MATERNAL FACTORS THAT INFLUENCE CHILDRENS POSITIVE BEHAVIOR - DEMONSTRATION OF A STRUCTURAL EQUATION ANALYSIS OF SELECTED DATA FROM THE BERKELEY GROWTH STUDY
MATERNAL FACTORS THAT INFLUENCE CHILDRENS POSITIVE BEHAVIOR - DEMONSTRATION OF A STRUCTURAL EQUATION ANALYSIS OF SELECTED DATA FROM THE BERKELEY GROWTH STUDY
复制标题
DOI：
10.2307/1130290
复制
发表时间：
1987-02-01
影响因子：
4.6
通讯作者：
MENDOZA, JL
中科院分区：
文献类型：
--
作者：
CRANO, WD;MENDOZA, JL
Experimentation can be inappropriate in research concerned with natural progressions, because the experimental method requires an artificial intervention into the ongoing processes it is designed to investigate. Nonetheless, when technical alternatives are not available, the research question sometimes is reshaped to fit the extant methodology. In such instances, the resulting data often do not speak to the issue that initially gave rise to the research. Structural equation modeling can be useful in solving the problem of lack of fit between issue and method, especially in developmental research. An analysis that makes use of data drawn from Bayley''s Berkeley Growth Study is performed for didactic purposes to illustrate the use of structural equation modeling on a child development data set. Alternatives to the standard latent factor approaches are suggested and demonstrated for use in research situations in which the subject-to-variable ratio is less than optimal. These preliminary analyses are intended to provide later confirmatory analyses with a more coherent empirical and theoretical foundation than would be possible if the data were left in their original form. The Bayley data provide a substantive context in which to consider the relevance of the proposed methodological approach for research on child development and an additional role for preliminary approximations of structural equation models in the early stages of a research project.
