Search Results
Title: Scaling Back Survey Scales: How Short is too Short?
Resulting in 1 citation.
| 1. |
Moore, Kristin Anderson Mariner, Carrie L. Halle, Tamara G. |
|
Scaling Back Survey Scales: How Short is too Short? Presented: Ann Arbor, MI, Health and Retirement Study, Institute for Social Research, Conference on Data Quality Issues in Longitudinal Surveys, October 1998 Cohort(s): Children of the NLSY79, NLSY79 Publisher: Institute for Social Research (ISR), University of Michigan Keyword(s): Attrition; Behavior Problems Index (BPI); Child Development; Data Quality/Consistency; Family Environment; Genetics; Home Observation for Measurement of Environment (HOME); Longitudinal Data Sets; Longitudinal Surveys; Marital Conflict; Marital Instability; Parent-Child Relationship/Closeness; Peabody Individual Achievement Test (PIAT- Reading); Peers/Peer influence/Peer relations; Research Methodology; Scale Construction; Test Scores/Test theory/IRT Permission to reprint the abstract has not been received from the publisher. Evolving research and theory indicate that the factors that affect child development are varied, and arise from multiple domains, including the family, the neighborhood, the school, and the peer group, in addition to genetic and physiological factors (Bronfenbrenner, 1977, 1979, 1986, 1989; Bronfenbrenner & Ceci, 1994). Even within a single domain -- the family-numerous processes have been posited and suggested to affect children's development in very different ways, some positive and some negative. For example, Day, Gavazzi, and Acock (1997) recommend multiple measures of family process constructs for inclusion in surveys, ranging from the intensity and frequency of marital conflict, to the use and creation of rituals. Similarly, Thornton (1998), summarizing the work of the NICHD Family and Child Well-being Research Network, lists dozens of candidate constructs for inclusion in studies of children and families. Collecting information on the varied factors that might potentially affe ct children's development places extraordinary data collection demands on a study. The respondent burden can become quite high, as data collection efforts can take an hour or more or even several hours. Repeated visits may be necessary at a single data collection point. Moreover, given a general consensus that longitudinal data are necessary in order to begin to address causal processes, there is a need to engage in these lengthy data collection efforts year after year. Thus, in addition to concerns about respondent burden, concerns must be addressed about the cumulative implications of very lengthy data collection studies on attrition, break-offs and respondent cooperation. |
|
|
Bibliography Citation
Moore, Kristin Anderson, Carrie L. Mariner and Tamara G. Halle. "Scaling Back Survey Scales: How Short is too Short?" Presented: Ann Arbor, MI, Health and Retirement Study, Institute for Social Research, Conference on Data Quality Issues in Longitudinal Surveys, October 1998. |