Two-tailed asymptotic inferences for a proportion. (English) Zbl 1514.62742
Summary: This paper evaluates 29 methods for obtaining a two-sided confidence interval for a binomial proportion (16 of which are new proposals) and comes to the conclusion that: Wilson’s classic method is only optimal for a confidence of 99%, although generally it can be applied when \(n \geq 50\); for a confidence of 95% or 90%, the optimal method is the one based on the arcsine transformation (when this is applied to the data incremented by 0.5), which behaves in a very similar manner to Jeffreys’ Bayesian method. A simpler option, though not so good as those just mentioned, is the classic-adjusted Wald method of Agresti and Coull.
MSC:
| 62-XX | Statistics |
Keywords:
arcsine transformation; asymptotic confidence interval; continuity correction; Jeffreys’ method; Wald method and adjusted Wald methods; Wilson score methodReferences:
| [1] | A. Agresti, Dealing with discreteness: making ‘exact’ confidence intervals for proportions, differences of proportions, and odds ratios more exact, Statist. Methods Med. Res. 12 (2003), pp. 3-21. doi: 10.1191/0962280203sm311ra · Zbl 1121.62564 · doi:10.1191/0962280203sm311ra |
| [2] | A. Agresti and B. Caffo, Simple and effective confidence intervals for proportions and difference of proportions result from adding two successes and two failures, Amer. Statist. 54(4) (2000), pp. 280-288. · Zbl 1250.62016 |
| [3] | A. Agresti and B.A. Coull, Approximate is better than “exact” for interval estimation of binomial proportions, Amer. Statist. 52(2) (1998), pp. 119-126. |
| [4] | F.J. Anscombe, The transformation of Poisson, binomial and negative-binomial data, Biometrika 35 (1948), pp. 246-254. · Zbl 0032.03702 · doi:10.1093/biomet/35.3-4.246 |
| [5] | F.J. Anscombe, On estimating binomial response relations, Biometrics 43 (1956), pp. 461-464. · Zbl 0074.14703 · doi:10.1093/biomet/43.3-4.461 |
| [6] | C.R. Blyth and H.A. Still, Binomial confidence intervals, J. Amer. Statist. Assoc. 78(381) (1983), pp. 108-116. doi: 10.1080/01621459.1983.10477938 · Zbl 0503.62028 |
| [7] | D. Böhning, Confidence interval estimation of a rate and the choice of sample size, Statist. Med. 7 (1988), pp. 865-875. doi: 10.1002/sim.4780070805 · doi:10.1002/sim.4780070805 |
| [8] | C.B. Borkowf, Constructing binomial confidence intervals with near nominal coverage by adding a single imaginary failure or success, Statist. Med. 25 (2006), 3679-3695. doi: 10.1002/sim.2469 · doi:10.1002/sim.2469 |
| [9] | L.D. Brown, T.T. Cai, and A. DasGupta, Interval estimation for a binomial proportion, Statist. Sci. 16(2) (2001), pp. 101-133. · Zbl 1059.62533 · doi:10.1214/ss/1009213286 |
| [10] | H. Chen, The accuracy of approximate intervals for a binomial parameter, J. Amer. Statist. Assoc. 85(410) (1990), pp. 514-518. doi: 10.1080/01621459.1990.10476229 · Zbl 0703.62041 |
| [11] | X. Chen, A quasi-exact method for the confidence intervals of the difference of two independent binomial proportions in small sample cases, Statist. Med. 21 (2002), 943-956. doi: 10.1002/sim.1053 · doi:10.1002/sim.1053 |
| [12] | C.J. Clopper and E.S. Pearson, The use of confidence or fiducial limits illustrated in the case of the binomial, Biometrika 26 (1934), pp. 404-413. doi: 10.1093/biomet/26.4.404 · JFM 60.1175.02 · doi:10.1093/biomet/26.4.404 |
| [13] | D.R. Cox, The continuity correction, Biometrika 57 (1970), pp. 217-219. doi: 10.1093/biomet/57.1.217 · Zbl 0193.17301 · doi:10.1093/biomet/57.1.217 |
| [14] | B.K. Ghosh, A comparison of some approximate confidence intervals for the binomial parameter, J. Amer. Statist. Assoc. 74(368) (1979), pp. 894-900. doi: 10.1080/01621459.1979.10481051 · Zbl 0429.62025 |
| [15] | Y. Guan, A generalized score confidence interval for a binomial proportion, J. Statist. Plann. Infer. 142 (2012), pp. 785-793. doi: 10.1016/j.jspi.2011.09.010 · Zbl 1428.62122 · doi:10.1016/j.jspi.2011.09.010 |
| [16] | M. Haber, A comparison of some continuity corrections for the chi-squared test on 2×2 tables, J. Amer. Statist. Assoc. 75 (1980), pp. 510-515. · Zbl 0455.62046 |
| [17] | A. Martín Andrés, M. Álvarez Hernández, and I. Herranz Tejedor, Inferences about a linear combination of proportions, Statist. Methods Med. Res. 20 (2011), pp. 369-387. doi: 10.1177/0962280209347953 · Zbl 1414.62152 · doi:10.1177/0962280209347953 |
| [18] | A. Martín Andrés and A. Silva Mato, Choosing the optimal unconditioned test for comparing two independent proportions, Comput. Statist. Data Anal. 17 (1994), pp. 555-574. doi: 10.1016/0167-9473(94)90148-1 · Zbl 0937.62534 · doi:10.1016/0167-9473(94)90148-1 |
| [19] | R.G. Newcombe, Interval estimation for the difference between independent proportions: Comparison of eleven methods, Statist. Med. 17 (1998), pp. 873-890. doi: 10.1002/(SICI)1097-0258(19980430)17:8<873::AID-SIM779>3.0.CO;2-I · doi:10.1002/(SICI)1097-0258(19980430)17:8<873::AID-SIM779>3.0.CO;2-I |
| [20] | R.M. Price and D.G. Bonett, An improved confidence interval for a linear function of binomial proportions, Comput. Statist. Data Anal. 45(3) (2004), 449-456. doi: 10.1016/S0167-9473(03)00007-0 · Zbl 1429.62102 · doi:10.1016/S0167-9473(03)00007-0 |
| [21] | D.E. Walters, An examination of the conservative nature of “classical” confidence limits for a proportion, Biometrical J. 8 (1985), pp. 851-861. doi: 10.1002/bimj.4710270804 · doi:10.1002/bimj.4710270804 |
| [22] | E.B. Wilson, Probable inference, the law of succession, and statistical inference, J. Amer. Statist. Assoc. 22 (1927), pp. 209-212. doi: 10.1080/01621459.1927.10502953 |
| [23] | B. Woolf, On estimating the relation between blood group and disease, Ann. Hum. Genetics 19(4) (1955), 251-352. doi: 10.1111/j.1469-1809.1955.tb01348.x · doi:10.1111/j.1469-1809.1955.tb01348.x |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
