Skip to main content

Open Access Missing Data: An Introductory Conceptual Overview for the Novice Researcher Les données manquantes: introduction aux concepts de base à l'intention du chercheur novice

Download Article:
 Download
(PDF 158.4 kb)
 

Abstract:

Missing data is a common issue in research that, if improperly handled, can lead to inaccurate conclusions about populations. A variety of statistical techniques are available to treat missing data. Some of these are simple while others are conceptually and mathematically complex. The purpose of this paper is to provide the novice researcher with an introductory conceptual overview of the issue of missing data. The authors discuss patterns of missing data, common missing-data handling techniques, and issues associated with missing data. Techniques discussed include listwise deletion, pairwise deletion, case mean substitution, sample mean substitution, group mean substitution, regression imputation, and estimation maximization.

French
Les données manquantes posent un problème fréquent en recherche; s'il n'est pas traité correctement, il peut fausser les conclusions concernant une population. Il existe un ensemble de méthodes statistiques permettant d'interpréter les données manquantes, certaines simples, et d'autres complexes, sur le plan théorique et mathématique. Le présent article propose une vue d'ensemble du problème des données manquantes à l'intention des chercheurs débutants. Les auteurs expliquent les modèles de données manquantes, discutent des questions qu'elles soulèvent et présentent certaines méthodes de traitement courantes. Parmi les techniques abordées, on compte la suppression dans la liste (listwise suppression), la suppression par paires (pairwise suppression), la substitution moyenne par cas, par échantillon ou par groupe, l'imputation par régression et la maximisation de l'estimation.

Keywords: CASE MEAN SUBSTITUTION; DELETION; GROUP MEAN SUBSTITUTION; IMPUTATION; MISSING DATA; PATTERNS OF MISSINGNESS

Document Type: Research Article

Publication date: December 1, 2005

More about this publication?
  • CJNR is a peer-reviewed, quarterly journal published by the McGill University School of Nursing since 1969. With world-wide circulation, CJNR's primary mandate is to publish original nursing research that develops basic knowledge for the discipline and examines the application of the knowledge in practice. Research related to education and history is also welcomed, as are methodological, theoretical, and review papers that advance nursing science. Letters or commentaries about published articles are encouraged. Learn more.
  • Editorial Board
  • Information for Authors
  • Subscribe to this Title
  • Archives
  • Overview / Permissions
  • ingentaconnect is not responsible for the content or availability of external websites
mcgill/cjnr/2005/00000037/00000004/art00010
dcterms_title,dcterms_description,pub_keyword
6
5
20
40
5

Access Key

Free Content
Free content
New Content
New content
Open Access Content
Open access content
Subscribed Content
Subscribed content
Free Trial Content
Free trial content
Cookie Policy
X
Cookie Policy
ingentaconnect website makes use of cookies so as to keep track of data that you have filled in. I am Happy with this Find out more