MCAR, MAR, and MNAR Values in the Same Dataset: A Realistic Evaluation of Methods for Handling Missing Data

Gomer, Brenna

doi:10.1080/00273171.2018.1557033

Cited by 10 publications

(8 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Multiple imputation was conducted for missing values based on other data through multiple versions of the data set, across a specified number of iterations (Gomer, 2019). Normality tests were conducted and deemed appropriate for multiple imputation.…”

Section: Methodsmentioning

confidence: 99%

“…This method provides a better estimate of missing values and standard errors than other estimation methods, as it is as unbiased as possible and can be used on any kind of data. Multiple imputation is the current ‘gold standard’ method for addressing missing data, which can account for up to 30% of data missing not at random (Gomer, 2019). A missing value analysis determined the missing data (29%) was not missing completely at random (MCAR) as little’s MCAR was significant (χ 2 (15,182, N = 599) = 15,509.75, p < .05); therefore, the data were missing either at random within subgroups of other variables in the data or as a result of a pattern within the missing data.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Across the wide prairie: Exploring fear of crime in a small Canadian municipality

Hodgkinson

Lunney

2021

Journal of Criminology

View full text Add to dashboard Cite

Although fear of crime is well-researched in urban domains, the predictors of fear of crime in non-urban contexts are less established. Using a sample of 559 people, this study aims to address this gap by evaluating the role of individual and ecological-level predictors on fear of crime in a small Canadian municipality. Key findings of this study include support for the influence of social cohesion, informal social control and social and physical disorder on fear in a small municipality. However, no clear relationship is found between gender and fear of crime. Additionally, nuanced relationships between social predictors and fear emerge that may be uniquely explained by non-metropolitan context. The findings have implications for the use of urban-based criminological theories of fear and for the use of crime prevention and fear reduction strategies in non-metropolitan contexts.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

Across the wide prairie: Exploring fear of crime in a small Canadian municipality

Hodgkinson

Lunney

2021

Journal of Criminology

View full text Add to dashboard Cite

show abstract

“…Their brief descriptions are given in Table 1, and these datasets can be downloaded from the link http://archive.ics.uci.edu/ml/index.php. For each benchmark dataset, incomplete datasets are artificially generated through missing completely at random (MCAR) mechanism [37].…”

Section: Consequence Parameter Identification and Missing Value Imput...mentioning

confidence: 99%

Joint clustering and missing value imputation for incomplete data via fuzzy modeling and alternate optimization

Wang

Pang

Zhang

et al. 2023

Preprint

View full text Add to dashboard Cite

Since missing information is an ordinary phenomenon in actual scenarios that increases the difficulty of data analysis, missing value imputation has attracted ever-growing attention in recent years, by exploiting data modeling. Particularly, missing information in engineering design and optimization is a challenging topic. In this work, an exquisite missing value imputation method based on Takagi-Sugeno (TS) fuzzy modeling is proposed, which first divides incomplete dataset by clustering into several fuzzy subsets and finally establishes global model with different regression models on each subset. Specifically, to improve the clustering performance of practical incomplete issues, we incorporate the measurement of common and uncommon observed feature subspace information of data instances into the fuzzy c-means (FCM) framework. Moreover, to better improve the model accuracy, feature selection is introduced to each fuzzy rule and update model parameters and imputations by co-training. The experiments reported both on UCI and tunnel boring machine (TBM) operation datasets demonstrate the proposed method can accurately accomplish missing value imputation of incomplete data.

show abstract

“…Missing data was de nition as ASCC patients with unknown survival months or dead with unknown reason. Further, we consider these data (n = 1422) are missing at random and employed the Multiple Imputation (MI) method via Markov Chain Monte Carlo to handle 9 . MI is a sophisticated but exible approach for handling missing data and is broadly applicable within the standard statistical software package of R 4.0.1.…”

Section: Missing Data De Nition and Handlingmentioning

confidence: 99%

Validation and Modification of the American Joint Commission on Cancer 8th Edition Staging System for Patients With Anal Squamous Cell Cancer: A Surveillance, Epidemiology and End Results Analysis

Yao

Yang

et al. 2021

Preprint

View full text Add to dashboard Cite

This study aimed to validate the clinical utility of American Joint Committee on Cancer (AJCC) 8th edition staging system for anal squamous cell cancer (ASCC). Further, Surveillance, Epidemiology, and End Results database (2004–2015) was used to validate the practicability of 8th AJCC stages. Disease-specific survival (DSS) and overall survival (OS) were evaluated using Kaplan-Meier analysis. Hazard ratios (HRs) were calculated using Cox regression analysis. Furthermore, the modification was validated using Kaplan-Meier analysis and calculating the concordance index (C-index). A total of 11891 patients with ASCC were analysed. the 5-year DSSs for stage I, IIA, IIB, IIIA, IIIB, IIIC, and IV were 93.8%, 87.4%, 76.2%, 76.9%, 72.0%, 66.1%, and 34.9%. We found the OS and DSS of IIIA patients showed better than IIB. After hazard analysis, modified IIIA patients showed worse OS and DSS than IIB. C-indexes were increased from 0.612 to 0.636 for DSS and from 0.566 to 0.582 for OS of stage II and III ASCCs. In conclusion, our study validated 8th AJCC staging system as overall suitable classification with a few defects. Thus, we suggested modifications in II and III stages to make it more suitable for ASCCs that could be adopted in clinical practice.

show abstract

MCAR, MAR, and MNAR Values in the Same Dataset: A Realistic Evaluation of Methods for Handling Missing Data

Cited by 10 publications

References 1 publication

Across the wide prairie: Exploring fear of crime in a small Canadian municipality

Across the wide prairie: Exploring fear of crime in a small Canadian municipality

Joint clustering and missing value imputation for incomplete data via fuzzy modeling and alternate optimization

Validation and Modification of the American Joint Commission on Cancer 8th Edition Staging System for Patients With Anal Squamous Cell Cancer: A Surveillance, Epidemiology and End Results Analysis

Contact Info

Product

Resources

About