Multiple imputation using chained equations for missing data in TIMSS: a case study

Bouhlila, Donia Smaali; Sellaouti, Fethi

doi:10.1186/2196-0739-1-4

Cited by 39 publications

(21 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…An alternative to deleting records with missing values is represented by multiple imputation methods, which have recently been extended to complex multilevel settings (Goldstein et al, 2014). In the framework of large-scale assessment data, Bouhlila and Sellaouti (2013) applied multiple imputation with chained equations to TIMSS 2007 data, while Foy and O'Dwyer (2013) applied single imputation to TIMSS 2011 data. Weirich et al (2014) conducted a simulation study to evaluate the performance of imputation methods to handle missing background variables in the IRT model for generating the plausible values.…”

Section: Model Selection and Resultsmentioning

confidence: 99%

Exploiting TIMSS and PIRLS combined data: Multivariate multilevel modelling of student achievement

Grilli¹,

Pennoni²,

Rampichini³

et al. 2016

Ann. Appl. Stat.

View full text Add to dashboard Cite

We exploit a multivariate multilevel model for the analysis of the Italian sample of the TIMSS&PIRLS 2011 Combined International Database on fourth grade students. The multivariate approach jointly considers educational achievement on Reading, Mathematics and Science, thus allowing us to test for differential associations of the covariates with the three outcomes, and to estimate the residual correlations between pairs of outcomes at student and class levels. Multilevel modelling allows us to disentangle student and contextual factors affecting achievement. We also account for territorial differences in wealth by means of an index from an external source. The model residuals point out classes with high or low performance. As educational achievement is measured by plausible values, the estimates are obtained through multiple imputation formulas. The results, while confirming the role of traditional student and contextual factors, reveal interesting patterns of achievement in Italian primary schools.

show abstract

Section: Model Selection and Resultsmentioning

confidence: 99%

Exploiting TIMSS and PIRLS combined data: Multivariate multilevel modelling of student achievement

Grilli¹,

Pennoni²,

Rampichini³

et al. 2016

Ann. Appl. Stat.

View full text Add to dashboard Cite

show abstract

“…Following MICE guidelines for the optimal overall number of imputations [34,35], we generated 20 complete data sets with all missing values imputed. The imputed data sets were then combined to produce an overall set of multiple imputation estimates for each analysis, reflecting both within- and between-imputation variance in the statistical estimates.…”

Section: Methodsmentioning

confidence: 99%

Depressive states among adults with diabetes: Findings from the National Health and Nutrition Examination Survey, 2007–2012

Albertorio-Díaz

Eberhardt

Oquendo

et al. 2017

Diabetes Research and Clinical Practice

View full text Add to dashboard Cite

Aims To determine (1) the prevalence of SubD states among adults with diabetes, and (2) whether evidence exists of an independent association between diabetes status and SubD, controlling for selected confounders. Methods Data from the 2007–2012 National Health and Nutrition Examination Surveys were combined to estimates of depressive states by diabetes status among the noninstitutionalized U.S. adult population, and to assess the association of diabetes status and depressive states using a polytomous logistic regression model. Results An estimated 17%, or 3.7 million, of U.S. adults with diabetes (diagnosed and undiagnosed) met criteria for either mD or ssD. The majority of SubD cases with diabetes were found to be ssD (10.1%) compared with mD (6.9%). After controlling for the effects of age, sex, race and ethnicity, education, body mass index, and poverty as covariates, an independent association persists between diagnosed diabetes and each SubD grouping (ssD: OR = 1.82, CIs 1.33, 2.47; mD: OR = 1.95, CIs 1.39, 2.74) compared with respondents having no diabetes. No association was found between depression and undiagnosed diabetes or prediabetes compared with those having no diabetes. Conclusion Milder forms of depression such as ssD and mD are more extant than major depressive episodes among adults with diabetes. The odds that an adult with diagnosed diabetes meets the criteria for ssD or mD are higher by 80% and 95%, respectively, after controlling for age, sex, race and ethnicity, education, body mass index, and poverty factors when compared against adults with no diabetes.

show abstract

“…MICE has several advantages over MVN such as mixed variable type (e.g. continuous, categorical), and skewed continuous variables, shown experimentally by Bouhlila et al [20]. MICE can impute when variables have different types of missingness, but not when multiple types of missingness occur within a single variable.…”

Section: Multiple Imputation (Mi)mentioning

confidence: 99%

Imputation of Missing Values in Time Series with Lagged Correlations

Rahman

Huang

Claassen

et al. 2014

2014 IEEE International Conference on Data Mining Workshop

View full text Add to dashboard Cite

Missing values are a common problem in real world data and are particularly prevalent in biomedical time series, where a patient's medical record may be split across multiple institutions or a device may briefly fail. These data are not missing completely at random, so ignoring the missing values can lead to bias and error during data mining. However, current methods for imputing missing values have yet to account for the fact that variables are correlated and that those relationships exist across time. To address this, we propose an imputation method (FLk-NN) that incorporates time lagged correlations both within and across variables by combining two imputation methods, based on an extension to k-NN and the Fourier transform. This enables imputation of missing values even when all data at a time point is missing and when there are different types of missingness both within and across variables. In comparison to other approaches on two biological datasets (simulated glucose in Type 1 diabetes and multi-modality neurological ICU monitoring) the proposed method has the highest imputation accuracy. This was true for up to half the data being missing and when consecutive missing values are a significant fraction of the overall time series length.

show abstract

Multiple imputation using chained equations for missing data in TIMSS: a case study

Cited by 39 publications

References 49 publications

Exploiting TIMSS and PIRLS combined data: Multivariate multilevel modelling of student achievement

Exploiting TIMSS and PIRLS combined data: Multivariate multilevel modelling of student achievement

Depressive states among adults with diabetes: Findings from the National Health and Nutrition Examination Survey, 2007–2012

Imputation of Missing Values in Time Series with Lagged Correlations

Contact Info

Product

Resources

About