Multiple Imputation Strategies for Multiple Group Structural Equation Models

Enders, Craig K.; Gottschall, Amanda C.

doi:10.1080/10705511.2011.532695

Cited by 71 publications

(57 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Missing data ranged between 1.7% and 10.2% and was handled via multiple imputation with M = 100 imputed data sets keeping two considerations in mind. First, acknowledging that our research question bears on group differences over time (Group × Time interaction), and consistent with accepted practice (Enders & Gottschall, 2011), missing data were imputed separately for the treatment and control groups. Second, consistent with research that stresses the importance of including auxiliary correlate data to better meet the assumption of missing at random (Enders, 2010; Graham, 2012), 69 additional variables were added to the imputation model.…”

Section: Methodsmentioning

confidence: 99%

Generating Attention, Inhibition, and Memory: A Pilot Randomized Trial for Preschoolers With Executive Functioning Deficits

Tamm

Epstein

Loren

et al. 2017

Journal of Clinical Child & Adolescent Psychology

View full text Add to dashboard Cite

This goal of this study was to assess the initial feasibility and efficacy of a play-based intervention targeting executive functions (EF) and parent–child relationships in preschoolers compared with an active control group. Preschoolers with EF deficits (M age = 3.7 ± 0.47, predominantly White boys) and their parents were randomized to intervention (n = 36) or active control (n = 32) conditions. Child performance on EF tasks, parent and masked teacher ratings of EF and behavior, and masked clinician ratings of severity were collected at baseline and at 3 and 6 months postbaseline. Partial eta-squared effect sizes at .02 or higher comparing performance across the two groups was considered evidence of meaningful, albeit small, intervention effects. Intervention effects were observed for parent ratings of inattention, hyperactivity/impulsivity, and number/severity of problems experienced in various home situations, teacher ratings of severity of problems in various school situations, parent and teacher ratings of overall impairment, and clinician ratings of impairment. Intervention effects for functional improvements were maintained at the 6-month follow-up. No effect of the intervention was observed on the objective EF measures, although parent ratings of emotional control were improved for children in the intervention group. An intervention utilizing play-based activities targeting EF, when administered in a structured way by parents, is a promising approach for improving behavior in preschoolers with self-regulation deficits. More work is needed to investigate potential impact on EF and to disentangle mechanisms of action. It may be that the intervention's focus on the structure and quality of parent–child interactions is a mediator of outcomes, rather than improved EFs.

show abstract

Section: Methodsmentioning

confidence: 99%

Generating Attention, Inhibition, and Memory: A Pilot Randomized Trial for Preschoolers With Executive Functioning Deficits

Tamm

Epstein

Loren

et al. 2017

Journal of Clinical Child & Adolescent Psychology

View full text Add to dashboard Cite

show abstract

“…JMI assumes multivariate normality which was upheld in this study but may be tenuous in applied research. Imputing for interactions and higher order terms and also properly centering variables can also be somewhat challenging with MI methods, particularly when attempting to specify the imputation model [10,13]. Failing to adequately account for the complexity these situations introduce into the imputation model can attenuate estimates so that estimates are biased towards 0, even if data are MCAR [10].…”

Section: Discussion and Limitationsmentioning

confidence: 99%

Missing data methods for arbitrary missingness with small samples

McNeish

2016

Journal of Applied Statistics

113

View full text Add to dashboard Cite

Missing data are a prevalent and widespread data analytic issue and previous studies have performed simulations to compare the performance of missing data methods in various contexts and for various models; however, one such context that has yet to receive much attention in the literature is the handling of missing data with small samples, particularly when the missingness is arbitrary. Prior studies have either compared methods for small samples with monotone missingness commonly found in longitudinal studies or have investigated the performance of a single method to handle arbitrary missingness with small samples but studies have yet to compare the relative performance of commonly implemented missing data methods for small samples with arbitrary missingness. This study conducts a simulation study to compare and assess the small sample performance of maximum likelihood, listwise deletion, joint multiple imputation, and fully conditional specification multiple imputation for a single-level regression model with a continuous outcome. Results showed that, provided assumptions are met, joint multiple imputation unanimously performed best of the methods examined in the conditions under study. ARTICLE HISTORY

show abstract

“…There was 2.3% missing data, so the robust FIML (i.e., full-information maximum likelihood) estimator was used for the CFA and SEM analyses (Enders & Gottschall, 2011). The CFA model included five latent constructs: the three climates, ownership, and empowerment.…”

Section: Measurement Invariancementioning

confidence: 99%