TY - JOUR
T1 - On the modeling of scaled measurement sequences
T2 - Implications for analyses of cognitive development
AU - Little, Todd D.
AU - Widaman, Keith F.
N1 - Funding Information:
During the last quarter century, cognitive developmentalists (e.g., Brainerd, 1978; Case, 1985; Fischer, 1980; Flavell, 1977; Levin, 1986; Wohlwill, 1960) have debated the relationship between patterns of data (or measurement se- Both authors contributed equally to the present paper, so they are listed in alphabetical order. The present study was supported in part by a Humanities and Social Sciences grant from the University of California at Riverside to the first author, and the following grants to the second author: Grants HD-14688, HD-21056, and HD-22953 from the National Institute of Child Health and Human Development, and an intramural grant from the Academic Senate, University of California at Riverside. Computing services were provided by the Academic Computing Center, University of California at Riverside. The authors would like to thank Carol Tomlinson-Keasey, Jerry Carlson, Richard Newman, Pierre Cormier, Carrie Saetermoe, and three anonymous reviewers for their enlightening discussions and courteous and helpful comments on earlier drafts of this manuscript. Special appreciation is extended to Debbie Retter, Mark Boyd, Mark Houser, Roberta Cox, and the faculty and staff of the Seventh Day Adventist School of Riverside for their assistance at various stages of the project.
PY - 1990
Y1 - 1990
N2 - Contrary to Brainerd's (1978) argument that measurement sequences do not provide a useful representation of hypothesized developmental sequences, the basic utility of measurement sequences as indices of age-related changes in cognition is demonstrated. Specifically, a scalable measurement sequence provides ordinal data describing a developmental phenomenon of interest. Because of this basic quality, a theoretically derived and empirically scaled task sequence can be used in robust statistical analyses, such as multiple regression or structural modeling, that enable tests of hypothesized relationships among indices of cognitive development. A scaled sequence of number tasks is presented to illustrate this argument and to test hypotheses concerning the relationship between a Piagetian measure of numerical functional relations and psychometric measures of numerical and perceptual ability. A structural equation model adequately represented the relationships among the four latent variables of age, numerical facility, perceptual speed, and numerical functional relations, demonstrating the utility of measurement sequences as indices of development. Extensions of the proposed procedures are discussed that would provide powerful tests of important theoretical questions regarding more comprehensive representations of developmental processes.
AB - Contrary to Brainerd's (1978) argument that measurement sequences do not provide a useful representation of hypothesized developmental sequences, the basic utility of measurement sequences as indices of age-related changes in cognition is demonstrated. Specifically, a scalable measurement sequence provides ordinal data describing a developmental phenomenon of interest. Because of this basic quality, a theoretically derived and empirically scaled task sequence can be used in robust statistical analyses, such as multiple regression or structural modeling, that enable tests of hypothesized relationships among indices of cognitive development. A scaled sequence of number tasks is presented to illustrate this argument and to test hypotheses concerning the relationship between a Piagetian measure of numerical functional relations and psychometric measures of numerical and perceptual ability. A structural equation model adequately represented the relationships among the four latent variables of age, numerical facility, perceptual speed, and numerical functional relations, demonstrating the utility of measurement sequences as indices of development. Extensions of the proposed procedures are discussed that would provide powerful tests of important theoretical questions regarding more comprehensive representations of developmental processes.
UR - http://www.scopus.com/inward/record.url?scp=0009364888&partnerID=8YFLogxK
U2 - 10.1016/S0160-2896(05)80015-7
DO - 10.1016/S0160-2896(05)80015-7
M3 - Article
AN - SCOPUS:0009364888
SN - 0160-2896
VL - 14
SP - 457
EP - 479
JO - Intelligence
JF - Intelligence
IS - 4
ER -