Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques

Missing categorical data presents a persistent challenge to data quality in quantitative sociological research, where simpler approaches can lead to biased estimates and incorrect conclusions. This article provides an empirically grounded evaluation of multiple imputation (MI) strategies for categor...

Full description

Saved in:
Bibliographic Details
Main Authors: Yaroslav Kostenko, Andrii Gorbachyk
Format: Article
Language:English
Published: Taras Shevchenko National University of Kyiv 2025-06-01
Series:Соціологічні студії
Subjects:
Online Access:https://sociostudios.vnu.edu.ua/index.php/socio/article/view/417
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849395282704334848
author Yaroslav Kostenko
Andrii Gorbachyk
author_facet Yaroslav Kostenko
Andrii Gorbachyk
author_sort Yaroslav Kostenko
collection DOAJ
description Missing categorical data presents a persistent challenge to data quality in quantitative sociological research, where simpler approaches can lead to biased estimates and incorrect conclusions. This article provides an empirically grounded evaluation of multiple imputation (MI) strategies for categorical survey data, specifically focusing on the complex, multi-category nominal variable "party voted for" using European Social Survey data from Sweden and Norway. We developed a simulation framework, introducing missingness under Missing Completely at Random, Missing at Random, derived from patterns of item nonresponse on auxiliary variables, and Missing Not at Random: linked to the undisclosed party choice itself. We systematically compared the performance of six imputation methods (Multinomial Logistic Regression, Random Forest, CART, KNN, Hot Deck, and Mode) across four distinct predictor set sizes, evaluating them using Accuracy, Cohen’s Kappa, and Macro F1-score with m=20 imputations. Results indicate that while imputing party choice is challenging, model-based MI techniques significantly outperform naive approaches. Multinomial Logistic Regression consistently emerged as the most robust and highest-performing method, often benefiting from larger predictor sets within the MI framework. K-Nearest Neighbors showed promise with smaller predictor sets, offering a computationally efficient alternative. The work emphasizes the importance of principled imputation and provides practical recommendations for sociologists regarding method selection, predictor set construction, and consideration of computational costs when addressing missing categorical data.
format Article
id doaj-art-e68750c1bbab43a2a5da337acdaec813
institution Kabale University
issn 2306-3971
2521-1056
language English
publishDate 2025-06-01
publisher Taras Shevchenko National University of Kyiv
record_format Article
series Соціологічні студії
spelling doaj-art-e68750c1bbab43a2a5da337acdaec8132025-08-20T03:39:40ZengTaras Shevchenko National University of KyivСоціологічні студії2306-39712521-10562025-06-011(26)8510910.29038/2306-3971-2025-01-32-32335Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation TechniquesYaroslav Kostenko0https://orcid.org/0009-0001-7878-5034Andrii Gorbachyk1https://orcid.org/0000-0003-1944-435XTaras Shevchenko National University of KyivTaras Shevchenko National University of KyivMissing categorical data presents a persistent challenge to data quality in quantitative sociological research, where simpler approaches can lead to biased estimates and incorrect conclusions. This article provides an empirically grounded evaluation of multiple imputation (MI) strategies for categorical survey data, specifically focusing on the complex, multi-category nominal variable "party voted for" using European Social Survey data from Sweden and Norway. We developed a simulation framework, introducing missingness under Missing Completely at Random, Missing at Random, derived from patterns of item nonresponse on auxiliary variables, and Missing Not at Random: linked to the undisclosed party choice itself. We systematically compared the performance of six imputation methods (Multinomial Logistic Regression, Random Forest, CART, KNN, Hot Deck, and Mode) across four distinct predictor set sizes, evaluating them using Accuracy, Cohen’s Kappa, and Macro F1-score with m=20 imputations. Results indicate that while imputing party choice is challenging, model-based MI techniques significantly outperform naive approaches. Multinomial Logistic Regression consistently emerged as the most robust and highest-performing method, often benefiting from larger predictor sets within the MI framework. K-Nearest Neighbors showed promise with smaller predictor sets, offering a computationally efficient alternative. The work emphasizes the importance of principled imputation and provides practical recommendations for sociologists regarding method selection, predictor set construction, and consideration of computational costs when addressing missing categorical data.https://sociostudios.vnu.edu.ua/index.php/socio/article/view/417data qualitymissing datadata imputationmultiple imputationlogistic regressionclustering
spellingShingle Yaroslav Kostenko
Andrii Gorbachyk
Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
Соціологічні студії
data quality
missing data
data imputation
multiple imputation
logistic regression
clustering
title Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
title_full Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
title_fullStr Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
title_full_unstemmed Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
title_short Missing Categorical Data in Sociological Surveys: An Experimental Evaluation of Imputation Techniques
title_sort missing categorical data in sociological surveys an experimental evaluation of imputation techniques
topic data quality
missing data
data imputation
multiple imputation
logistic regression
clustering
url https://sociostudios.vnu.edu.ua/index.php/socio/article/view/417
work_keys_str_mv AT yaroslavkostenko missingcategoricaldatainsociologicalsurveysanexperimentalevaluationofimputationtechniques
AT andriigorbachyk missingcategoricaldatainsociologicalsurveysanexperimentalevaluationofimputationtechniques