Online Program Home
My Program

Abstract Details

Activity Number: 448
Type: Contributed
Date/Time: Tuesday, August 2, 2016 : 2:00 PM to 2:45 PM
Sponsor: Biometrics Section
Abstract #321698
Title: A Conceptual Framework for Initial Data Analysis
Author(s): Marianne Huebner* and Saskia le Cessie and Werner Vach
Companies: Michigan State University and Leiden University Medical Center and University of Freiburg Medical Center
Keywords: initial data analysis ; data screening ; reproducibility

Scientists perform initial data analyses (IDA) as part of their research studies, often informally and unstructured. For some researchers this may mean data cleaning, or basic data summaries, or explorative data analysis. Systematically checking for errors or getting an understanding of the underlying features of the data are necessary steps. How to perform IDA in a structured and strategic way needs to be discussed and reporting guidelines need to be developed. We suggest that IDA can be seen as a process of five steps: 1.Data cleaning 2.Data screening that consists of understanding the properties of the data 3.Initial data reporting that informs all potential collaborators working with the data in future about insights 4.Refining and updating the analysis plan that translates the relevant findings into adaptations to the analysis plan 5.Reporting of IDA in research papers that document steps that impact the interpretation of results Nontransparent changes or data driven hypotheses are to be avoided in IDA. Modern challenges for IDA are size and complexity of datasets, including data from different sources, data that were collected for administrative instead of research purposes.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2016 program

Copyright © American Statistical Association