Multiple Imputation of Missing Data for Multilevel Models

Grund, Simon; Lüdtke, Oliver; Robitzsch, Alexander

doi:10.1177/1094428117703686

Cited by 167 publications

(175 citation statements)

References 62 publications

(116 reference statements)

Supporting

Mentioning

162

Contrasting

Unclassified

Order By: Relevance

“…To evaluate changes on relevant outcome measurements from baseline to discharge and discharge to follow‐up, we used multilevel models, conducted in R using the nlme package (Pinheiro, Bates, DebRoy, Sarkar, & R Core Team, ). To handle high levels of missing data, we used full information maximum likelihood estimation, which is considered a robust method for generation of model estimates with missing data (Allison, ) and generates estimates similar to multiple imputation methods (Ferro, ; Grund, Lüdtke, & Robitzsch, ; Peyre, Leplege, & Coste, ). Multilevel models included repeated measurements of dependent variables (EDE‐Q, binge eating and purging frequency, %EBW, BDI‐II, STAI‐Trait) nested within participants.…”

Section: Methodsmentioning

confidence: 99%

Naturalistic outcomes for a day‐hospital programme in a mixed diagnostic sample of adolescents with eating disorders

Reilly

Rockwell

Ramirez

et al. 2020

Euro Eating Disorders Rev

View full text Add to dashboard Cite

Despite initial data suggesting positive treatment outcomes for adolescent eating disorder day-hospital programmes (DHPs), existing studies have included limited follow-up, small samples, and a focus on restricting-type eating disorders. To address these gaps, we explored naturalistic outcomes for an adolescent eating disorders DHP. Adolescent participants (N = 265) completed measurements at treatment admission, discharge (n = 170), and various lengths of follow-up (n = 126; M follow up = 278.87 days). Results from multilevel models indicated significant increases in body weight for the anorexia nervosa group throughout treatment and maintenance of increased body weight from discharge to follow-up. In bulimic spectrum disorders, binge eating and purging significantly decreased from intake to discharge and did not change from discharge to follow-up. Across the entire sample, eating disorder symptoms decreased from intake to discharge and did not change from discharge to follow-up. Further, anxiety and depression decreased over the course of treatment and continued to decrease over the follow-up period. The current investigation represents the first study to explore longitudinal DHP outcomes within adolescent bulimic spectrum eating disorders. Our findings also highlight many challenges inherent in conducting naturalistic research; it is critical that the field continue to develop solutions to the barriers inherent in conducting longitudinal research on eating disorder treatment. K E Y W O R D Sadolescent, day-hospital care, eating disorders, treatment

show abstract

Section: Methodsmentioning

confidence: 99%

Naturalistic outcomes for a day‐hospital programme in a mixed diagnostic sample of adolescents with eating disorders

Reilly

Rockwell

Ramirez

et al. 2020

Euro Eating Disorders Rev

View full text Add to dashboard Cite

show abstract

“…Recent multilevel research demonstrated that multiple imputation is preferable over listwise deletion (Grund, Lüdtke and Robitzsch, 2016, 2018). We applied a reversed multiple imputation procedure for missing level‐1 data with the mice package (van Buuren and Groothuis‐Oudshoorn, 2011), taking into account the multilevel structure (Grund, Lüdtke and Robitzsch, 2018). Following Graham, Olchowski and Gilreath (2007), 20 datasets were imputed.…”

Section: Methodsmentioning

confidence: 99%

Managers’ Resources for Authentic Leadership – a Multi‐study Exploration of Positive Psychological Capacities and Ethical Organizational Climates

Fladerer¹,

Braun

2020

British J of Management

View full text Add to dashboard Cite

While authentic leadership is highly valued in today's business world, managers do not necessarily have the resources to attain it. Building on conservation of resources theory, we propose a conceptual model to address how personal and contextual resources predict authentic leadership. Study 1 analyses the day-to-day variability in managers' positive psychological capacities as personal resources in relation to changes in authentic leadership. In addition, it tests ethical organizational climates as stable, contextual resources for authentic leadership. In Study 2, we replicate our results on the between-person level and extend the research model by exploring promotion focus as a link in the relationship between personal resources and authentic leadership.Evidence from an experience sampling study with 89 managers surveyed daily on 10 consecutive working days (Study 1) and a field survey of 130 managers at two points in time (Study 2) supports the hypothesized role of personal resources and promotion focus for authentic leadership. In both studies, only principled but not benevolent ethical organizational climates emerged as a contextual resource for authentic leadership. We discuss the implications for current management research and practice.

show abstract

“…Finally, in line with other published studies, we may not have done enough to minimize the proportion of nonresponses. A recent study proposed advanced solutions for missing data to maximize the use of data in analysis [70]. We suggest that future studies could apply for processing missing data by S Grund (2018) [70] to achieve a better quality of research design and a robust conclusion.…”

Section: Limitations and Future Studymentioning

confidence: 98%

Work–Family Conflict on Sustainable Creative Performance: Job Crafting as a Mediator

2020

View full text Add to dashboard Cite

The psychology of sustainability in terms of sustainable development for work–family conflict (WFC) in organizations has become increasingly relevant in sustainable human resource management research, and the pursuit of sustainable creative performance is an important aspect of corporate sustainable development. Taking a regulatory focus perspective, this study integrates a moderated-mediation model to examine the relationship between work–family conflict and sustainable creative performance. Data were collected from 203 supervisor–subordinate dyads from two branches of a high-end manufacturing enterprise in Shanghai, China. The results reveal that work–family conflict has a positive effect on sustainable creative performance through job crafting, and the effect is stronger when individuals show higher promotion regulatory focus. We discuss the theoretical and practical implications of the study, along with potential future research directions.

show abstract

Multiple Imputation of Missing Data for Multilevel Models

Cited by 167 publications

References 62 publications

Naturalistic outcomes for a day‐hospital programme in a mixed diagnostic sample of adolescents with eating disorders

Naturalistic outcomes for a day‐hospital programme in a mixed diagnostic sample of adolescents with eating disorders

Managers’ Resources for Authentic Leadership – a Multi‐study Exploration of Positive Psychological Capacities and Ethical Organizational Climates

Work–Family Conflict on Sustainable Creative Performance: Job Crafting as a Mediator

Contact Info

Product

Resources

About