Abstract Details
Activity Number:
|
471
|
Type:
|
Invited
|
Date/Time:
|
Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
|
Sponsor:
|
ASA
|
Abstract #314130
|
View Presentation
|
Title:
|
Reproducibility and the Scientific Method
|
Author(s):
|
Philip B. Stark*+
|
Companies:
|
University of California
|
Keywords:
|
Reproducibility ;
version control ;
open science ;
big data ;
data-driven discovery
|
Abstract:
|
Assessing the scientific merit of a claim--determining whether it is real, accidental, or generalizable---requires knowing the data and a recipe to reproduce the analysis. To check a claim might require knowing details of the instrument used to collect data, the raw data, data cleaning, pre-processing, model selection, model fitting, statistical tests, scripts and code, and software package versions; settings and tuning parameters; and even the software build environment. And results can be extremely sensitive to small changes to any of those things. Computationally driven research and data intensive research have largely abandoned the scientific method, and now rely more on trusting authority than on direct evidence. It is crucial to rectify this situation by developing tools, processes, and habits that allow our collaborators, referees, and others to check, use, and extend our work. Some of these exist already, and science has much to learn from current practice in software development. I will discuss a pilot course on Reproducible and Collaborative Statistical Data Science aimed to help the next generation develop better habits than my generation of scientists learned.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2014 program
|
2014 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Professional Development program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.