology and Psychiatry, 23, 105-116.
Leave a ReplyWant to join the discussion?
Feel free to contribute!