Cross‐Lagged Panel Design

University of Connecticut, Storrs, CT, USA
DOI: 10.1002/0470013192.bsa156 In book: Encyclopedia of Statistics in Behavioral Science
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The capacity for teacher expectation effects to interact and compound across a child's schooling offers a largely untested mechanism for magnifying or minimizing effects. This study examined four types of long-term teacher expectation effects: within-year effects of single teachers, cross-year effects of single teachers, mediated effects of single and multiple teachers, and compounded effects of multiple teachers. Participants were 110 students tracked from preschool through Grade 4 on measures of achievement and teacher expectations. Evidence was found for within-year but not direct cross-year effects. However, path models demonstrated enduring indirect effects of teacher expectations on cross-year achievement. Multiple years of teacher expectation effects were additive in predicting student achievement at fourth grade, with similar effects for teachers' over- and underestimates of student ability. The study extends understanding of longer-term teacher expectation effects.
    Journal of Applied Developmental Psychology 05/2014; 35(3):181–191. DOI:10.1016/j.appdev.2014.03.006 · 1.85 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Much research has focused on physical disorder in urban neighborhoods as evidence that the community does not maintain local norms and spaces. Little attention has been paid to the opposite: indicators of proactive investment in the neighborhood's upkeep. This manuscript presents a methodology that translates a database of approved building permits into an ecometric of investment by community members, establishing basic content, criteria for reliability, and construct validity. A database from Boston, MA contained 150,493 permits spanning 2.5 years, each record including the property to be modified, permit type, and date issued. Investment was operationalized as the proportion of properties in a census block group that underwent an addition or renovation, excluding larger developments involving the demolition or construction of a building. The reliability analysis found that robust measures could be generated every 6 months, and that longitudinal analysis could differentiate between trajectories across neighborhoods. The validity analysis supported two hypotheses: investment was best predicted by homeownership and median income; and maintained an independent relationship with measures of physical disorder despite controlling for demographics, implying that it captures the other end of a spectrum of neighborhood maintenance. Possible uses for the measure in research and policy are discussed.
    American Journal of Community Psychology 10/2014; 55(1-2). DOI:10.1007/s10464-014-9685-8 · 1.74 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Objective : Substance use and delinquency among adolescents has been shown to be positively associated; however, the temporal relationship is not well understood. Examining the association between delinquency and substance use is especially relevant among adolescents with a first-time substance use related offense as they are at-risk for future problems. Method : Data from 193 adolescents at time of diversion program entry and six months later was examined using cross-lagged path analysis to determine whether substance use and related consequences were associated with other types of delinquency across time. Results : Results demonstrated that delinquency at program entry was related to subsequent reports of heavy drinking and alcohol consequences, but not marijuana use or its consequences. In contrast, alcohol and marijuana use at program entry was not related to future reports of delinquency. Conclusions : Findings emphasize the need to build in comprehensive assessments and interventions for youth with a first time offense in order to prevent further escalation of substance use and criminal behaviors.
    Addictive behaviors 06/2014; DOI:10.1016/j.addbeh.2014.03.002 · 2.44 Impact Factor