2011
DOI: 10.1111/j.1741-3737.2011.00861.x
|View full text |Cite
|
Sign up to set email alerts
|

Toward Best Practices in Analyzing Datasets with Missing Data: Comparisons and Recommendations

Abstract: Although several methods have been developed to allow for the analysis of data in the presence of missing values, no clear guide exists to help family researchers in choosing among the many options and procedures available. We delineate these options and examine the sensitivity of the findings in a regression model estimated in three random samples from the National Survey of Families and Households ( n = 250 -2,000). These results, combined with findings from simulation studies, are used to guide answers to a… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

5
309
0
3

Year Published

2014
2014
2023
2023

Publication Types

Select...
9

Relationship

0
9

Authors

Journals

citations
Cited by 447 publications
(317 citation statements)
references
References 34 publications
5
309
0
3
Order By: Relevance
“…First, although the missing data concerning classroom observations was large, various simulation studies have shown little change in the findings, based on MAR assumptions, for levels of missing data to 50%; although, beyond that level, there might be differences in the estimators using different missing data strategies (van Buuren, 2010;Johnson & Young, 2011). However, the small sample size of the observed teachers is likely to have diminished the power of our statistical testing.…”
Section: Limitations and Direction For Future Researchmentioning
confidence: 99%
See 1 more Smart Citation
“…First, although the missing data concerning classroom observations was large, various simulation studies have shown little change in the findings, based on MAR assumptions, for levels of missing data to 50%; although, beyond that level, there might be differences in the estimators using different missing data strategies (van Buuren, 2010;Johnson & Young, 2011). However, the small sample size of the observed teachers is likely to have diminished the power of our statistical testing.…”
Section: Limitations and Direction For Future Researchmentioning
confidence: 99%
“…Although researchers have indicated feeling more confident imputing their data, there is still no consensus about the maximum number of missing in multilevel data that can be safely imputed or handled by using FIML (van Buuren, 2010). Previous simulation studies show little change in findings based on MAR assumptions for levels of missing data to 50%, although beyond that level, there might be differences in the estimators using different missing data strategies (Johnson & Young, 2011;van Buuren, 2010). Since the missing data were consistent with the assumption of MAR in this study, statistical analyses were carried out using the FIML, which allows all available information to be used without imputing data (Muthén & Muthén, 1998-2015.…”
Section: Aims and Hypothesesmentioning
confidence: 99%
“…Research demonstrates that this method yields less-biased results and more efficient estimates than complete case analysis or other traditional approaches (Young and Johnson 2015;Johnson and Young 2011). Datasets were generated by means of an imputation model, regressing incomplete covariates on the other covariates in the analysis (including interaction terms) and the outcome.…”
Section: Analytic Strategymentioning
confidence: 99%
“…Because the scales were highly reliable and the data were missing completely at random, there was very little loss of information (Allison 2002). We used Full Information Maximum Likelihood (FIML) in Plus (Methuen and Methuen 2007) to handle the missing data because it is one of the best methods for handling data that is missing completely at random (Johnson and Young 2011). The analytic sample includes all women for whom we had full information on the exogenous variables.…”
Section: Samplementioning
confidence: 99%