Missing data in bioarchaeology I: A review of the literature

Wissler, Amanda; Blevins, Kelly E.; Buikstra, Jane E.

doi:10.1002/ajpa.24609

Cited by 3 publications

(2 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Missing data are a common problem in forensic and archeological anthropology (Kenyhercz et al, 2019; Wissler et al, 2022a). In both fields, researchers often encounter bones and artifacts that are incomplete or fragmented.…”

Section: Introductionmentioning

confidence: 99%

Evaluation of missing data imputation methods for human osteometric measurements

Pang

Liu

2023

American Journal of Biological Anthropology

View full text Add to dashboard Cite

It is not uncommon for biological anthropologists to analyze incomplete bioarcheological or forensic skeleton specimens. As many quantitative multivariate analyses cannot handle incomplete data, missing data imputation or estimation is a common preprocessing practice for such data. Using William W. Howells' Craniometric Data Set and the Goldman Osteometric Data Set, we evaluated the performance of multiple popular statistical methods for imputing missing metric measurements. Results indicated that multiple imputation methods outperformed single imputation methods, such as Bayesian principal component analysis (BPCA). Multiple imputation with Bayesian linear regression implemented in the R package norm2, the Expectation-Maximization (EM) with Bootstrapping algorithm implemented in Amelia, and the Predictive Mean Matching (PMM) method and several of the derivative linear regression models implemented in mice, perform well regarding accuracy, robustness, and speed.Based on the findings of this study, we suggest a practical procedure for choosing appropriate imputation methods.

show abstract

Section: Introductionmentioning

confidence: 99%

Evaluation of missing data imputation methods for human osteometric measurements

Pang

Liu

2023

American Journal of Biological Anthropology

View full text Add to dashboard Cite

show abstract

“…Additionally, archaeo-historical data can suffer from the problem of forward bias (Beheim et al 2021), as the first observed instance of evidence is unlikely to be the true first occurrence, with the true first occurrence happening potentially hundreds or thousands of years prior. In bioarchaeology there is little consideration given to how missing data may bias findings, and the majority of studies do not have a stated approach for handling missing data (Wissler, Blevins, and Buikstra 2022).…”

Section: Introductionmentioning

confidence: 99%

Predicting the Past: Imputation of Historical Data

Spicer¹,

Monroe²,

Slingerland³

et al. 2023

Preprint

View full text Add to dashboard Cite

Research using historical data is becoming more common across the social sciences. However, approaches using historical data suffer from a more acute missing data problem than approaches using contemporary data. Complete datasets are required to use many common statistical and machine learning techniques, the majority of which cannot handle missing data. One approach to handling missing data is to impute the missing values. Using the Database of Religious History (DRH), a large historical database, comprised of both quantitative and qualitative data, this research compares the accuracy and bias of 10 imputation methods. k-nearest neighbors (k-NN), missForest, Generalized low rank models (GLRM), Factorial analysis for mixed data (FAMD), DataWig, and five methods implemented in the R package mice (multiple imputation by chained equations). On average, a missForest approach had higher imputation accuracy and among the lowest bias compared to the other methods. The accuracy and bias of different methods implemented in mice were highly variable, highlighting the importance of researchers stating the precise method used for imputation, rather than general phrases such as ‘data were imputed using mice’.

show abstract

Imputation methods for mixed datasets in bioarchaeology

Ryan-Despraz,

Wissler

2024

Archaeol Anthropol Sci

View full text Add to dashboard Cite

Missing data is a prevalent problem in bioarchaeological research and imputation could provide a promising solution. This work simulated missingness on a control dataset (481 samples × 41 variables) in order to explore imputation methods for mixed data (qualitative and quantitative data). The tested methods included Random Forest (RF), PCA/MCA, factorial analysis for mixed data (FAMD), hotdeck, predictive mean matching (PMM), random samples from observed values (RSOV), and a multi-method (MM) approach for the three missingness mechanisms (MCAR, MAR, and MNAR) at levels of 5%, 10%, 20%, 30%, and 40% missingness. This study also compared single imputation with an adapted multiple imputation method derived from the R package “mice”. The results showed that the adapted multiple imputation technique always outperformed single imputation for the same method. The best performing methods were most often RF and MM, and other commonly successful methods were PCA/MCA and PMM multiple imputation. Across all criteria, the amount of missingness was the most important parameter for imputation accuracy. While this study found that some imputation methods performed better than others for the control dataset, each imputation method has advantages and disadvantages. Imputation remains a promising solution for datasets containing missingness; however when making a decision it is essential to consider dataset structure and research goals.

show abstract

Missing data in bioarchaeology I: A review of the literature

Cited by 3 publications

References 42 publications

Evaluation of missing data imputation methods for human osteometric measurements

Evaluation of missing data imputation methods for human osteometric measurements

Predicting the Past: Imputation of Historical Data

Imputation methods for mixed datasets in bioarchaeology

Contact Info

Product

Resources

About