[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
"Daniel Waxman" <dan@amplecat.com> |

To |
<statalist@hsphsun2.harvard.edu> |

Subject |
st: confidence intervals for ratio of predictions-- bootstrap vs. parametric methods? |

Date |
Wed, 10 Oct 2007 20:05:03 -0400 |

Statalist, A couple of months ago, Maarten Buis was kind enough to answer my questions: How to correctly use -predictnl- to calculate confidence intervals for a ratio of two adjusted predictions (i.e. relative risk) after logistic regression. To paraphrase his solution with my situation: . global mbpos_predictors _b[_cons] + _b[int_zlog_pos]*zlog + /* */ _b[int_zero_pos]*zero + /* */ _b[zlog]*zlog + _b[zero]*zero + _b[mbpos] . global mbneg_predictors _b[_cons] + _b[zlog]*zlog + _b[zero]*zero . predictnl rr=invlogit($mbpos_predictors)/invlogit($mbneg_predictors),se(se_rr) . gen ub = rr + 1.96*se_rr . gen lb = rr - 1.96*se_rr The problem is that these confidence intervals appear unreasonably wide, and the lower bound can be negative, which is nonsensical. So I did a bootstrap of the following program, and the bias-corrected bootstraps give much happier results: (note that I am bootstrapping the relative risk at a specific value of the covariates (zero=1, zlog=`1') ) . program newboot, rclass . local median_zlog=log10(`1') . global mbpos_predictors _b[_cons] + _b[int_zlog_pos]*`median_zlog' + /* *\ _b[int_zero_pos] + _b[zlog]*`median_zlog' + _b[zero] + _b[mbpos] . global mbneg_predictors _b[_cons] + _b[zlog]*`median_zlog' + _b[zero] . logistic outcome zlog zero mbpos int_zlog_pos int_zero_pos . assert e(sample)==1 . return scalar rr_posneg=invlogit($mbpos_predictors)/invlogit($mbneg_predictors) end The sample sizes range from 1500 to 6000, with an event rate of ~ 1.5% - 3.5% depending on the population. Most subjects are at low risk (i.e. the distribution of the predictors is highly skewed). Example of the different results: parametric: 3.1 (0.7 ,5.4); bc-bootstrap: 3.7 (2.2,7.4) My questions: 1. Can anybody explain why the results are so different, and whether the bias-corrected bootstraps can reasonably be thought to be much closer to the truth? 2. If reporting parametric CIs, what to do when the results get negative? 3. In one of my subpopulations, the bootstrap process returned a few red 'x's instead of dots, meaning, I think that in some samples, one of the covariate patterns didn't exist or the regression couldn't be performed. No bc-CI was calculated. Any thoughts on the real meaning of this? (are the CIs truly infinite?) Thanks. Daniel Waxman * * For searches and help try: * http://www.stata.com/support/faqs/res/findit.html * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**Re: st: calculating nearest neighbors; looping back to the beginning of observations** - Next by Date:
**Re: st: confidence intervals for ratio of predictions-- bootstrap vs. parametric methods?** - Previous by thread:
**st: Prais-Winsten regression: problem with coefficient estimates** - Next by thread:
**Re: st: confidence intervals for ratio of predictions-- bootstrap vs. parametric methods?** - Index(es):

© Copyright 1996–2016 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |