Details
| Original language | English |
|---|---|
| Article number | e202300197 |
| Number of pages | 15 |
| Journal | Biometrical journal |
| Volume | 66 |
| Issue number | 5 |
| Publication status | Published - 2 Jul 2024 |
Abstract
In biomedical research, the simultaneous inference of multiple binary endpoints may be of interest. In such cases, an appropriate multiplicity adjustment is required that controls the family-wise error rate, which represents the probability of making incorrect test decisions. In this paper, we investigate two approaches that perform single-step (Formula presented.) -value adjustments that also take into account the possible correlation between endpoints. A rather novel and flexible approach known as multiple marginal models is considered, which is based on stacking of the parameter estimates of the marginal models and deriving their joint asymptotic distribution. We also investigate a nonparametric vector-based resampling approach, and we compare both approaches with the Bonferroni method by examining the family-wise error rate and power for different parameter settings, including low proportions and small sample sizes. The results show that the resampling-based approach consistently outperforms the other methods in terms of power, while still controlling the family-wise error rate. The multiple marginal models approach, on the other hand, shows a more conservative behavior. However, it offers more versatility in application, allowing for more complex models or straightforward computation of simultaneous confidence intervals. The practical application of the methods is demonstrated using a toxicological dataset from the National Toxicology Program.
Keywords
- bootstrap, family-wise error rate, generalized linear models, multiple comparisons, power
ASJC Scopus subject areas
- Mathematics(all)
- Statistics and Probability
- Decision Sciences(all)
- Statistics, Probability and Uncertainty
Cite this
- Standard
- Harvard
- Apa
- Vancouver
- BibTeX
- RIS
In: Biometrical journal, Vol. 66, No. 5, e202300197, 02.07.2024.
Research output: Contribution to journal › Article › Research › peer review
}
TY - JOUR
T1 - Simultaneous Inference of Multiple Binary Endpoints in Biomedical Research
T2 - Small Sample Properties of Multiple Marginal Models and a Resampling Approach
AU - Budig, Sören
AU - Jung, Klaus
AU - Hasler, Mario
AU - Schaarschmidt, Frank
N1 - Publisher Copyright: © 2024 The Author(s). Biometrical Journal published by Wiley-VCH GmbH.
PY - 2024/7/2
Y1 - 2024/7/2
N2 - In biomedical research, the simultaneous inference of multiple binary endpoints may be of interest. In such cases, an appropriate multiplicity adjustment is required that controls the family-wise error rate, which represents the probability of making incorrect test decisions. In this paper, we investigate two approaches that perform single-step (Formula presented.) -value adjustments that also take into account the possible correlation between endpoints. A rather novel and flexible approach known as multiple marginal models is considered, which is based on stacking of the parameter estimates of the marginal models and deriving their joint asymptotic distribution. We also investigate a nonparametric vector-based resampling approach, and we compare both approaches with the Bonferroni method by examining the family-wise error rate and power for different parameter settings, including low proportions and small sample sizes. The results show that the resampling-based approach consistently outperforms the other methods in terms of power, while still controlling the family-wise error rate. The multiple marginal models approach, on the other hand, shows a more conservative behavior. However, it offers more versatility in application, allowing for more complex models or straightforward computation of simultaneous confidence intervals. The practical application of the methods is demonstrated using a toxicological dataset from the National Toxicology Program.
AB - In biomedical research, the simultaneous inference of multiple binary endpoints may be of interest. In such cases, an appropriate multiplicity adjustment is required that controls the family-wise error rate, which represents the probability of making incorrect test decisions. In this paper, we investigate two approaches that perform single-step (Formula presented.) -value adjustments that also take into account the possible correlation between endpoints. A rather novel and flexible approach known as multiple marginal models is considered, which is based on stacking of the parameter estimates of the marginal models and deriving their joint asymptotic distribution. We also investigate a nonparametric vector-based resampling approach, and we compare both approaches with the Bonferroni method by examining the family-wise error rate and power for different parameter settings, including low proportions and small sample sizes. The results show that the resampling-based approach consistently outperforms the other methods in terms of power, while still controlling the family-wise error rate. The multiple marginal models approach, on the other hand, shows a more conservative behavior. However, it offers more versatility in application, allowing for more complex models or straightforward computation of simultaneous confidence intervals. The practical application of the methods is demonstrated using a toxicological dataset from the National Toxicology Program.
KW - bootstrap
KW - family-wise error rate
KW - generalized linear models
KW - multiple comparisons
KW - power
UR - http://www.scopus.com/inward/record.url?scp=85197241471&partnerID=8YFLogxK
U2 - 10.1002/bimj.202300197
DO - 10.1002/bimj.202300197
M3 - Article
C2 - 38953619
AN - SCOPUS:85197241471
VL - 66
JO - Biometrical journal
JF - Biometrical journal
SN - 0323-3847
IS - 5
M1 - e202300197
ER -