Abstract
A regression analysis usually consists of several stages, such as variable selection, transformation and residual diagnosis. Inference is often made from the selected model without regard to the model selection methods that preceeded it. This can result in overoptimistic and biased inferences. We first characterize data-analytic actions as functions acting on regression models. We investigate the extent of the problem and test bootstrap, jackknife, and sample-splitting methods for ameliorating it. We also demonstrate an interactive LISP-STAT system for assessing the cost of the data analysis while it is taking place.
| Original language | English |
|---|---|
| Pages (from-to) | 213-229 |
| Number of pages | 17 |
| Journal | Journal of Computational and Graphical Statistics |
| Volume | 1 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 1 Jan 1992 |
Keywords
- Bootstrap
- Data splitting
- Jackknife
- Model selection
- Regression analysis
ASJC Scopus subject areas
- Statistics and Probability
- Statistics, Probability and Uncertainty
- Discrete Mathematics and Combinatorics
Cite this
On The Cost Of Data Analysis. / Faraway, Julian J.
In: Journal of Computational and Graphical Statistics, Vol. 1, No. 3, 01.01.1992, p. 213-229.Research output: Contribution to journal › Article
}
TY - JOUR
T1 - On The Cost Of Data Analysis
AU - Faraway, Julian J.
PY - 1992/1/1
Y1 - 1992/1/1
N2 - A regression analysis usually consists of several stages, such as variable selection, transformation and residual diagnosis. Inference is often made from the selected model without regard to the model selection methods that preceeded it. This can result in overoptimistic and biased inferences. We first characterize data-analytic actions as functions acting on regression models. We investigate the extent of the problem and test bootstrap, jackknife, and sample-splitting methods for ameliorating it. We also demonstrate an interactive LISP-STAT system for assessing the cost of the data analysis while it is taking place.
AB - A regression analysis usually consists of several stages, such as variable selection, transformation and residual diagnosis. Inference is often made from the selected model without regard to the model selection methods that preceeded it. This can result in overoptimistic and biased inferences. We first characterize data-analytic actions as functions acting on regression models. We investigate the extent of the problem and test bootstrap, jackknife, and sample-splitting methods for ameliorating it. We also demonstrate an interactive LISP-STAT system for assessing the cost of the data analysis while it is taking place.
KW - Bootstrap
KW - Data splitting
KW - Jackknife
KW - Model selection
KW - Regression analysis
UR - http://www.scopus.com/inward/record.url?scp=0000039381&partnerID=8YFLogxK
U2 - 10.1080/10618600.1992.10474582
DO - 10.1080/10618600.1992.10474582
M3 - Article
VL - 1
SP - 213
EP - 229
JO - Journal of Computational and Graphical Statistics
JF - Journal of Computational and Graphical Statistics
SN - 1061-8600
IS - 3
ER -