<p>Missing data are common in social and clinical sciences and understanding the causes and patterns of missing data is important for selecting analysis approach and for the interpretation of the remaining data. Yet, knowledge about the factors influencing data loss is limited. Here, we assessed the contribution of genes and environments to data missingness across three experiments of infant brain and behavioural development. The sample consisted of 594 infant twins (330 monozygotic, 152 female, 178 male infants; 264 dizygotic, 132 female, 132 male infants) who were assessed with electroencephalography (EEG), pupillometry, and gaze tracking technologies at 5 months of age. Substantial familial factors (additive genetics and/or shared environment) for data missingness were found across all experiments. The amount of missing data showed only a low correlation across the experiments, suggesting a high degree of specificity in the factors contributing to missingness. The results underscore the need to adopt and improve procedural and analytical strategies that minimise data loss and its negative impacts on study conclusions.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Genetic and environmental influences on data missingness in developmental cognitive neuroscience

  • G. Bussu,
  • A. M. Portugal,
  • C. Viktorsson,
  • I. Hardiansyah,
  • T. Falck-Ytter

摘要

Missing data are common in social and clinical sciences and understanding the causes and patterns of missing data is important for selecting analysis approach and for the interpretation of the remaining data. Yet, knowledge about the factors influencing data loss is limited. Here, we assessed the contribution of genes and environments to data missingness across three experiments of infant brain and behavioural development. The sample consisted of 594 infant twins (330 monozygotic, 152 female, 178 male infants; 264 dizygotic, 132 female, 132 male infants) who were assessed with electroencephalography (EEG), pupillometry, and gaze tracking technologies at 5 months of age. Substantial familial factors (additive genetics and/or shared environment) for data missingness were found across all experiments. The amount of missing data showed only a low correlation across the experiments, suggesting a high degree of specificity in the factors contributing to missingness. The results underscore the need to adopt and improve procedural and analytical strategies that minimise data loss and its negative impacts on study conclusions.