Who participated in the studies and how?

Sixty-six research reports were identified and examined using systematic procedures (see the Methods section of the Research Guide). All reports represented research that examined the links between seven factors of child care and children’s cognitive, language, and behavioural development. The reports were sorted into the factors studied: type of care, age of entry into child care, time in care, stability of care, quality of care, teacher education, and adult-child ratio.

• The majority of research reports are based on studies conducted in the United States. Many of the authors of the 47 US studies collected and analysed data together through interrelated research networks including the National Institute of Child Health and Human Development.
   

• In addition to those who collected their own data, some researchers made use of national or other large-scale surveys of child care and child development. The databases are listed with citations in the Resource section of the Research Guide. Forty-one percent of the articles used large-scale surveys.
   

• Researchers measured characteristics of children to assess the impact of child care. They measured characteristics of families and children to account for possible influences on child development over and above the factors of child care. Sometimes they created their own measuring devices but often they used already-published and previously used measuring tools. The 109 previously published measuring tools are listed with full citations in the Resource section of the Research Guide.
   

• Most researchers used previously published and standardised inventories when they wanted to assess the quality of child care conditions. Twenty-three published measures of child care quality are listed with full citations in the Resource section of the Research Guide.
   

• Most of the data were obtained from observational studies. Researchers assessed the developmental status of the children in their samples and correlated their scores with child care factors. The assessments were sometimes repeated as the child advanced from preschool to school entry and beyond. Studies conducted in this way are called, in the jargon of research design, prospective, follow-up, observations. If information about children’s child care environments was obtained after the fact, the design is called, retrospective. If children were measured only once, there are no follow-up assessments.
   

• All things being equal, it is known that well-designed studies typically yield fewer significant findings than poorly designed studies. This is because poorly designed studies may contain unknown factors that influence children’s development but are unrelated to child care. For example, if the families of children in home care had significantly lower incomes than those of children in centre care, any differences in development said to be due to child care type could be challenged as differences in the repercussions of poverty.
   

• The studies in the Research Guide were well designed and conducted. Sample sizes were large and most researchers carefully controlled for countless variables and events relating to family and child characteristics that could masquerade as outcomes of child care. The researchers used selection and statistical techniques to remove the possible influence on children’s development of variables other than the factors of child care.
   

• In the 6 studies (9% of 66 studies) that were designed as experiments, children participated in one of several types of child care conditions as a result of jurisdictional assignment. Assignment was based on availability or perceived need, and not as a result of random assignment to the child care conditions.
   

• Approximately 27,879 children participated in the studies. Sometimes children were assessed more than once and over the course of several years. Their data were sometimes reported in more than one report and by different groups of researchers. Only the number of children in the original sample was included in our calculations of 27,879. In other words, children were counted only once.

• In each study, every measurement of children’s development was identified and coded as positive, negative, or null based on statistical significance (see Methods section of the Research Guide for details) for each category of children's development: cognitive, language, and behaviour. A positive result means that that factor was related to better scores on the child development measure. For example, a positive result for the teacher education factor on a language development measure means that greater levels of teacher education are associated with higher scores on a test of children’s language development.
   

• The results from each measuring tool used in each study received only one credit (positive, negative, or null). Subscales of the measuring tool were given fractional scores so that studies in which children were measured in fine detail did not overly influence the summary of results for that factor. After all, that one study might be unique in other ways (country in which it was conducted, political climate, etc.). Across the 66 research reports there were 563 results of the relationship of child care factors to children’s development.
   

• Some reports contain several measures of the same category of development for the same children. These individual reports may also unduly influence one’s overall conclusions about the advantages or disadvantages of factors of child care simply because the researchers measured the same children many times using similar tests. Therefore, in addition to bar graphs of the 563 results, tables were constructed that describe the results of a report as a whole. Did the report present only positive results, negative results, or mixed (some positive, some negative) results, yes or no?

Go to Scorecards for the results of each report or continue on to What's better for children's development, home or centre care?