Direction Preferring Confidence Intervals
Abstract: Confidence intervals (CIs) are instrumental in statistical analysis, providing a range estimate of the parameters. In modern statistics, selective inference is common, where only certain parameters are highlighted. However, this selective approach can bias the inference, leading some to advocate for the use of CIs over p-values. To increase the flexibility of confidence intervals, we introduce direction-preferring CIs, enabling analysts to focus on parameters trending in a particular direction. We present these types of CIs in two settings: First, when there is no selection of parameters; and second, for situations involving parameter selection, where we offer a conditional version of the direction-preferring CIs. Both of these methods build upon the foundations of Modified Pratt CIs, which rely on non-equivariant acceptance regions to achieve longer intervals in exchange for improved sign exclusions. We show that for selected parameters out of m > 1 initial parameters of interest, CIs aimed at controlling the false coverage rate, have higher power to determine the sign compared to conditional CIs. We also show that conditional confidence intervals control the marginal false coverage rate (mFCR) under any dependency.
- Gwas for discovery and replication of genetic loci associated with sudden cardiac arrest in patients with coronary artery disease. BMC cardiovascular disorders 11, 1–10.
- Multiple testing of pairs of one-sided hypotheses. Metrika 33, 121–127.
- Benjamini, Y. (2020). Selective inference: The silent killer of replicability.
- Confidence intervals with more power to determine the sign: Two ends constrain the means. Journal of the American Statistical Association 93(441), 309–317.
- Benjamini, Y. and P. B. Stark (1996). Nonequivariant simultaneous confidence intervals less likely to contain zero. Journal of the American Statistical Association, 329–337.
- False discovery rate–adjusted multiple confidence intervals for selected parameters. Journal of the American Statistical Association 100(469), 71–81.
- Bohrer, R. (1979). Multiple three-decision rules for parametric signs. Journal of the American Statistical Association 74(366a), 432–437.
- Cumming, G. (2014). The new statistics: Why and how. Psychological science 25(1), 7–29.
- Hayter, A. J. and J. C. Hsu (1994). On the relationship between stepwise decision procedures and confidence sets. Journal of the American Statistical Association 89(425), 128–136.
- Simultaneous directional inference. Journal of the Royal Statistical Society Series B: Statistical Methodology.
- On the length of post-model-selection confidence intervals conditional on polyhedral constraints. Journal of the American Statistical Association 116(534), 845–857.
- Testing statistical hypotheses, Volume 3. Springer.
- Pratt, J. W. (1961). Length of confidence intervals. Journal of the American Statistical Association 56(295), 549–567.
- Sun, W. and T. T. Cai (2007). Oracle and adaptive compound decision rules for false discovery rate control. Journal of the American Statistical Association 102(479), 901–912.
- Tukey, J. W. (1991). The philosophy of multiple comparisons. Statistical science, 100–116.
- Selection adjusted confidence intervals with more power to determine the sign. Journal of the American Statistical Association 108(501), 165–176.
- Selective sign-determining multiple confidence intervals with fcr control. Statistica Sinica 30(1), 531–555.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.