A Review of
Meaningful Differences in the Everyday Experience of Young American Children.

Reviewed by Joseph Cautilli

Meaningful Differences in the Everyday Experience of Young American Children.
Hart, B. & Risley, T. R. Baltimore: Paul H. Brooks, 1995.

Context

One of the longer lasting programs was Project Head Start. It began as a result of the Economic Opportunity Act (Zigler & Valentine, 1979). Its goal was to ameliorate the problems that may develop from living in poverty. Head Start focused not only on school interventions but on parent participation as well (Valentine & Stark, 1979).

The long-term effects of Head Start have been widely debated (e.g., Herrnstein & Murray, 1994). Supporters of Head Start and Follow Through (which followed Head Start in elementary school) point to the positive gains from the Distar reading program (Biloine, 1968; Gersten & Gutkin, 1983) or to studies showing other gains, such as Meyer's (1984) finding that of three schools exposed to Distar the immediate gains of one school persisted as children continued in school. The Pigetian Perry School Preschool project reported two long term effects: a decrease in special education placements and criminal arrests (Barrent, 1985). Herrnstein & Murray (1994) pointed to a study by Westinghouse Learning Corporation (1969) showing that all 10 effects and educational outcomes of Follow Through faded within five years and that “sleeper effects.” Finally the effects of the Perry Preschool project were smaller than initially predicted and often failed to reach statistical significance (Herrnstein & Murray, 1994; Spitz, 1986) Discussions of politics and research surrounding Head Start still continue.

One issue remains clear: despite the successes of Head Start, it failed to raise 1Q scores (Haskins, 1989; Herrnstein & Murray, 1994; Lazure & Darlington, 1982; Lee, Brookes, Gunn, Schnur, & Liaw 1990; Westinghouse Learning Corporation, 1969; Zigler & Muchenchow, 1992). While some have questioned the relevance of using 1Q (Locurto, 1991), the importance of studying IQ is that it is the single best predictor of academic achievement. An intervention that could raise IQ would lead to better prediction of educational outcome for children living in poverty. In addition, society and the educational establishment attend to IQ changes.

Review

Hart and Risley began their research against the background of the apparent failure of Head Start to raise IQ scores and the general failure of the War on Poverty to overcome the transmission of poverty across generations. They directed a Head Start program at Turner House in Kansas City, KA. Their program also failed to produce lasting IQ changes, but more importantly, the children were learning more slowly, regardless of starting level, than other preschool children from academic families. Observation and information about the children's families suggested that the children were not dissimilar from their parents and siblings (Hart & Risley, 1995, p. 12). The research effort then centered on how children living in poverty continued to develop at one rate (despite intensive intervention effort from many different approaches) while middle class children were developing at a different rate. The importance of this book stems from the authors' refusal to merely accept this difference, or to attribute it to socioeconomic or intellectual variables. They sought to learn how these differences between groups developed.

Vocabulary growth rate was chosen as a directly observable criterion measure, rather than IQ, an indirect measure. Hart and Risley (1995) state: “We wanted the children to know more, but we also wanted to see them applying that knowledge, using language to elicit information and learning opportunities from their teachers in preschool. We watched what the children were doing to guide what we were doing (p. 5).”

The choice of vocabulary growth rate rather than IQ as the measure of accumulated experience avoided certain practical pitfalls in standard longitudinal studies (see Patterson, 1993). Hart and Risley were able to observe the speech of children within a matrix of changing social conditions of which the child's behavior was a part. The project was labor-intensive but the outcome justified the effort. The authors followed 42 families from each child's first words through the next two and one half years. The data set would encompass approximately 1300 hour-long observations. They wanted to record everything that went on in the children's homes. They trained observers by making repeated measures of interobserver agreement in the daycare centers.

First, the observers recorded for 10 minutes and then compared their observations. Eventually they could observe for one half hour with 90% or more agreement among their recordings. The observation training sessions resulted in refinement and elimination of some behavioral definitions. Observers next began four months observations in two homes to establish reliability.

They obtained frequencies for each episode type for each of six speaker categories: child speaking, parent speaking to the child, parent speaking to someone other than the child, other adults speaking to the child, other children speaking to the child and others speaking to others.

Their data analysis began by examining the number of nouns and descriptive words the children used. But they did not stop there. They linked these rates to the amount of social interaction in the context of the words.

“We had seen that some parents talked much more than others; they exposed their children to many more words than did parents who talked less. We needed to see whether the greater amount of talking that these parents did actually provided their children more experience with the quality features of language and interaction, Talking a lot may be more repetitious and, as a result less rich in quality features. Fewer interactions richer in nouns, questions, responses, and approval, for example, may contribute more to children’s experience than more frequent interactions less rich in these quality features” (pp. 119-120).

Hart and Risley compared measures of speech across socioeconomic status, grouping families into welfare families, working class families, and professional families. Welfare family interactions included fewer nouns, modifiers, verbs, verb tenses, and clauses. The length and follow up of interactions were shorter. Total words and time spent talking to children was lower in the welfare families. Children in welfare families received less than a third of the language experience of the working class children. Prohibitions (don’t do ..., that is bad, etc., differed between the groups. Children of professional parents received an average of five prohibitions per hour; children of welfare parents received an average of eleven. However, the difference in amount of talk was also large, thus the ratio of prohibitions to total talk directed at the child in the utterances of welfare families was seven times the ratio of prohibitions in the utterances of the professional families. As a result, a high proportion of the language interactions "missed" by children from welfare families, compared with children from professional families, was of the positive or encouraging type.

At 36 months, the child exhibiting the smallest vocabulary of the children from professional families had a significantly larger vocabulary than any of the children from the welfare families; there was no overlap whatsoever. In addition, the rates of vocabulary growth indicated that this language gap was still getting larger every month.

Statistical analysis of data collected during this time, along with follow-up data at age 9-10, indicated that, in each group, 60% of the variance in vocabulary and IQ scores at age 3, and IQ scores at age 9, were accounted for by variables derived from the parent language experience provided in the first 3 years. Race and SES did not predict accomplishment within each group.

The authors calculated the amount of language experience that would be needed to correct the drastic differences discovered between the working class and welfare groups. Given an optimal learning environment for verbal skills, equalizing differences between working class and welfare children would require approximately 41 hours per week of language intervention and of social interaction. Programs whose goal was to be effective in achieving equity would be intensive and time consuming. This could be expensive, but the alternative would be even more costly to society. This may be a multi-generational project. Another option might be to enlist and train the families and the children (cf., Quigley, Morris, & Hammett, 1976). Such programs would draw on the family's resources. In addition, training the child to occasion more positive verbal responses from the parent might also be helpful.

Summary

I found the book very enjoyable. It highlighted one of the basic issues of our time: how to bring equity to those who live in poverty. The book will undoubtedly have a major impact on those who are in the process of studying language and language intervention. It raises many questions, a major one being “is there a way to intervene more efficiently?” In our ongoing worker shortage, we will need as many basically literate workers as possible. This book suggests a way to ensure that this workforce will be there when we call on them.

Thanks to Ken Lloyd for editorial suggestions

Return Home